Recent #GPU news in the semiconductor industry
12/28/2024, 01:37 PM UTC
小米投资10,000-GPU集群以扩大人工智能能力[News] Xiaomi to Scale AI Capabilities with New 10,000-GPU Cluster Investment
➀ 小米正在大力投资建设10,000个GPU集群，以增强其人工智能能力；➁ 公司于2023年4月成立了人工智能大模型团队；➂ 团队由一位CEO领导，拥有1,200名研发人员。➀ Xiaomi is investing heavily in building a 10,000 GPU cluster to enhance its AI capabilities; ➁ The company has established an AI large model team in April 2023; ➂ The team is led by a CEO with 1,200 R&D personnel.
12/26/2024, 10:04 PM UTC
Clevo新一代X580游戏笔记本曝光：18英寸显示屏，Arrow Lake-HX处理器，RTX 50系列笔记本GPUClevo's next-gen X580 laptop teased: 18-inch display, Arrow Lake-HX CPU, RTX 50 Laptop GPU
➀ Clevo即将推出配备18英寸显示屏、Arrow Lake-HX处理器和RTX 50系列GPU的X580游戏笔记本；➁ 笔记本支持高达192GB的RAM、Gen5 SSD，并包含两个Thunderbolt 5端口；➂ 提供4K和QHD显示选项，重量约为3.6公斤。➀ Clevo is set to launch the X580 gaming laptop with an 18-inch display, Arrow Lake-HX processor, and RTX 50 series GPU; ➁ The laptop supports up to 192GB RAM, Gen5 SSD, and includes two Thunderbolt 5 ports; ➂ It offers 4K and QHD display options and weighs approximately 3.6kg.
12/26/2024, 01:31 PM UTC
未发布的RTX 5090显卡图片曝光：搭载巨大GPU芯片，配备32GB GDDR7内存Nvidia's unreleased RTX 5090 pictured with huge GPU die, sports 32GB of GDDR7 memory
➀ 已泄露的RTX 5090显卡图片显示，其搭载大尺寸GPU芯片和32GB GDDR7内存；➁ 这款GPU芯片可能是自2018年以来NVIDIA最大的消费级GPU，尺寸为744平方毫米；➂ 预计RTX 5090将配备21,760个CUDA核心和32GB内存。➀ An image of the upcoming RTX 5090 GPU has been leaked, featuring a large GPU die and 32GB of GDDR7 memory; ➁ The GPU die is rumored to be Nvidia's largest consumer-grade GPU since 2018, with a size of 744mm-squared; ➂ The RTX 5090 is expected to feature 21,760 CUDA cores and 32GB of memory.
12/26/2024, 12:27 PM UTC
AMD新一代Radeon RX 9070 XT曝光：旗舰RDNA 4 GPU提升至3.1GHz，功率260-330WAMD's next-gen Radeon RX 9070 XT leaks: flagship RDNA 4 GPU boosts to 3.1GHz, 260-330W power
➀ AMD新一代Radeon RX 9070 XT显卡采用RDNA 4架构，传闻其加速频率可达3.1GHz，功耗在260W到330W之间；➁ 预计在RT游戏性能上比RX 7900 XTX提升45%，并将于CES 2025上发布；➂ 基础GPU频率预计为2.8GHz，旗舰RDNA 4芯片的加速频率可达3.0GHz至3.1GHz。➀ AMD's next-gen Radeon RX 9070 XT graphics card with RDNA 4 architecture is rumored to have a boost clock of up to 3.1GHz and a TBP ranging from 260W to 330W; ➁ The card is expected to outperform the RX 7900 XTX by 45% in RT gaming and will be unveiled at CES 2025; ➂ The base GPU clock is expected to be 2.8GHz with boost clocks up to 3.0GHz to 3.1GHz.
12/26/2024, 11:03 AM UTC
AMD和NVIDIA在1月20日新关税前将新一代游戏GPU从中国运往美国AMD and NVIDIA ship next-gen gaming GPUs from China to US before Jan 20 to escape tariffs
➀ AMD和NVIDIA在新关税于2025年1月20日生效前，将最新的GeForce RTX 50和Radeon RX 9000系列GPU运往美国。➁ 两家公司将在CES 2025上展示它们的新GPU。➂ NVIDIA计划发布多个RTX 50系列型号，而AMD将推出RDNA 4架构。➀ NVIDIA and AMD are shipping their latest GeForce RTX 50 and Radeon RX 9000 GPUs to the US before the new tariffs on January 20, 2025. ➁ Both companies will showcase their new GPUs at CES 2025. ➃ NVIDIA plans to release multiple RTX 50 series models, while AMD introduces the RDNA 4 architecture.
12/26/2024, 10:14 AM UTC
NVIDIA下一代GeForce RTX 5090 PCB再次泄露：这次搭载了三星28Gbps GDDR7芯片NVIDIA's next-gen GeForce RTX 5090 PCB leaked again: this time with Samsung 28Gbps GDDR7 chips
➀ 泄露的NVIDIA GeForce RTX 5090 PCB展示了GB202 GPU和三星28Gbps GDDR7内存模块；➁ 该卡预计将配备32GB GDDR7内存，通过512位接口提供1792GB/秒的带宽；➂ PCB包括28个VRM和一个16针电源连接器。➀ A leak reveals NVIDIA's GeForce RTX 5090 PCB featuring the GB202 GPU and Samsung's 28Gbps GDDR7 memory modules; ➁ The card is expected to have 32GB of GDDR7 memory on a 512-bit interface, offering 1792GB/sec bandwidth; ➂ The PCB includes 28 VRMs and a single 16-pin power connector.
12/25/2024, 03:01 PM UTC
苹果下一代M5系列处理器曝光：TSMC N3P节点，服务器级SoIC先进封装Apples next-gen M5 series processor leaks: TSMC N3P, server-grade SoIC advanced packaging
➀ 苹果下一代M5系列处理器正在使用TSMC的N3P节点进行原型设计；➁ 预计M5、M5 Pro、M5 Max和M5 Ultra将在2025年和2026年实现量产；➂ 采用服务器级SoIC封装和独立的CPU和GPU设计，以提高产量和热性能。➀ Apple's next-gen M5 series processors are in the prototype phase using TSMC's N3P node; ➁ Mass production expected in 2025 and 2026 for M5, M5 Pro, M5 Max, and M5 Ultra; ➂ Server-grade SoIC packaging and separate CPU and GPU designs for improved yields and thermal performance.
12/25/2024, 01:01 PM UTC
NVIDIA RTX 5070 Ti和RTX 5070：规格泄露，性能分析Nvidia's RTX 5070 Ti and RTX 5070 allegedly sport 16GB and 12GB of GDDR7 memory, respectively — Up to 8960 CUDA cores, 256-bit memory bus, and 300W TDP
➀ 据传闻，NVIDIA的RTX 5070 Ti和RTX 5070分别配备16GB和12GB的GDDR7内存；➁ RTX 5070 Ti拥有8960个CUDA核心和300W的TDP，而RTX 5070则有6144个CUDA核心和250W的TDP；➂ 三星预计明年年初开始大规模生产GDDR7内存。➀ Nvidia's RTX 5070 Ti and RTX 5070 are rumored to have 16GB and 12GB of GDDR7 memory, respectively; ➁ The RTX 5070 Ti features 8960 CUDA cores and a 300W TDP, while the RTX 5070 has 6144 CUDA cores and a 250W TDP; ➂ Samsung is expected to start mass production of GDDR7 memory early next year.
12/25/2024, 12:40 PM UTC
泄露的Nvidia RTX 5090 PCB设计图揭示巨大GPU封装和高性能特性Leaked Nvidia RTX 5090 PCB design reveals massive GPU package and high-performance features
➀ 泄露的Nvidia RTX 5090 PCB设计图揭示了巨大的GPU封装和高性能特性；➁ 预计将配备32GB下一代GDDR7内存；➂ 包含一个单12V-2x6电源连接器，可提供高达600W的电力➀ Leaked PCB design of Nvidia RTX 5090 reveals a massive GPU package and high-performance features; ➁ Expected to feature 32GB of next-generation GDDR7 memory; ➂ Includes a single 12V-2x6 power connector capable of delivering up to 600W of power
12/24/2024, 02:00 PM UTC
如果我是英特尔CEO，我会怎么做？What would you do if you were the CEO of Intel?
➀ 英特尔在技术宣布中转向更加透明的策略；➁ 英特尔PowerVia与台积电Super Power Rail之间的竞争；➂ 安迪·格鲁夫保持‘适度忧虑’的哲学。➀ Intel's shift to a more transparent strategy in technology announcements; ➁ The competition between Intel's PowerVia and TSMC's Super Power Rail; ➂ Andy Grove's philosophy of maintaining a 'healthy amount of paranoia'.
12/24/2024, 12:58 PM UTC
CES 2025前瞻：搭载英特尔Arrow Lake HX和NVIDIA GeForce RTX 5080的HP Omen 16 Max游戏笔记本泄露Nvidia RTX 5080 mobile powered HP Omen 16 Max leaks ahead of CES 2025 — listing mentions a tasty new Intel Arrow Lake HX CPU, too
➀ CES 2025前夕，HP即将推出的Omen 16 Max游戏笔记本泄露，搭载英特尔Arrow Lake HX处理器和NVIDIA GeForce RTX 5080移动GPU；➁ 英特尔Core Ultra 9 275HX处理器预计将成为英特尔即将推出的产品线的一部分，拥有八个P核心和十六个E核心；➂ 笔记本预计将配备16英寸显示屏、32GB DDR5-5600 RAM和1TB PCIe Gen 4 NVMe SSD。➀ HP's upcoming Omen 16 Max gaming laptop leaks ahead of CES 2025, featuring Intel's Arrow Lake HX processor and Nvidia's GeForce RTX 5080 mobile GPU; ➁ The Intel Core Ultra 9 275HX processor is set to be part of Intel's upcoming lineup, with eight P-cores and 16 E-cores; ➂ The laptop is expected to have a 16-inch display, 32GB of DDR5-5600 RAM, and a 1TB PCIe Gen 4 NVMe SSD.
12/23/2024, 04:00 PM UTC
Nvidia RTX 50笔记本电脑GPU规格曝光：16GB VRAM和最高175W TDPNvidia RTX 50 laptop GPU specs claimed to include 16GB VRAM and up to 175W TDP
➀ 消息泄露显示，即将推出 的Nvidia RTX 50系列笔记本电脑GPU将配备16GB VRAM，最高功耗达175W；➁ 采用RTX 50系列GPU的华硕ROG笔记本电脑将由AMD和Intel的CPU提供动力；➂ RTX 5090笔记本电脑GPU仅配备16GB GDDR7 VRAM，是其桌面版容量的一半。➀ Leaks suggest the upcoming Nvidia RTX 50 series laptop GPUs will feature 16GB VRAM and up to 175W TDP; ➁ Asus ROG laptops with RTX 50 series GPUs will be powered by both AMD and Intel CPUs; ➂ The RTX 5090 Laptop GPU is listed with only 16 GB of GDDR7 VRAM, half the capacity of its desktop counterpart.
12/23/2024, 01:02 AM UTC
埃德嗅到欧洲的机会Ed Sniffs Euro-Wonga
➀ 埃德认为欧盟的‘ECS经纪’是一个鼓励英国技术人才获取欧洲资金的方法；➁ 他已经说服部门为研究人员、发明家、初创企业和中小企业参加这些活动提供预算；➂ 他正在探索斯肯索普量子利用欧洲发展基金的机会。➀ Ed sees the EU's 'ECS Brokerage' as a way to encourage UK techies to access European funds; ➁ He has convinced the Department to allocate a budget for researchers, inventors, startups, and SMEs to attend these events; ➂ He is exploring opportunities for Scunthorpe Quantum to benefit from European development funds.
12/22/2024, 01:25 PM UTC
英特尔显卡驱动更新解决《战法师》错误 - Arc B580修复和已知问题成为最新更新的亮点Intel Game On Graphics Driver addresses Battlemage bloopers – Arc B580 fixes and known issues are the highlights of the latest update
➀ 英特尔发布更新驱动程序，解决《战法师》游戏的问题；➁ 包含对五款游戏的修复和一系列已知问题；➂ 英特尔正在解决流行软件和多个GPU配置中的已知问题。➀ Intel releases an updated driver addressing issues with the Battlemage game; ➁ Fixes for five games and a list of known issues are highlighted; ➂ Intel is working on fixes for known issues in popular software and multi-GPU configs.
12/21/2024, 03:16 AM UTC
NVIDIA GeForce RTX 5080笔记本电脑GPU泄露：比RTX 4080笔记本电脑GPU快40-60%NVIDIA GeForce RTX 5080 Laptop GPU leaks: 40-60% faster than RTX 4080 Laptop GPU
➀ 据传闻，NVIDIA的GeForce RTX 5080笔记本电脑GPU比RTX 4080快40-60%；➁ 该GPU预计将配备16GB的GDDR7内存、175W的最大TGP和7680个CUDA核心；➂ 新GPU预计将在CES 2025年亮相。➀ NVIDIA's GeForce RTX 5080 Laptop GPU is rumored to be 40-60% faster than the RTX 4080; ➁ The GPU is expected to feature 16GB of GDDR7 memory, a 175W max TGP, and 7680 CUDA cores; ➂ The new GPU is rumored to debut at CES 2025.
12/20/2024, 01:33 PM UTC
ASIC会取代GPU吗？Will ASIC Replace GPU?
<p>➀ 资本市场上ASIC的崛起正在挑战GPU在AI计算中的主导地位；</p><p>➁ ASIC和GPU都是用于计算功能的半导体芯片，ASIC专注于特定任务；</p><p>➂ ASIC的性能和效率与任务算法高度匹配，使其比通用芯片如GPU更可靠、更节能；</p><p>➃ 由于性能、生态系统和集成能力强大，NVIDIA 的GPU占据了市场主导地位，但ASIC的兴起为AI计算提供了多样化的机会。</p><p>➀ The rise of ASICs in the capital market is challenging the dominance of GPUs in AI computing;</p><p>➁ ASICs and GPUs are both semiconductor chips used for computing, with ASICs being dedicated to specific tasks;</p><p>➂ The performance and efficiency of ASICs are highly matched to the task algorithms, making them more reliable and energy-efficient than general-purpose chips like GPUs;</p><p>➃ NVIDIA's GPUs have a strong market position due to their performance, ecosystem, and integration capabilities, but the rise of ASICs presents opportunities for diversification in AI computing.</p>
12/19/2024, 08:39 PM UTC
CXL技术终于在2025年到来CXL is Finally Coming in 2025
➀ 计算表达式链接（CXL）技术预计将在2025年从利基市场转向主流应用；➁ CXL对内存扩展的支持是主要驱动力，现在有各种服务器和内存解决方案可用；➂ CXL 2.0和未来的PCIe/CXL版本将实现更高级的应用，如交换和动态内存分配。➀ Compute Express Link (CXL) technology is expected to move from a niche to mainstream use in 2025; ➁ CXL's support for memory expansion is a significant driver, with various server and memory solutions now available; ➂ CXL 2.0 and future generations of PCIe/CXL will enable more advanced use cases like switching and dynamic memory allocation.
12/19/2024, 11:56 AM UTC
英国公司向中国GPU设计师提供关键技术：报告称Chinese GPU designers received key technologies from British company Imagination Technology: Report
➀ 英国公司Imagination Technologies向中国公司提供关键GPU技术；➁ 英国中国透明度组织指责Imagination的高管促成了技术转移；➂ Imagination Technologies否认任何不当行为，并声称遵循行业标准。➀ British company Imagination Technologies supplies key GPU technologies to Chinese companies; ➁ UKCT accuses Imagination's managers of facilitating technology transfer; ➂ Imagination Technologies denies any wrongdoing and maintains industry standards were followed.
12/18/2024, 01:57 PM UTC
英特尔Arc B580显卡需求旺盛，零售商每周补货——249美元的新GPU冠军供不应求High demand for Intel's Arc B580 as retailers receive weekly restocks — demand outstrips supply for the new $249 GPU champion
➀ 英特尔Arc B580显卡因零售商每周补货而需求旺盛；➁ 249美元的GPU冠军供不应求；➂ 在英特尔近期挑战之后，Arc B580的成功是一个受欢迎的转变。➀ Intel's Arc B580 graphics card is in high demand as retailers receive weekly restocks; ➁ The demand outstrips supply for the $249 GPU champion; ➂ The success of the Arc B580 is a welcome change for Intel after recent challenges.
12/17/2024, 09:11 PM UTC
NVIDIA新一代GeForce RTX 5080显卡：搭载更快30Gbps GDDR7内存，功耗400WNVIDIA's new GeForce RTX 5080 expected to feature faster 30Gbps GDDR7 memory, 400W power
➀ 预计NVIDIA的GeForce RTX 5080将配备16GB GDDR7内存，提供高达960GB/s的带宽和400W的功耗；➁ 该显卡在光线追踪和4K游戏方面承诺将带来显著性能提升；➂ 新的GeForce RTX 5080传闻拥有10,752个CUDA核心，将比RTX 4090提供重大升级。➀ NVIDIA's GeForce RTX 5080 is expected to feature 16GB GDDR7 memory, offering up to 960GB/sec bandwidth and 400W power consumption; ➁ The GPU promises significant performance improvements, especially in ray tracing and 4K gaming; ➂ The new GeForce RTX 5080 is rumored to have 10,752 CUDA cores and will offer a significant upgrade over the RTX 4090.
12/17/2024, 04:41 PM UTC
NVIDIA RTX 5080 GPU可能配备比RTX 5090更快的VRAM，但16GB容量对未来保障似乎不足NVIDIA RTX 5080 GPU may get faster VRAM than RTX 5090, but 16GB looks shaky for future-proofing
➀ RTX 5080可能配备16GB的30Gbps GDDR7 VRAM，使其在RTX 5000系列中独一无二；➁ RTX 5060可能仅配备8GB VRAM，可能让用户失望；➂ RTX 5080的960GB/s带宽令人印象深刻，但16GB的VRAM可能不足以应对未来的需求。➀ The RTX 5080 may feature 16GB of 30Gbps GDDR7 VRAM, making it unique in the RTX 5000 series; ➁ The RTX 5060 might come with only 8GB VRAM, potentially disappointing users; ➂ The RTX 5080's 960GB/s bandwidth is impressive, but the 16GB VRAM may not be enough for future-proofing.
12/17/2024, 02:00 PM UTC
NVIDIA RTX 5080 据称采用更快的 30 Gbps GDDR7 模块，提供 960 GB/s 的带宽 — 剩余的 Blackwell 系列预计将继续使用较慢的 28 Gbps 内存Nvidia RTX 5080 allegedly adopts faster 30 Gbps GDDR7 modules, delivering 960 GB/s of bandwidth — The remaining Blackwell lineup is expected to stick with slower 28 Gbps memory
➀ NVIDIA 的 RTX 5080 据称配备 30 Gbps 的 GDDR7 内存，提供 960 GB/s 的带宽；➁ 剩余的 Blackwell 系列预计将继续使用较慢的 28 Gbps 模块；➂ NVIDIA 的 RTX 50 系列预计将成为首个使用 GDDR7 内存的产品➀ Nvidia's RTX 5080 rumored to feature 30 Gbps GDDR7 memory for 960 GB/s bandwidth; ➁ Remaining Blackwell lineup to use slower 28 Gbps modules; ➂ Nvidia's RTX 50 series expected to be first with GDDR7 memory
12/17/2024, 12:06 AM UTC
ZOTAC确认NVIDIA新一代GeForce RTX 5090显卡配备32GB GDDR7内存ZOTAC confirms NVIDIA's next-gen GeForce RTX 5090 has 32GB of GDDR7 memory
➀ ZOTAC确认新一代GeForce RTX 5090显卡配备32GB GDDR7内存；➁ RTX 5090拥有超过20,000个CUDA核心和高达2.0TB/sec的内存带宽；➂ RTX 5090在4K游戏和光线追踪方面承诺显著性能提升。➀ ZOTAC confirms NVIDIA's GeForce RTX 5090 with 32GB of GDDR7 memory; ➁ The RTX 5090 features over 20,000 CUDA cores and up to 2.0TB/sec memory bandwidth; ➂ The RTX 5090 promises significant performance improvements in 4K gaming and ray tracing.
12/16/2024, 07:16 PM UTC
ASRock Rack 6U8X-EGS2 H200 NVIDIA HGX H200 AI服务器评测ASRock Rack 6U8X-EGS2 H200 NVIDIA HGX H200 AI Server Review
➀ ASRock Rack 6U8X-EGS2 H200是一款全新的NVIDIA HGX H200 8-GPU平台，专为AI训练和推理设计。它配备英特尔Xeon可扩展处理器、16个DDR5内存插槽以及12个带有PCIe Gen5和SATA选项的硬盘托架。服务器还采用了定制的主板设计，以实现PCIe交换机和GPU连接。 ➁ 评测详细介绍了服务器的前后I/O、CPU、内存和存储配置。 ➂ ASRock 6U8X-EGS2 H200展示了ASRock在GPU计算平台方面的持续创新。➀ The ASRock Rack 6U8X-EGS2 H200 is a new NVIDIA HGX H200 8-GPU platform designed for AI training and inference. It features an Intel Xeon Scalable processor, 16 DDR5 DIMMs, and 12 drive bays with PCIe Gen5 and SATA options. The server also includes a custom motherboard design for PCIe switch and GPU connectivity. ➁ The review details the server's front and rear I/O, CPU, memory, and storage configurations. ➂ The ASRock Rack 6U8X-EGS2 H200 showcases ASRock's continued innovation in GPU compute platforms.
12/16/2024, 12:20 PM UTC
蔡崇信与埃隆·马斯克联手机器人C.C.Wei and Elon Musk hooking up on robots
➀ 蔡崇信与埃隆·马斯克正在合作开发多功能机器人；➁ 马斯克计划于2026年推出名为Optimus的多功能人形机器人；➂ 蔡崇信强调无人机和人工智能在水资源、电力等公用事业领域的应用的重要性。➀ C.C. Wei and Elon Musk are collaborating on multifunctional robots; ➁ Musk plans to launch a multifunctional humanoid robot named Optimus in 2026; ➂ Wei emphasizes the importance of drones and AI applications in utilities such as water and electricity provision.
12/16/2024, 01:29 AM UTC
英特尔Arc B580显卡已售罄，更多库存将于2025年1月到来Intel Arc B580 is already sold out at retailers, more stock coming in January 2025
➀ 英特尔Arc B580显卡，售价249.99美元，已在各大零售商处售罄；➁ 预计更多库存将于2025年1月到来；➂ 该显卡性能强劲，性价比高，超越了GeForce RTX 4060和Radeon RX 7600等竞争对手。➀ The Intel Arc B580 GPU, priced at $249.99, is already sold out at major retailers; ➁ More stock is expected to arrive in January 2025; ➂ The GPU offers strong performance and value, surpassing competitors like the GeForce RTX 4060 and Radeon RX 7600.
12/15/2024, 01:31 PM UTC
双英特尔Arc B580电脑配置分享 —— 将与Nvidia Titan Xp一同用于快速FluidX3D模拟Dual Intel Arc B580 PC details shared — they will be used alongside an Nvidia Titan Xp for fast FluidX3D simulations
➀ 英特尔Arc B580显卡与Nvidia Titan Xp一同用于FluidX3D模拟；➁ 系统总共有36GB VRAM；➂ FluidX3D的开发者Moritz Lehmann博士同时也是英特尔的一名GPU软件开发工程师。➀ The Intel Arc B580 GPUs are used for FluidX3D simulations alongside an Nvidia Titan Xp; ➁ The system boasts a total of 36GB of VRAM; ➂ Dr. Moritz Lehmann, the developer of FluidX3D, is also a GPU Software Development Engineer at Intel.
12/13/2024, 02:15 PM UTC
英国在塑料电子领域的领先地位UK Leads In Plastic Electronics
➀ 英国在塑料电子领域是全球的领导者；➁ 曼德尔森勋爵强调了为行业制定路线图的重要性，以从尖端技术走向大众市场；➂ 斯旺西大学和威尔士印刷与涂覆中心参与开发使用塑料电子的新产品线。➀ The UK has been a global leader in the Plastic Electronics sector; ➁ Lord Mandelson emphasized the importance of a roadmap for the industry to move from cutting-edge to mass market; ➂ Swansea University and the Welsh Centre for Printing and Coating are involved in developing new product lines using plastic electronics.
12/13/2024, 01:02 PM UTC
NVIDIA暗示下月发布RTX 50 Blackwell游戏GPU —— 《巫师4》首部预告片可能使用了即将推出的RTX 5090Nvidia teases RTX 50 Blackwell Gaming GPUs for launch next month — The Witcher IV's first cinematic trailer likely leveraged the upcoming RTX 5090
➀ 暗示RTX 50 Blackwell GPU发布；➁ CES 2025活动宣布；➂ 《巫师4》预告片暗示RTX 5090➀ Nvidia tease RTX 50 Blackwell GPUs; ➁ CES 2025 event announcement; ➂ The Witcher IV cinematic trailer hinting at RTX 5090
12/13/2024, 09:18 AM UTC
利用人工智能实现更好的光伏材料Using AI to Achieve Better Photovoltaic Materials
➀ 卡尔鲁厄理工学院的研究人员通过人工智能和高通量合成找到了新的有机分子，以提高钙钛矿太阳能电池的效率。➁ 开发的策略可以应用于其他材料研究领域，如新型电池材料。➂ 该团队使参考太阳能电池的效率提高了2%，达到26.2%。➀ Researchers at KIT have found new organic molecules to improve the efficiency of perovskite solar cells with the help of AI and high-throughput synthesis. ➁ The strategy developed can be applied to other areas of material research, such as new battery materials. ➂ The team achieved a two percent efficiency increase for a reference solar cell, reaching 26.2 percent.
12/12/2024, 02:00 PM UTC
寓言：牛仔CEO的故事Fable: The Cowboy CEO
➀ 这位CEO以他的十加仑帽子和牛仔靴闻名，创立的公司收购了60多家电信公司，市值一度高达1860亿美元。➁ 在一个季度亏损的情况下，公司报告了138亿美元的净利润。➂ 他从公司借款4.08亿美元来弥补保证金调用，最终违约，并在监狱中度过了13年。➀ The CEO, known for his ten gallon hats and cowboy boots, founded a company that acquired over 60 telecoms companies and had a peak market cap of $186 billion. ➁ In one quarter with a loss, the company reported a net profit of $1.38 billion. ➂ He borrowed $408 million from the company to cover margin calls and defaulted on the debt, spending 13 years in jail.
12/12/2024, 12:17 PM UTC
GMG的THERMAL-XR(R)在电子散热片微型化和效率方面展现潜力GMG's THERMAL-XR(R) Demonstrates Potential for Electronics Heat Sink Miniaturization and Efficiency
➀ GMG的THERMAL-XR®涂层散热片展示了微型化和效率的潜力；➁ 该技术可以将散热片尺寸减少高达39%，同时保持等效的散热性能；➂ THERMAL-XR®的应用可以提高散热片的效率，实现更小的散热片和更紧凑的PCB组装。➀ GMG's THERMAL-XR® coated heat sinks demonstrate the potential for miniaturization and efficiency; ➁ The technology could reduce heat sink size by up to 39% while maintaining equivalent thermal performance; ➂ The application of THERMAL-XR® could improve the efficiency of heat sinks, enabling smaller heat sinks and more compact PCB assemblies.
12/10/2024, 06:34 PM UTC
游戏显卡销量第三季度暴跌14.5%——Nvidia RTX 5000系列即将发布Gaming GPU sales plummet 14.5% in third quarter — Nvidia's RTX 5000-series launch looms large
➀ 2024年第三季度游戏显卡销量下降了14.5%；➁ Nvidia的市场份额从88%增长到90%；➂ 关税和经济衰退可能影响显卡销售。➀ Gaming GPU sales dropped 14.5% in Q3 2024; ➁ Nvidia's market share increased from 88% to 90%; ➂ Concerns over tariffs and economic recession may impact GPU sales.
12/09/2024, 03:51 PM UTC
英特尔Arc B580在早期基准测试中与RTX 4060和RX 7600交锋 —— B580在OpenCL和Vulkan工作负载中比A580快30%Intel Arc B580 trades blows with the RTX 4060 and RX 7600 in early benchmarks — B580 beats A580 by up to 30% in OpenCL and Vulkan workloads
➀ 英特尔Arc B580在早期基准测试中比其前代产品A580快30%以上；➁ 基准测试在OpenCL和Vulkan API上完成；➂ Arc B580的价格低于RTX 4060，但性能快10%。➀ Intel's Arc B580 outperforms its predecessor, A580, by up to 30% in early benchmarks; ➁ The benchmarks were conducted across the OpenCL and Vulkan APIs; ➂ The Arc B580 is 10% faster than the RTX 4060 at a lower price point.
12/09/2024, 06:02 AM UTC
Farnell年终促销Farnell end-of-year sale
➀ 慧荣科技全球宣布在EMEA地区推出测试与工具（T&T）产品的年终促销活动；➁ 该促销活动包括广泛测试与工具产品的显著折扣；➂ 促销活动将于2025年1月31日结束。➀ Farnell Global has announced end-of-year offers on Test & Tools (T&T) products across the EMEA; ➁ The sale includes significant discounts on a wide range of T&T products; ➂ The promotional period ends on January 31, 2025.
12/06/2024, 01:23 PM UTC
被暗杀的联合健康CEO涉嫌使用AI拒绝患病者保险Assassinated UnitedHealthcare CEO allegedly used AI to deny sick people coverage
➀ 一项集体诉讼指控联合健康保险公司使用一个有缺陷的算法拒绝患者保险，由两名现已去世的个人提起。➁ 联合健康保险公司首席执行官布莱恩·汤普森本周在曼哈顿中城被杀，嫌疑人目前仍在逃。➂ 诉讼声称联合健康保险公司促使员工使用一个错误率约为90%的算法来拒绝保险。➀ A class-action lawsuit alleges that UnitedHealthcare uses a faulty algorithm to deny patient coverage, filed by two now-deceased individuals. ➁ UnitedHealthcare CEO Brain Thompson was killed in Midtown Manhattan earlier this week, and the suspect is currently on the run. ➂ The lawsuit claims UnitedHealthcare pushed employees to use an algorithm with a 90% error rate to deny coverage.
12/06/2024, 03:02 AM UTC
英特 尔Xe3 'Battlemage-next'架构已定稿，团队已开始开发Xe4Intel's Xe3 'Battlemage-next' architecture has been finalized, team already working on Xe4
➀ 英特尔已定稿Xe3 'Celestial'架构，并已经开始开发Xe4；➁ 英特尔Arc B580 'Battlemage'显卡将于2024年12月13日上市；➂ Arc团队已经转向开发下一代Xe架构。➀ Intel has finalized the Xe3 'Celestial' architecture and is already working on Xe4; ➁ The Intel Arc B580 'Battlemage' GPU is set to launch on December 13, 2024; ➂ The Arc team has already moved on to developing the next generation of Xe architecture.
12/06/2024, 02:08 AM UTC
2024年第三季度GPU出货量下降，玩家等待NVIDIA新GeForce RTX 50系列GPUGPU shipments in Q3 2024 dip, as gamers wait for NVIDIA's new GeForce RTX 50 series GPUs
➀ 2024年第三季度JPR报告显示，GPU出货量为7360万；➁ 2025年至2028年，GPU出货量预计将以-1.9%的复合年增长率下降；➂ 第三季度AMD和英特尔发布了新的CPU。➀ JPR reports 73.6 million GPUs shipped in Q3 2024; ➁ GPU shipments are expected to decline at a -1.9% CAGR from 2025 to 2028; ➂ AMD and Intel released new CPUs in Q3.
12/05/2024, 04:15 PM UTC
AMD确认Radeon RX 8800和RX 8600显卡通过ROCm补丁AMD confirms Radeon RX 8800 and RX 8600 GPUs via ROCm patches
➀ 通过ROCm补丁确认了Radeon RX 8800和RX 8600显卡；➁ ROCm库代码揭示了下一代显卡的命名；➂ RX 8800系列将由Navi 48 GPU提供动力，而RX 8600系列将使用Navi 44 GPU。➀ AMD confirms Radeon RX 8800 and RX 8600 GPUs via ROCm patches; ➁ The ROCm library code reveals the naming of the next-generation graphics cards; ➂ The RX 8800 series will be powered by the Navi 48 GPU, and the RX 8600 series will use the Navi 44 GPU.
12/05/2024, 02:30 PM UTC
英特尔确认Arc Xe4“德鲁伊”显卡正在开发中——Xe3“天际”即将发布，软件优化是最后一步Intel says Arc Xe4 Druid GPUs are already in the works — Software optimization is the only remaining step for Xe3 Celestial as it approaches launch with Panther Lake
➀ 英特尔确认Arc Xe4“德鲁伊”显卡的开发；➁ 即将发布的Xe3“天际”显卡的最后一步是软件优化；➂ 英特尔硬件团队正在开发下一代Xe4显卡。➀ Intel confirms the development of Arc Xe4 'Druid' GPUs; ➁ Software optimization is the last step for the upcoming Xe3 'Celestial' GPUs; ➂ Intel's hardware teams are working on next-generation Xe4 GPUs.
12/05/2024, 08:53 AM UTC
首个可编程连接模块First Programmable Connectivity Module
➀ 本文讨论了首个可编程连接模块，这是电子行业的一项重大发展；➁ 可能探讨了其功能和潜在应用；➂ 该模块可能与连接技术方面的进步有关，影响游戏、人工智能和汽车等各个领域。➀ The article discusses the first programmable connectivity module, a significant development in the electronics industry; ➁ It likely explores its features and potential applications; ➂ The module could be related to advancements in connectivity technology, impacting various sectors such as gaming, AI, and automotive.
12/04/2024, 04:45 PM UTC
独家AMD合作伙伴可能加入英特尔Arc显卡阵营——Onix品牌疑似与Sapphire有关联Exclusive AMD partner reportedly hops on Intel's Arc bandwagon — new Onix brand is seemingly affiliated with Sapphire
➀ 据报道，Onix Technology Limited与Sapphire Technology Limited有关联；➁ Onix的新品牌似乎与Sapphire有关联；➂ 由于资源分配紧张，Sapphire可能选择了英特尔Arc显卡而不是英伟达GPU。➀ Onix Technology Limited is reportedly affiliated with Sapphire Technology Limited; ➁ Onix's new brand is seemingly associated with Sapphire; ➂ Sapphire may have chosen Intel Arc over Nvidia GPUs due to tight allocations.
12/04/2024, 02:00 PM UTC
与Ansys和Synopsys的大师班：多芯片设计最新进展A Master Class with Ansys and Synopsys, The Latest Advances in Multi-Die Design
➀ 2.5D和3D多芯片设计在主流应用中的兴起；➁ Synopsys和Ansys为多芯片项目提供的全面设计流程；➂ Marc Swinnen和Keith Lanier就技术知识和引人入胜的演示提供的专家见解。➀ The rise of 2.5D and 3D multi-die design in mainstream applications; ➁ Comprehensive design flows provided by Synopsys and Ansys for multi-die projects; ➂ Expert insights from Marc Swinnen and Keith Lanier on technical knowledge and engaging presentation.
12/04/2024, 05:11 AM UTC
NVIDIA下一代Rubin AI GPU或因HBM4提前6个月发布NVIDIA's next-gen Rubin AI GPU could be pushed up 6 months ahead of schedule with HBM4
➀ NVIDIA的Rubin AI GPU架构，原定于2026年发布，现在预计提前六个月；➁ 采用台积电的3nm工艺和下一代HBM4内存；➂ 继承了Blackwell架构，NVIDIA正与台湾合作伙伴合作开发基于R100的AI服务器。➀ NVIDIA's Rubin AI GPU architecture, originally set for 2026, is now expected six months earlier; ➁ Utilizing TSMC's 3nm process and next-gen HBM4 memory; ➂ Follows the Blackwell architecture, with NVIDIA collaborating with Taiwanese partners on R100-powered AI servers.
12/04/2024, 04:29 AM UTC
《印第安纳琼斯：大圆圈》需GeForce RTX 4090才能实现4K 60FPS光线追踪Indiana Jones and the Great Circle will need a GeForce RTX 4090 for 4K 60FPS ray tracing
➀ 《印第安纳琼斯：大圆圈》需要GeForce RTX 4090才能在4K 60FPS下实现光线追踪；➁ 游戏已公布PC配置要求，全光线追踪模式需要高端硬件；➂ 游戏定于12月9日发售。➀ Indiana Jones and the Great Circle requires a GeForce RTX 4090 for 4K 60FPS ray tracing; ➁ The game's PC requirements have been announced, with high-end hardware needed for full ray tracing mode; ➂ The game is set to launch on December 9.
12/04/2024, 04:10 AM UTC
NVIDIA新一代智能电动 汽车芯片“Thor”被比亚迪、上汽大众等中国电动汽车制造商采用NVIDIA's new Thor chip for smart EVs picked up by BYD, SAIC Motors, other Chinese EV companies
➀ 恩智浦正在开发面向未来电动汽车的Thor芯片，与中国电动汽车制造商如比亚迪和上汽大众合作。➁ Thor平台有五个版本，目标在2025年底或2026年初实现全面过渡。➂ 恩智浦正通过其新的Drive Thor芯片进入汽车市场。➀ NVIDIA is developing the Thor chip for future EVs, partnering with Chinese EV companies like BYD and SAIC Motors. ➁ The Thor platform, with five versions, aims for a full transition by late 2025 or early 2026. ➂ NVIDIA is moving into the automotive market with its new Drive Thor chip.
12/02/2024, 09:54 PM UTC
AMD Radeon RX 8800 XT传闻：旗舰RDNA 4 GPU比7900 XTX在RT游戏性能上快45%AMD Radeon RX 8800 XT rumor: flagship RDNA 4 GPU is 45% faster in RT gaming than 7900 XTX
➀ AMD的Radeon RX 8800 XT，作为顶级RDNA 4 GPU，预计12月开始量产，功耗比7900 XTX低25%，光线追踪性能提升45%；➁ NVIDIA即将推出的GeForce RTX 5070预计将在CES 2025年超越RDNA 4系列；➂ RX 8800 XT是UDNA之前最后一个RDNA 4显卡，性能和功耗的传闻显示有显著提升。➀ AMD's Radeon RX 8800 XT, the top RDNA 4 GPU, is set for mass production in December with 25% less power consumption and 45% better ray tracing performance than the 7900 XTX; ➁ NVIDIA's upcoming GeForce RTX 5070 is expected to surpass the RDNA 4 stack at CES 2025; ➂ The RX 8800 XT is the last RDNA 4 card before UDNA, with performance and power consumption rumors suggesting significant improvements.
12/02/2024, 06:00 PM UTC
Breker如何帮助解决RISC-V认证问题How Breker is Helping to Solve the RISC-V Certification Problem
➀ RISC-V核心的兴起和认证的挑战；➁ Breker验证系统在认证过程中的作用；➂ 认证RISC-V ISA实现的复杂性以及RISC-V国际组织的努力。➀ The rise of RISC-V cores and the challenges of certification; ➁ The role of Breker Verification Systems in the certification process; ➂ The complexity of certifying RISC-V ISA implementations and the efforts of RISC-V International.
12/02/2024, 12:51 PM UTC
几乎三年后，英特尔Arc显卡终于登上Steam最受欢迎GPU榜单Nearly 3 years after Intel Arc graphics cards arrived, they're finally on Steam's top GPU list
➀ 英特尔Arc品牌以0.18%的市场份额进入Steam的GPU排名；➁ 榜单包括所有Alchemist笔记本电脑GPU；➂ 这些集成显卡解决方案被归为一个标签。➀ Intel's Arc brand has entered Steam's GPU rankings with a 0.18% share; ➁ The ranking includes all Alchemist laptop GPUs; ➂ The integrated graphics solutions are grouped under one label.
12/02/2024, 02:01 AM UTC
英特尔发布两款战神系列显卡 - Arc B570配备10GB VRAM和Arc B580配备12GB VRAMIntel launching two Battlemage GPUs - Arc B570 with 10GB and Arc B580 with 12GB of VRAM
➀ 英特尔即将推出两款新的战神系列显卡，Arc B570配备10GB VRAM，Arc B580配备12GB VRAM；➁ B570拥有18个Xe2核心和10GB VRAM，而B580拥有20个Xe2核心和12GB VRAM；➂ 两款显卡均支持PCIe Gen 4.0 x 8、DisplayPort 2.1和HDMI 2.1。➀ Intel is launching two new Battlemage GPUs, the Arc B570 with 10GB VRAM and the Arc B580 with 12GB VRAM; ➁ The B570 has 18 Xe2 cores and 10GB VRAM, while the B580 has 20 Xe2 cores and 12GB VRAM; ➂ Both GPUs feature PCIe Gen 4.0 x 8, DisplayPort 2.1, and HDMI 2.1 connectivity.
12/02/2024, 12:58 AM UTC
英特尔Arc战神显卡正式发布，全面揭晓将于12月3日到来Intel Arc Battlemage is official, full reveal coming December 3
➀ 英特尔计划于12月3日正式发布下一代战神显卡Arc B580；➁ 该卡预计将与GeForce RTX 4060和Radeon RX 7600竞争，提供12GB VRAM，售价约250美元；➂ 英特尔需要展示其价值主张，以应对GeForce RTX 3060、RTX 4060和Radeon RX 6600和RX 7600等竞争对手。➀ Intel is set to announce its next-generation Battlemage graphics card, the Arc B580, on December 3rd; ➁ The card is expected to compete with GeForce RTX 4060 and Radeon RX 7600, offering 12GB VRAM at around $250; ➂ Intel will need to showcase its value proposition against competitors like GeForce RTX 3060, RTX 4060, and Radeon RX 6600 and RX 7600.
11/29/2024, 05:32 AM UTC
英特尔计划在Arc Battlemage显卡发布后推出更多桌面GPUIntel plans to release more desktop GPUs after the first Arc Battlemage card debuts in December
➀ 英特尔预计将在12月12日左右发布Arc B580 Battlemage显卡，售价约250美元；➁ 英特尔计划继续开发桌面GPU，未来将推出Arc Celestial和Arc Druid型号；➂ 英特尔的重点正在转向预算型桌面GPU。➀ Intel is expected to release the Arc B580 Battlemage GPU around December 12, priced at approximately $250; ➁ Intel plans to continue developing desktop GPUs with future Arc Celestial and Arc Druid models; ➂ Intel's focus is shifting towards budget desktop GPUs.
11/27/2024, 12:36 AM UTC
NVIDIA新款GeForce RTX 5090与RTX 5090D：相同GPU，相同频率，其他部分进行调整NVIDIA's new GeForce RTX 5090, RTX 5090D rumor: same GPU, same clocks, other parts being tuned
➀ NVIDIA的GeForce RTX 5090和RTX 5090D拥有相同的GPU配置和时钟速度；➁ RTX 5090D的推出是为了绕过出口限制，类似于之前的RTX 4090D；➂ 两款型号都将于2025年1月发布，价格相似。➀ NVIDIA's GeForce RTX 5090 and RTX 5090D share the same GPU configuration and clock speeds; ➁ The RTX 5090D was developed to bypass export restrictions, similar to the RTX 4090D; ➂ Both models are set to launch in January 2025 with similar pricing.
11/24/2024, 09:51 PM UTC
未来AI服务器：搭载NVIDIA下一代Rubin Ultra AI GPU，机架密度达1000kW+AI servers of the future: 'rack density' of 1000kW+ with NVIDIA's next-gen Rubin Ultra AI GPUs
➀ 由于NVIDIA即将推出的Rubin Ultra AI GPU和HBM4内存，预计AI服务器将消耗超过1000 kW的电力；➁ 当前AI服务器使用130-250 kW，但未来的Blackwell Ultra和Rubin AI服务器将需要250-900 kW，到2026-2027年，每个机架的GPU数量可能达到576个；➂ NVIDIA未来的AI服务器，配备Rubin Ultra AI GPU，到2028-2029+年将使用1000 kW或更多的电力。➀ AI servers are expected to consume over 1000 kW of power due to NVIDIA's upcoming Rubin Ultra AI GPU and HBM4 memory; ➁ Current AI servers use 130-250 kW, but future models like Blackwell Ultra and Rubin AI servers will require 250-900 kW, with up to 576 GPUs per rack by 2026-2027; ➂ NVIDIA's future AI servers with Rubin Ultra AI GPUs will use 1000 kW or more by 2028-2029+.
11/24/2024, 12:34 AM UTC
NVIDIA将于2025年初推出4款新的RTX 50系列GPU：RTX 5090、5080、5070 Ti和5070NVIDIA to launch 4 new RTX 50 series GPUs in early 2025: RTX 5090, 5080, 5070 Ti and 5070
➀ NVIDIA计划在2025年第一季度推出四款GeForce RTX 50系列GPU；➁ 这些GPU将采用新的Blackwell架构和下一代GDDR7内存；➂ 发布时间表与之前的传闻一致，RTX 5090和RTX 5080预计将在1月份推出。➀ NVIDIA is set to launch four GeForce RTX 50 series GPUs in Q1 2025; ➁ The GPUs will feature the new Blackwell architecture and next-generation GDDR7 memory; ➂ The release schedule aligns with previous rumors, with the RTX 5090 and RTX 5080 expected in January.
11/23/2024, 11:47 PM UTC
NVIDIA GeForce RTX 5070 Ti规格泄露：8960 CUDA核心和300W功耗NVIDIA GeForce RTX 5070 Ti leaked specs tease 8960 CUDA cores and 300W power
➀ 据报道，NVIDIA正在开发GeForce RTX 5070 Ti，配备8960 CUDA核心和70个流处理器；➁ 该GPU预计为GB203 'Blackwell'；➂ 该卡据称具有300W的TDP。➀ NVIDIA is reportedly developing the GeForce RTX 5070 Ti with 8960 CUDA cores and 70 Streaming Multiprocessors; ➁ The GPU is expected to be the GB203 'Blackwell'; ➂ The card is rumored to have a 300W TDP.
11/23/2024, 10:19 AM UTC
您的免费试用Feed URL已于2024年11月23日星期六到期，请创建新的免费URL或升级订阅计划以获得持久URLYour trial feed URL has expired on Sat Nov 23, 2024
➀ 您的免费试用Feed URL已于2024年11月23日星期六到期；➁ 请免费创建新的URL或升级您的订阅计划；➂ 在<a href="https://www.mysitemapgenerator.com/rss/index.html">MySitemapGenerator.com</a>获取持久URL➀ Your trial feed URL has expired on Sat Nov 23, 2024; ➁ Please create a new one for free or upgrade your subscription plan; ➂ Get a persistent URL at <a href="https://www.mysitemapgenerator.com/rss/index.html">MySitemapGenerator.com</a>
11/23/2024, 03:15 AM UTC
NVIDIA GeForce RTX 5090细节：GB202 GPU面积达744mm²，比RTX 4090的AD102大22%NVIDIA GeForce RTX 5090 details: GB202 GPU is 744mm2, around 22% larger than AD102 in RTX 4090
➀ NVIDIA GeForce RTX 5090将使用GB202 'Blackwell' GPU，比RTX 4090中的AD102 'Ada Lovelace' GPU大22%；➁ GB202 GPU的晶圆面积为744mm²，比RTX 4090中的AD102 GPU大22%；➂ 预计RTX 5090将配备20,000+ CUDA核心，32GB的下一代GDDR7内存，以及高达2.0TB/s的内存带宽。➀ The NVIDIA GeForce RTX 5090 will use the GB202 'Blackwell' GPU, 22% larger than the AD102 'Ada Lovelace' GPU in the RTX 4090; ➁ The GB202 GPU has a die size of 744mm², making it 22% larger than the AD102 GPU in the RTX 4090; ➂ The RTX 5090 is expected to have 20,000+ CUDA cores, 32GB of next-gen GDDR7 memory, and up to 2.0TB/sec memory bandwidth.
11/22/2024, 05:15 PM UTC
Comino服务器搭载8块NVIDIA RTX 5090显卡已开始预订 - 这是否意味着旗舰GPU即将到来？Comino servers with 8 x NVIDIA RTX 5090s are on pre-order - a sign the flagship GPU is close?
➀ 慧荣科技推出了搭载NVIDIA RTX 5090显卡的液冷工作站和服务器，这可能意味着Blackwell旗舰GPU即将上市。➁ 慧荣科技为潜在买家提供优先预订权，目前已经开始接受预订。➂ 这些服务器专为高性能设计，可在高达40°C的恶劣环境中运行。➀ Comino has revealed liquid-cooled workstations and servers with Nvidia RTX 5090 graphics cards, indicating that the Blackwell flagship GPU might be close. ➁ Comino is offering priority access to potential buyers, with pre-orders available. ➃ The servers are designed for high performance and can operate in harsh environments up to 40°C.
11/22/2024, 03:36 AM UTC
NVIDIA确认 假期期间GeForce RTX GPU短缺NVIDIA confirms that there's a GeForce RTX GPU shortage as we head into the holidays
➀ NVIDIA第三季度总收入为351亿美元，毛利率为74.6%，其中数据中心业务贡献颇多；➁ GeForce RTX游戏和AI PC部门收入同比增长15%，但当前季度预计将出现短缺；➂ NVIDIA首席财务官科莱特·克雷斯指出，公司正在将重点转向下一代Blackwell驱动的GeForce RTX 50系列，预计2025年初推出。➀ NVIDIA reported a Q3 revenue of $35.1 billion with a gross margin of 74.6%, much of which comes from its Data Center business; ➁ NVIDIA's GeForce RTX Gaming and AI PC sector saw a 15% revenue growth year-over-year, but a shortage is expected in the current quarter; ➂ NVIDIA CFO Colette Kress indicated a shift in focus to the next-gen Blackwell-powered GeForce RTX 50 Series, with the launch expected in early 2025.
11/21/2024, 06:03 AM UTC
NVIDIA确认Blackwell GPU已全面生产，尽管面临冷却挑战NVIDIA confirms Blackwell GPUs are in 'full production' despite cooling challenges
➀ 尽管面临冷却挑战，NVIDIA已确认Blackwell GPU已全面生产；➁ 公司的AI芯片对其成功做出了重大贡献；➂ NVIDIA的AI数据中心业务占其季度收入的307亿美元。➀ NVIDIA has confirmed that Blackwell GPUs are in full production despite cooling challenges; ➁ The company's AI chips have significantly contributed to its success; ➂ NVIDIA's AI data center business accounted for a staggering $30.7 billion of its quarterly earnings.
11/20/2024, 09:01 AM UTC
英特尔正在测试下一代Xe3图形架构，但会出现在台式机上吗？Intel is already testing its next-gen Xe3 graphics architecture, but will it come to desktops?
➀ 英特尔正在测试下一代Xe3图形架构在即将推出的APU中；➁ 公司仍在等待推出第二代Arc台式机“Battlemage”显卡；➂ 英特尔首席执行官帕特·格尔辛格提到Arc未来的重点是集成图形。➀ Intel is testing its next-gen Xe3 graphics architecture in upcoming APUs; ➁ The company is still waiting to launch its second-gen Arc desktop 'Battlemage' graphics cards; ➂ Intel CEO Pat Gelsinger mentioned a focus on integrated graphics for the future of Arc.
11/19/2024, 05:26 PM UTC
AMD RDNA 4可能只是“修复补丁”，如果全新UDNA GPU如传言在2026年第二季度到来AMD RDNA 4 really might be just a 'bug fix' if all-new UDNA GPUs arrive in Q2 2026 as rumored
➀ AMD计划在RX 8000系列之后从RDNA过渡到UDNA架构；➁ UDNA将统一游戏和数据中心GPU，RX 9000预计将在2026年第二季度使用该架构；➂ 消息人士称，MI400和RX 9000使用相同的UDNA，该架构使用类似于GCN的ALU设计。➀ AMD plans to transition from RDNA to UDNA architecture after the RX 8000 series; ➁ UDNA will unify gaming and data center GPUs, with RX 9000 set to use this architecture in Q2 2026; ➂ The leaker claims that MI400 and RX 9000 use the same UDNA, and the architecture uses an ALU design similar to GCN.
11/19/2024, 07:29 AM UTC
散热供应商称GeForce RTX 50系列GPU将‘占领整个市场’GeForce RTX 50 Series GPUs will 'seize the whole market,' cooling supplier claims
➀ 慧荣科技，一家GPU散热供应商，预测即将推出的GeForce RTX 50系列将从12月开始对市场产生重大影响；➁ 该公司将供应NVIDIA GeForce RTX 50 'Blackwell'系列产品的零部件；➂ RTX 50系列预计将于明年年初推出。➀ Auras Technology, a GPU cooling supplier, predicts the upcoming GeForce RTX 50 Series will have a significant market impact starting in December; ➁ The company will supply parts for NVIDIA's GeForce RTX 50 'Blackwell' Series; ➂ The RTX 50 Series is expected to launch early next year.
11/19/2024, 07:03 AM UTC
NVIDIA GeForce RTX 40系列显卡供应紧张，为RTX 50系列发布做准备GeForce RTX 40 Series GPUs are in short supply as NVIDIA prepares for the RTX 50 Series launch
➀ GeForce RTX 4070、RTX 4080和RTX 4090的供应即将在2025年1月RTX 50系列发布前耗尽；➁ 据传闻，NVIDIA将于2025年初推出GeForce RTX 50系列'Blackwell' GPU；➂ GeForce RTX 40系列的生产正在减少或暂停。➀ GeForce RTX 4070, RTX 4080, and RTX 4090 supply is about to run out ahead of the RTX 50 Series launch in January 2025; ➁ NVIDIA rumored to launch GeForce RTX 50 Series 'Blackwell' GPUs in early 2025; ➂ GeForce RTX 40 Series production is being reduced or halted.
11/19/2024, 02:34 AM UTC
英特尔Arc B580 'Battlemage'显卡亮相，预算卡或于12月上市Intel Arc B580 'Battlemage' GPU appears, and the budget card could be here in December
➀ 新的英特尔Arc B580显卡包装已泄露，暗示即将发布；➁ 预计该卡将于2024年12月发布，早于CES 2025；➂ 英特尔Arc B580基于Battlemage架构，是英特尔Arc A580的直接继任者。➀ The packaging for the new Intel Arc B580 GPU has been leaked, suggesting an upcoming launch; ➁ The card is expected to release in December 2024, ahead of CES 2025; ➂ The Intel Arc B580 is based on the Battlemage architecture and is a successor to the Intel Arc A580.
11/14/2024, 04:33 AM UTC
全球GPU收入年底将达1000亿美元，AI推动增长Global GPU revenue to reach $100 billion by the end of the year, driven by AI
➀ 预计到年底全球GPU市场将达到近1000亿美元；➁ 由于AI的兴起和对NVIDIA GPU的高需求，GPU在各种行业中越来越普遍；➂ NVIDIA的GPU销售额可能超过900亿美元，而AMD的AI GPU销售额预计将达到30亿美元。➀ The global GPU market is projected to reach nearly $100 billion by the end of the year; ➁ GPUs are becoming increasingly prevalent in various sectors, driven by the AI boom and high demand for NVIDIA GPUs; ➂ NVIDIA's GPU sales could top $90 billion, with AMD's AI GPU sales expected to reach $3 billion.
11/13/2024, 12:06 AM UTC
NVIDIA转向GeForce RTX 50系列GPU生产，仅剩一个Ada GPU继续生产NVIDIA reportedly shifts production to GeForce RTX 50 series GPUs, only one Ada GPU lives on
➀ 英伟达已转向生产新的GeForce RTX 50系列'Blackwell' GPU；➁ 仅RTX 40系列的AD107生产线仍在运营；➂ RTX 40系列的生产即将结束。➀ NVIDIA has shifted production to the new GeForce RTX 50 series 'Blackwell' GPUs; ➁ Only the AD107 line for the RTX 40 series remains in operation; ➂ The RTX 40 series is nearing the end of its production.
11/08/2024, 08:11 AM UTC
NVIDIA 注册新一代GeForce RTX 60系列显卡商标：RTX 6090 SUPER、RTX 5090 SUPER等NVIDIA trademarks future-gen GeForce RTX 6090 SUPER graphics card, RTX 5090 SUPER teased
➀ NVIDIA已注册GeForce RTX 60系列显卡的商标，包括RTX 6090 SUPER、RTX 6090 Ti、RTX 6090以及RTX 5090 SUPER、RTX 5090 Ti等。➁ 向欧亚经济委员会的注册文件表明，NVIDIA正准备应对AMD即将推出的RDNA架构。➂ NVIDIA为未来的GPU发布做准备，暗示了新旗舰显卡的推出。➀ NVIDIA has trademarked names for the GeForce RTX 60 series, including the RTX 6090 SUPER, RTX 6090 Ti, and RTX 6090, as well as the RTX 5090 SUPER, RTX 5090 Ti GPUs. ➁ The filings with the Eurasian Economic Commission suggest NVIDIA's strategy to stay ahead of AMD's upcoming RDNA. ➃ NVIDIA is preparing for future GPU releases, indicating the launch of new flagship cards.
11/06/2024, 10:46 AM UTC
英伟达超越苹果，成为全球最有价值公司NVIDIA surpasses Apple as the single most valuable company in the world
➀ 英伟达已成为全球最有价值的上市公司，市值达到3.43万亿美元，超越苹果；➁ 英伟达股价在过去5年飙升了2700%，仅2024年就增长了近三倍；➂ 英伟达的增长归功于其在AI硬件领域的统治地位以及抓住AI早期机会的能力。➀ NVIDIA has become the world's most valuable publicly traded company, surpassing Apple with a market cap of $3.43 trillion; ➁ NVIDIA's share price has surged 2,700% over a 5-year period, nearly tripling in 2024 alone; ➂ NVIDIA's growth is attributed to its dominance in AI hardware and its ability to capitalize on the early opportunity of AI.
11/05/2024, 12:39 PM UTC
NVIDIA 即将揭晓下一代 GeForce RTX 50 'Blackwell' 显卡NVIDIA to unveil its next-gen GeForce RTX 50 'Blackwell' GPUs 'soon' says leaker
➀ NVIDIA 的下一代 GeForce RTX 50 系列 'Blackwell' 显卡可能很快就会揭晓；➁ RTX 5090、5080 和 5070 本预计将在 CES 2025 上宣布；➂ 一位可靠的泄露者表示，揭晓可能会在 CES 2025 之前发生。➀ NVIDIA's next-generation GeForce RTX 50 series 'Blackwell' GPUs may be revealed soon; ➁ The RTX 5090, 5080, and 5070 were expected to be announced at CES 2025; ➂ A reputable leaker suggests the reveal could happen before CES 2025.
11/04/2024, 04:46 AM UTC
PC玩家中排名前34的GPU全部为GeForce，NVIDIA在自我竞争Top 34 GPUs among PC gamers are all GeForce, NVIDIA is competing against itself
➀ 2024年10月的Steam硬件与软件调查结果显示，NVIDIA占据了GPU市场的77.37%份额，前34款GPU均为GeForce产品，最受欢迎的是GeForce RTX 3060；➁ 最受欢迎的桌面Radeon选项是Radeon RX 6600，排名第35，其次是Radeon RX 580；➂ AMD在列表中有两款RDNA 3选项——AMD Radeon RX 7900 XTX和AMD Radeon RX 7700 XT，而NVIDIA最新的GeForce RTX 40系列在Top 20中占据了五席。➀ The October 2024 Steam Hardware & Software Survey shows NVIDIA with a 77.37% share of the GPU market, with the top 34 GPUs being GeForce products. The most popular GPU is the GeForce RTX 3060; ➁ The Radeon RX 6600 is the most popular desktop Radeon option, ranking 35th, followed by the Radeon RX 580; ➂ AMD has two RDNA 3 options in the list - the AMD Radeon RX 7900 XTX and AMD Radeon RX 7700 XT, while NVIDIA's latest GeForce RTX 40 Series has five entries in the Top 20.
11/01/2024, 12:07 PM UTC
Meta使用超过10万台NVIDIA H100 AI GPU训练Llama 4，扎克伯格：比我所见过的都要大Meta using over 100,000 NVIDIA H100 AI GPUs for Llama 4, Zuck: 'bigger than anything I've seen'
➀ 马克·扎克伯格宣布了Meta的新Llama 4模型，该模型在超过10万台NVIDIA H100 GPU集群上训练；➁ 据报道，该设置的成本超过20亿美元；➂ Meta正在与xAI和埃隆·马斯克在AI超级集群竞赛中竞争。➀ Mark Zuckerberg announced Meta's new Llama 4 model, trained on a cluster of over 100,000 NVIDIA H100 GPUs; ➁ The setup is reported to cost over $2 billion; ➂ Meta is competing with xAI and Elon Musk in the AI supercluster race.
10/31/2024, 12:30 AM UTC
数据中心GPU的使用寿命可能惊人地短，仅为1到3年[News] Datacenter GPUs May Have an Astonishingly Short Lifespan of Only 1 to 3 Years
➀ 据汤姆的硬件报道，数据中心GPU的使用寿命可能惊人地短，仅为1到3年。 ➁ 这个寿命比预期要短，可能是因为使用频繁和技术进步。 ➂ 这对数据中心运营的可持续性和成本效益有影响。➀ A report from Tom's Hardware suggests that datacenter GPUs have a surprisingly short lifespan of only 1 to 3 years. ➁ The lifespan is shorter than expected, which could be due to heavy use and technological advancements. ➂ This has implications for the sustainability and cost-effectiveness of datacenter operations.
10/29/2024, 08:33 AM UTC
马斯克xAI的Colossus AI超级计算机：搭载10万个NVIDIA H100 AI GPU的深入剖析Elon Musk xAI Colossus AI supercomputer with 100,000 NVIDIA H100 AI GPUs gets in-depth look
➀ 马斯克的xAI仅用122天就建成了Colossus AI超级计算机；➁ 该系统由10万个NVIDIA H100 AI GPU供电；➂ 每个包含64个GPU的Supermicro液冷机架形成小型集群。➀ Elon Musk's xAI has built the Colossus AI supercomputer in just 122 days; ➁ The system is powered by 100,000 NVIDIA H100 AI GPUs; ➂ Supermicro liquid-cooled racks, each containing 64 GPUs, form mini clusters.
10/24/2024, 12:00 PM UTC
谷歌或采用台积电N3E工艺替代2nm制造Tensor G6[News] Google Reportedly Adopts TSMC’s N3P Process instead of 2nm for Tensor G6
➀ 据传闻，谷歌将采用台积电的N3E工艺制造Tensor G5；➁ 报告还明确指出，谷歌没有选择为Tensor G6使用2nm技术；➂ 这一举措可能影响AI和智能手机芯片市场的竞争。➀ Google is rumored to switch to TSMC's N3E process for Tensor G5; ➁ The report also clarifies that Google has chosen not to use 2nm technology for Tensor G6; ➂ The move could impact the competition in the AI and smartphone chip markets.
10/24/2024, 03:25 AM UTC
NVIDIA GeForce RTX 5080笔记本电脑GPU泄露：预计比RTX 4080笔记本电脑GPU快45-65%NVIDIA GeForce RTX 5080 Laptop GPU leaks: estimated 45-65% faster than RTX 4080 Laptop GPU
➀ 据传闻，NVIDIA新一代GeForce RTX 5080笔记本电脑GPU比RTX 4080笔记本电脑GPU快45-65%；➁ 该GPU预计将配备8000+ CUDA核心和16GB的下一代GDDR7内存；➂ 预计RTX 5080笔记本电脑GPU的性能将比RTX 4080笔记本电脑GPU快45-65%，甚至达到桌面RTX 4070 Ti或RTX 4080的水平。➀ NVIDIA's next-gen GeForce RTX 5080 Laptop GPU is rumored to be 45% to 65% faster than the RTX 4080 Laptop GPU; ➁ The GPU is expected to feature 8000+ CUDA cores and 16GB of next-gen GDDR7 memory; ➂ The performance of the RTX 5080 Laptop GPU is predicted to be 45-65% faster than the RTX 4080 Laptop GPU, and even around desktop RTX 4070 Ti or RTX 4080 levels.
10/23/2024, 03:10 AM UTC
摩根士丹利指出NVIDIA 2U空冷MGX GB200 NVL2仍存在散热问题Morgan Stanley says NVIDIA 2U air-cooled MGX GB200 NVL2 still suffers from 'thermal issues'
➀ 摩根士丹利报告称，NVIDIA的2U空冷MGX GB200 NVL2服务器继续面临‘散热问题’；➁ 该服务器在同一PCB板上集成了2颗Grace CPU和2颗Blackwell B200 AI GPU；➂ 由于散热限制，公司可能选择4U的机箱规格。➀ Morgan Stanley reports that NVIDIA's 2U air-cooled MGX GB200 NVL2 server continues to face 'thermal issues'; ➁ The server features 2x Grace CPUs and 2x Blackwell B200 AI GPUs on the same PCB board; ➂ The company may opt for a 4U form factor due to thermal constraints.
10/23/2024, 12:34 AM UTC
泄露视频中所谓的GeForce RTX 5090其实是ZOTAC GAMING GeForce RTX 4070 Ti SUPER SOLIDGeForce RTX 5090 in leaked footage was just a massive RTX 4070 Ti SUPER
➀ 来自印度尼西亚工厂的视频错误地将GeForce RTX 4070 Ti SUPER识别为GeForce RTX 5090；➁ ZOTAC GAMING GeForce RTX 4070 Ti SUPER SOLID是该公司最大的GPU之一；➂ NVIDIA的GeForce RTX 5090和GeForce RTX 5080预计将在CES 2025年展出。➀ Footage from an Indonesian factory mistakenly identified a GeForce RTX 4070 Ti SUPER as a GeForce RTX 5090; ➁ The RTX 4070 Ti SUPER SOLID is one of ZOTAC's largest GPUs; ➂ NVIDIA's GeForce RTX 5090 and GeForce RTX 5080 are expected to be unveiled at CES 2025.
10/20/2024, 08:33 AM UTC
英特尔Arc Battlemage GPU泄露：AIB们对Arc再次‘几乎毫无兴趣’Intel Arc Battlemage GPU leaks: AIBs have 'almost ZERO desire to gamble on Arc' again
➀ 英特尔Arc Battlemage GPU泄露揭示了G31和G21 SKU的预期性能；➁ AIB们对Arc兴趣不大，AMD据报道正在推广Ryzen + RDNA 4捆绑包；➂ 英特尔在Battlemage上面临挑战，包括预算限制和与RDNA 4的竞争。➀ Intel's Arc Battlemage GPU leaks reveal the G31 and G21 SKUs with expected performance levels; ➁ AIBs show little interest in Arc, with AMD reportedly pushing Ryzen + RDNA 4 bundles; ➂ Intel faces challenges with Battlemage, including budget constraints and competition with RDNA 4.
10/18/2024, 02:33 AM UTC
三星预告下一代24Gb GDDR7内存：支持更高容量VRAM，最高42.5Gbps速度Samsung teases next-gen 24Gb GDDR7 memory: enables higher-capacity VRAM, up to 42.5Gbps speeds
➀ 三星开发了下一代24Gbps GDDR7内存模块；➁ 为未来消费级GPU提供更高的VRAM容量；➂ 最高42.5Gbps的速度。➀ Samsung has developed next-gen 24Gbps GDDR7 memory modules; ➁ Offering higher VRAM capacities on future consumer GPUs; ➂ Up to 42.5Gbps speeds.
10/17/2024, 03:49 AM UTC
戴尔PowerEdge XE9712：基于NVIDIA GB200 NVL72的人工智能GPU集群，用于LLM训练和推理Dell PowerEdge XE9712: NVIDIA GB200 NVL72-based AI GPU cluster for LLM training, inference
➀ 戴尔推出了搭载NVIDIA GB200 NVL72 AI服务器的PowerEdge XE9712，其实时LLM性能比H100 AI GPU快30倍；➁ 系统包含72个B200 AI GPU，通过NVLink技术连接，提供超快连接性；➂ 戴尔强调液冷系统可以最大化数据中心电力利用率，并快速部署AI集群。➀ Dell has launched the new PowerEdge XE9712 with NVIDIA GB200 NVL72 AI servers, offering 30x faster real-time LLM performance over the H100 AI GPU; ➁ The system features 72 x B200 AI GPUs connected with NVLink technology, providing lightning-fast connectivity; ➂ Dell highlights the liquid-cooled system for maximizing datacenter power utilization and rapid deployment of AI clusters.
10/17/2024, 01:32 AM UTC
英特尔和AMD正在测试下一代Radeon和Arc GPU的12V-2x6 PCIe电源连接器Intel and AMD are testing 12V-2x6 PCIe power connectors for their next-gen Radeon and Arc GPUs
➀ 英特尔和AMD正在测试下一代Radeon和Arc GPU的12V-2x6 PCIe电源连接器；➁ 预计这一新标准将在即将推出的GPU中使用；➂ 尽管新标准有优势，但由于传统连接器的可靠性，英特尔和AMD仍犹豫不决。➀ Intel and AMD are testing 12V-2x6 PCIe power connectors for their next-gen Radeon and Arc GPUs; ➁ The new standard is expected to be used in upcoming GPUs; ➂ Despite the new standard's benefits, Intel and AMD are hesitant to change due to the reliability of traditional connectors.
10/16/2024, 01:00 PM UTC
移动LLM不只是技术。实际应用案例才是关键Mobile LLMs Aren’t Just About Technology. Realistic Use Cases Matter
➀ 谷歌正在探索在移动设备上运行大型语言模型（LLM）的可行性；➁ 谷歌在移动设备上优化LLM的技术；➂ 移动设备上LLM实际应用案例的重要性➀ Arm is exploring the feasibility of running LLMs on mobile devices; ➁ Arm's optimization techniques for LLMs on mobile; ➂ The importance of practical use cases for LLMs in mobile devices
10/15/2024, 12:55 AM UTC
NVIDIA GeForce RTX 5070 据传将配备18GB GDDR7版本（真的吗？）NVIDIA GeForce RTX 5070 rumored with an 18GB GDDR7 version (huh?!)
➀ 据传，NVIDIA下一代GeForce RTX 5070将配备18GB的GDDR7内存；➁ 初始版本将提供12GB内存，打破VRAM的传统；➂ 18GB版本可能为RTX 5070 SUPER 18GB。➀ NVIDIA's next-generation GeForce RTX 5070 is rumored to launch with 18GB of GDDR7 memory; ➁ Initially, it will be available in a 12GB model, breaking the VRAM mold; ➂ The 18GB version could be the RTX 5070 SUPER 18GB.
10/14/2024, 02:01 AM UTC
NVIDIA Blackwell AI GPU 下12个月将面临缺货NVIDIA Blackwell GPUs for AI are effectively 'sold out' for the next 12 months
➀ 由于需求旺盛，NVIDIA的Blackwell AI GPU未来12个月将面临缺货；➁ 包括谷歌、Meta和微软在内的主要科技公司对Blackwell GPU需求极高；➂ 尽管生产延迟，但Blackwell预计将为AI训练和推理带来显著的性能提升。➀ NVIDIA's Blackwell AI GPUs are 'sold out' for the next 12 months due to high demand; ➁ The GPUs are in high demand among major tech companies like Google, Meta, and Microsoft; ➂ Despite the delay in production, Blackwell is expected to deliver significant performance improvements for AI training and inference.
10/13/2024, 11:44 PM UTC
NVIDIA下一代GeForce RTX 5090价格不会比RTX 4090高很多NVIDIA's next-gen GeForce RTX 5090 won't have huge price increase over RTX 4090 says leaker
➀ 爆料者 'kopite7kimi' 对RTX 5090的1999至2499美元的传言价格表示怀疑；➁ 预计RTX 5090的性能将比RTX 4090有显著提升；➂ RTX 5090据称将配备32GB GDDR7内存、超过20K CUDA核心和600W的功耗。➀ Leaker 'kopite7kimi' disputes the rumored $1999 to $2499 pricing for the RTX 5090; ➁ The RTX 5090 is expected to have significant performance upgrades over the RTX 4090; ➂ The RTX 5090 is rumored to feature 32GB GDDR7 memory, over 20K CUDA cores, and a 600W power requirement.
10/13/2024, 11:13 PM UTC
NVIDIA下一代Blackwell Ultra 'B300' AI GPU：GB300 AI服务器将采用插槽设计NVIDIA's next-gen Blackwell Ultra 'B300' AI GPU for GB300 AI servers: socketed design rumored
➀ 据传闻，NVIDIA的下一代Blackwell Ultra 'B300' AI GPU将在GB300 AI服务器上采用插槽设计；➁ 这种设计变更预计将简化维护和升级；➂ B300系列预计将在2025年下半年成为NVIDIA的主流产品。➀ NVIDIA's next-gen Blackwell Ultra 'B300' AI GPUs are rumored to use a socketed design on GB300 AI servers; ➁ This design change is expected to simplify maintenance and upgrades; ➂ The B300 series is set to become the mainstream product for NVIDIA in the second half of 2025.
10/10/2024, 12:07 AM UTC
OpenAI获得NVIDIA新Blackwell DGX B200 AI系统的首批工程样品OpenAI gets one of the first engineering builds of NVIDIA's new Blackwell DGX B200 AI system
➀ OpenAI获得了NVIDIA新Blackwell DGX B200 AI系统的首批工程样品；➁ 该系统配备8个B200 AI GPU，最高1.4TB的HBM3内存，以及72 petaFLOPS的训练性能；➂ OpenAI对新的系统及其加速AI计算的能力表示兴奋。➀ OpenAI receives one of the first engineering builds of NVIDIA's new Blackwell DGX B200 AI system; ➁ The system features 8 x B200 AI GPUs with up to 1.4TB of HBM3 memory and 72 petaFLOPS of training performance; ➂ OpenAI expresses excitement about the new system and its potential to accelerate AI compute.
10/09/2024, 02:33 AM UTC
NVIDIA GeForce RTX 5080 泄露规格：32Gbps GDDR7，1TB/s 内存带宽NVIDIA GeForce RTX 5080 leaked specs: 16GB GDDR7 at fastest 32Gbps, 1TB/sec bandwidth
➀ 预计 NVIDIA 的 GeForce RTX 5080 将配备最快的 32Gbps GDDR7 内存，16GB GDDR7 提供 1TB/s 的内存带宽；➁ 该 GPU 将拥有 10,752 个 CUDA 核心并预计将与 RTX 4090 性能相匹配；➂ 新的 RTX 5080 将在性能上显著优于 RTX 4080 和 RTX 4080 SUPER。➀ NVIDIA's GeForce RTX 5080 is expected to feature the fastest GDDR7 memory at 32Gbps, with 16GB GDDR7 at 1TB/sec memory bandwidth; ➁ The GPU will have 10,752 CUDA cores and is expected to match RTX 4090 performance; ➂ The new RTX 5080 will offer significant performance improvements over the RTX 4080 and RTX 4080 SUPER.
10/09/2024, 01:34 AM UTC
NVIDIA展示令人惊叹的乐高版《堡垒之夜》PC，搭载GeForce RTX 4080 SUPER GPUNVIDIA showcases stunning LEGO Fortnite PC with GeForce RTX 4080 SUPER GPU
➀ NVIDIA展示了搭载GeForce RTX 4080 SUPER GPU的惊人乐高版《堡垒之夜》PC；➁ 该构建包含超过1000块乐高积木，耗时14天完成；➂ PC拥有完整的乐高《堡垒之夜》场景，包括角色、生物群落和标志性 的《堡垒之夜》巴士。➀ NVIDIA showcases a stunning LEGO Fortnite PC with GeForce RTX 4080 SUPER GPU; ➁ The build includes over 1,000 LEGO bricks and took 14 days to complete; ➂ The PC features a full LEGO Fortnite scene with characters, biomes, and the iconic Fortnite bus.
10/09/2024, 12:31 AM UTC
Prime Day大促销：这些Radeon RX 7800 XT和GeForce RTX 4070 Ti SUPER显卡大优惠GPU Deals - Save big on these Radeon RX 7800 XT and GeForce RTX 4070 Ti SUPER cards
➀ Prime Day促销期间，可购买XFX Radeon RX 7800 XT和MSI GeForce RTX 4070 Ti SUPER等显卡；➁ MSI GeForce RTX 4070 Ti SUPER Ventus 3X OC型号折扣11%，降至749.99美元；➂ XFX Speedster Radeon RX 7800 XT CORE QICK319型号折扣18%，售价427.47美元。➀ With Prime Day sales, save on GPUs like the XFX Radeon RX 7800 XT and MSI GeForce RTX 4070 Ti SUPER; ➁ The MSI GeForce RTX 4070 Ti SUPER Ventus 3X OC is discounted by 11% to $749.99 USD; ➂ The XFX Speedster Radeon RX 7800 XT CORE QICK319 is 18% off at $427.47 USD.
10/07/2024, 03:01 AM UTC
NVIDIA 停产 GeForce RTX 4080 和 RTX 4090，为 RTX 50 系列让路NVIDIA is discontinuing the GeForce RTX 4080 and RTX 4090 to make way for the RTX 50 Series
➀ NVIDIA 正在停产 GeForce RTX 4080 和 RTX 4090；➁ GeForce RTX 4090 库存减少，导致价格上涨；➂ 停产以腾出空间为即将推出的 RTX 50 系列做准备。➀ NVIDIA is discontinuing the GeForce RTX 4080 and RTX 4090; ➁ GeForce RTX 4090 stock is dwindling, causing prices to rise; ➂ Production halt to make room for the upcoming RTX 50 Series.
10/05/2024, 09:55 AM UTC
AMD RX 7000 显卡价格大跳水，现在购买是否划算？AMD's RX 7000 GPUs are getting price cuts left, right and center - so should you buy now?
➀ AMD RX 7000 系列显卡正经历大幅降价；➁ Radeon RX 7900 XTX 和 RX 7700 XT 都已降价；➂ 是否应该等待 RDNA 4 的到来以期待更深的折扣？➀ AMD's RX 7000 GPUs are experiencing significant price cuts; ➁ The Radeon RX 7900 XTX and RX 7700 XT have seen discounts; ➂ Should you wait for deeper discounts with the arrival of RDNA 4?
10/03/2024, 01:59 AM UTC
MSI 新款 RTX 4070 Ti SUPER STALKER 2 显卡免费赠送《STALKER 2：切尔诺贝利之心》游戏MSI's new RTX 4070 Ti SUPER STALKER 2 graphics card includes the game for FREE
➀ MSI 新款 RTX 4070 Ti SUPER STALKER 2 显卡免费附赠《STALKER 2：切尔诺贝利之心》游戏；➁ 该显卡采用军事绿色主题和磨砂金属质感；➂ 新款显卡采用了 MSI 的 GAMING SLIM 设计，是该公司 GAMING SLIM 系列中最强大的定制显卡。➀ MSI's new RTX 4070 Ti SUPER STALKER 2 graphics card includes STALKER 2: Heart of Chernobyl for FREE; ➁ The card features a military green color theme with brushed metal finish; ➂ MSI's GAMING SLIM design is used for the new card, which is the most powerful custom card in the company's GAMING SLIM lineup.
10/02/2024, 06:55 PM UTC
NVIDIA rumored to be halting production of RTX 4060 GPU - but it's nothing to worry aboutNVIDIA's rumored to be halting production of RTX 4060 GPU - but it's nothing to worry about
➀ 据传闻，NVIDIA将暂停RTX 4060和RTX 4060 Ti显卡的生产；➁ 这次暂停是临时的，目的是调整库存水平；➂ 这次行动与RTX 4090的停产无关。➀ NVIDIA is rumored to be halting production of the RTX 4060 and RTX 4060 Ti GPUs; ➁ The halt is temporary and aimed at correcting inventory levels; ➂ The move is not related to the discontinuation of the RTX 4090.
09/30/2024, 04:33 AM UTC
Thermal Grizzly的WireView Pro GPU监控器：在连接器处监控功耗和温度Thermal Grizzly's WireView Pro GPU monitors power usage and temperatures at the connector
➀ 慧荣科技推出WireView Pro GPU监控器，用于测量和监控GPU的功耗和温度。 ➁ 该设备旨在保护显卡免受过热或电源问题造成的潜在损害。 ➂ 它具有OLED显示屏、内部温度传感器和外部温度传感器，以实现精确的测量。➀ Thermal Grizzly introduces the WireView Pro GPU, a tool for measuring and monitoring GPU power usage and temperatures. ➁ The device is designed to protect graphics cards from potential damage due to overheating or power issues. ➂ It features an OLED display, internal temperature sensors, and an external temperature sensor for accurate measurements.
09/29/2024, 10:14 PM UTC
NVIDIA已统治GPU市场20年，AMD和Intel望尘莫及NVIDIA has enjoyed GPU dominance for over 20 years now, leaving AMD and Intel in its GPU dust
➀ NVIDIA已统治GPU市场20年之久；➁ AMD无法竞争，而Intel甚至没有真正进入这场竞赛；➂ 自2010年GeForce 500系列GPU推出以来，NVIDIA的市场份额一直在增加。➀ NVIDIA has been dominating the GPU market for over 20 years; ➁ AMD has been unable to compete and Intel didn't start in the race; ➂ NVIDIA's market share has been increasing since 2010 with the GeForce 500 series GPUs.
09/28/2024, 08:25 AM UTC
AMD下一代RDNA 4 GPU传闻推迟至2025年第一季度：因RDNA 3供过于求AMD's next-gen RDNA 4 GPU rumored delay: now Q1 2025 because of RDNA 3 oversupply
➀ 据传闻，AMD的RDNA 4 GPU因RDNA 3 GPU供过于求而推迟至2025年第一季度；➁ 该问题归因于高端Radeon RX 7000系列GPU的过剩；➂ AMD的RDNA 4策略是针对中端市场，让NVIDIA主导高端和超发烧友市场。➀ AMD's RDNA 4 GPU is rumored to be delayed to Q1 2025 due to an oversupply of RDNA 3 GPUs; ➁ The issue is attributed to an excess of high-end Radeon RX 7000 series GPUs; ➂ AMD's strategy with RDNA 4 is to target the mid-range market, leaving NVIDIA to dominate the high-end and ultra-enthusiast segments.
09/28/2024, 05:33 AM UTC
NVIDIA GeForce RTX 5090 Founders Edition：600W显卡的双插槽设计GeForce RTX 5090 Founders Edition features a dual-slot design, impressive for a 600W GPU
➀ GeForce RTX 5090 Founders Edition将配备双插槽散热器和设计；➁ NVIDIA下一代GeForce RTX 5090和GeForce RTX 5080预计将于明年年初发布；➂ GeForce RTX 5090将具有600W的功耗，并配备紧凑型双插槽散热器。➀ The GeForce RTX 5090 Founder Edition features a dual-slot cooler and build; ➁ NVIDIA's next-generation GeForce RTX 5090 and GeForce RTX 5080 are expected to launch early next year; ➂ The GeForce RTX 5090 will have a 600W power draw and a compact dual-slot cooler.
09/27/2024, 01:01 AM UTC
MSI发布GeForce RTX 4070 Ti SUPER GAMING Slim 'STALKER 2'版显卡MSI reveals GeForce RTX 4070 Ti SUPER GAMING Slim 'STALKER 2' Edition graphics card
➀ MSI发布了GeForce RTX 4070 Ti SUPER GAMING Slim 'STALKER 2'版显卡；➁ 该显卡与GSC Games合作设计，为STALKER 2: 心脏之核游戏打造；➂ 拥有定制设计和工厂超频功能。➀ MSI reveals the GeForce RTX 4070 Ti SUPER GAMING Slim 'STALKER 2' Edition graphics card; ➁ The card is designed in collaboration with GSC Games for STALKER 2: Heart of Chernobyl; ➂ It features a custom design and factory overclocking.
09/26/2024, 04:34 AM UTC
由于价格小幅下降，GPU销量迎来爆发GPU sales are experiencing a boom thanks to a 'slight' drop in price
➀ 2024年第二季度AIB的GPU出货量增加；➁ 英伟达和AMD实现显著增长；➂ GPU销量增长归因于新卡发布和价格下降➀ GPU shipments from AIBs increased in Q2 2024; ➁ NVIDIA and AMD see significant growth; ➂ GPU sales attributed to new card releases and price drop
09/02/2024, 10:00 AM UTC
NVIDIA展示其增强型H200 AI GPU超越AMD最新发布的Instinct MI300XNVIDIA shows off its beefed-up H200 AI GPU beating AMD's just-released Instinct MI300X
➀ NVIDIA的H200 AI GPU在其前代H100的基础上展示了显著的性能提升。 ➁ 与H100相比，H200提供了80%更多的HBM内存和40%更高的带宽。 ➂ NVIDIA的H200在AI性能基准测试中超越了AMD的Instinct MI300X。➀ NVIDIA's H200 AI GPU shows significant performance improvements over its predecessor H100. ➁ The H200 offers 80% more HBM memory and 40% higher bandwidth compared to H100. ➂ NVIDIA's H200 outperforms AMD's Instinct MI300X in AI performance benchmarks.
08/29/2024, 05:03 AM UTC
NVIDIA 调整 Blackwell AI GPU，因 'GPU 掩模' 问题需重新设计 B200NVIDIA says it will tweak Blackwell AI GPUs, issues with the 'GPU mask' needing B200 re-spin
➀ NVIDIA 承认 Blackwell AI GPU 存在问题；➁ 需要重新设计 B200 AI GPU 层以提高产量；➂ B100 AI GPU 被取消，由 B200A 取代➀ NVIDIA admits issues with Blackwell AI GPUs; ➁ Requires re-spin of B200 AI GPU layers to improve yields; ➂ B100 AI GPU effectively canceled, replaced by B200A
08/25/2024, 09:42 PM UTC
MSI 将在所有 AMD X870E 主板配备专用 8 针电源接口，为 RTX 50 系列 GPU 供电MSI wants to feed RTX 50 series GPUs with dedicated 8-pin power on all AMD X870E motherboards
➀ MSI 的 X870E 主板将配备专用的 8 针 PCIe 电源接口，用于 RTX 50 系列 GPU。➁ 这一设计旨在支持 NVIDIA 的下一代 GeForce RTX 50 系列显卡。➂ 其他主板制造商也预计将采用类似的电源接口，以适应高端配置。➀ MSI's X870E motherboards will feature dedicated 8-pin PCIe power connectors for RTX 50 series GPUs. ➁ This design aims to support NVIDIA's next-gen GeForce RTX 50 series cards. ➂ Other motherboard makers are also expected to adopt similar power connectors for high-end configurations.
08/25/2024, 11:02 AM UTC
华硕限量版《战锤40K：太空陆战队2》Radeon 7900 XTX显卡惊艳亮相ASUS's limited edition Warhammer 40K Space Marine 2 Radeon 7900 XTX GPU looks incredible
➀ 华硕与AMD合作推出限量版《战锤40K：太空陆战队2》Radeon RX 7900 XTX显卡。➁ 仅生产20台，采用定制的TUF Gaming设计，融入《战锤40,000》美学元素。➂ 该显卡不对外销售，将通过《战锤》促销网站赠送。➀ ASUS and AMD collaborate to create a limited-edition Warhammer 40K Space Marine 2 Radeon RX 7900 XTX GPU. ➁ Only 20 units are produced, featuring a custom TUF Gaming design with Warhammer 40,000 aesthetics. ➂ The GPU will not be sold but given away through a Warhammer promotion website.
08/24/2024, 01:48 PM UTC
NVIDIA将在Hot Chips 2024深入探讨Blackwell GPU架构NVIDIA to deep dive into the Blackwell GPU architecure at Hot Chips 2024 next week
➀ NVIDIA将在Hot Chips 2024讨论Blackwell GPU架构；➁ 特点是将两个受光罩限制的GPU合并为一个；➂ 拥有2080亿个晶体管和20 PetaFLOPS FP4 AI性能。➀ NVIDIA to discuss Blackwell GPU architecture at Hot Chips 2024; ➁ Features two reticle-limited GPUs merged into one; ➂ Boasts 208 billion transistors and 20 PetaFLOPS FP4 AI performance.
08/24/2024, 01:18 PM UTC
NVIDIA将在即将举行的Hot Chips活动中讨论如何利用AI制造AI芯片NVIDIA to discuss building AI to build chips for AI at upcoming Hot Chips event
➀ NVIDIA将在Hot Chips 2024活动中讨论如何利用AI制造下一代AI芯片。➁ 公司将重点介绍拥有2080亿个晶体管的新Blackwell B200 AI GPU。➂ NVIDIA将探讨使用生成式AI优化用于芯片设计的Verilog代码。➀ NVIDIA will discuss using AI to build next-generation chips for AI at Hot Chips 2024. ➁ The company will focus on the new Blackwell B200 AI GPU with 208 billion transistors. ➂ NVIDIA will explore using generative AI to optimize Verilog code for chip design.
08/24/2024, 12:46 PM UTC
NVIDIA展示在其数据中心实时运行的Blackwell AI GPU：首次制作FP4 GenAI图像NVIDIA shows off Blackwell AI GPUs running in its data center: makes first-ever FP4 GenAI image
➀ NVIDIA展示了其新Blackwell AI GPU在其数据中心实时运行的情景。➁ 公司预告将在Hot Chips上透露更多关于Blackwell GPU架构的 细节。➂ NVIDIA展示了使用这些GPU制作的第一个FP4 GenAI图像。➀ NVIDIA demonstrated its new Blackwell AI GPUs running in real-time in its data centers. ➁ The company teased more details about the Blackwell GPU architecture at Hot Chips. ➂ NVIDIA showcased the first-ever FP4 GenAI image created with these GPUs.
08/24/2024, 06:31 AM UTC
单个NVIDIA RTX 3090显卡可运行超过1,000用户的AI聊天机器人Single NVIDIA RTX 3090 found capable of running 1,000+ user AI chatbot
➀ 一家爱沙尼亚初创公司展示了单个NVIDIA RTX 3090显卡可以运行一个客户聊天机器人AI系统。➁ RTX 3090提供的AI性能与客户服务AI聊天机器人相当，每秒处理12.88个令牌。➂ 这个AI聊天机器人可以支持数千用户，并在任何时候满足100个用户的请求。➀ An Estonian startup demonstrated that a single NVIDIA RTX 3090 can run a customer chatbot AI system. ➁ The RTX 3090 provided AI performance comparable to a customer service AI chatbot, serving 12.88 tokens per second. ➂ This AI chatbot could support thousands of users and fulfill the requests of 100s at any given time.
08/23/2024, 01:59 AM UTC
NVIDIA 赠送定制版《黑神话：悟空》与《夺宝奇兵》GeForce RTX 4080 SUPER 显卡NVIDIA is giving away custom Black Myth Wukong and Indiana Jones GeForce RTX 4080 SUPER cards
➀ NVIDIA 正在赠送两款定制版 GeForce RTX 4080 SUPER 显卡，主题分别为《黑神话：悟空》与《夺宝奇兵》。➁ 这些显卡是 #GeForceSummer of RTX 2024 促销活动的一部分。➂ 赠品活动的报名截止日期为 2024 年 9 月 20 日。➀ NVIDIA is giving away two custom GeForce RTX 4080 SUPER cards themed after Black Myth Wukong and Indiana Jones. ➁ These cards are part of the #GeForceSummer of RTX 2024 promotion. ➂ Entries for the giveaway are open until September 20, 2024.
08/21/2024, 03:36 AM UTC
适用于《黑神话：悟空》、《最终幻想16》及《星球大战：亡命之徒》的GeForce游戏驱动发布GeForce Game Driver for Black Myth Wukong, Final Fantasy 16, and Star Wars Outlaws is here
➀ 最新GeForce Game Ready 560.94 WHQL驱动支持《黑神话：悟空》、《星球大战：亡命之徒》和《最终幻想16》等大型游戏。➁ 特性包括全光线追踪、DLSS 3和Reflex技术。➂ 同时支持如配备GDDR6内存的GeForce RTX 4070等新硬件。➀ New GeForce Game Ready 560.94 WHQL driver supports major games like Black Myth: Wukong, Star Wars Outlaws, and Final Fantasy 16. ➁ Features include Full Ray Tracing, DLSS 3, and Reflex technologies. ➂ Also supports new hardware like the GeForce RTX 4070 with GDDR6 memory.
08/12/2024, 02:46 AM UTC
高通Adreno GPU发现漏洞：Snapdragon芯片加入问题重重的Intel和AMD CPU行列Qualcomm's Adreno GPU vulnerability found: Snapdragon chips join problematic Intel, AMD CPUs
➀ 谷歌研究人员在高通的Adreno GPU中发现了9个漏洞；➁ 这些漏洞可能允许黑客完全控制高通驱动的设备；➂ 高通已经修补了这些问题，但用户仍应保持警惕。➀ Google researchers discovered 9 vulnerabilities in Qualcomm's Adreno GPU; ➁ These vulnerabilities could allow hackers full control over Qualcomm-powered devices; ➂ Qualcomm has already patched the issues, but users should still be cautious.
08/08/2024, 08:31 AM UTC
AMD将于8月23日举办神秘的Gamescom活动，新一代Radeon显卡即将发布？AMD is hosting a mysterious Gamescom Event on August 23, is a new Radeon announcement coming?
➀ AMD将于8月23日举办一场神秘的Gamescom活动；➁ 该活动可能会宣布或预告新一代Radeon显卡；➂ ASUS宣布了一款新的限量版Warhammer 40,000: Space Marine 2主题显卡，暗示了AMD Gamescom活动。➀ AMD is hosting a mysterious Gamescom Event on August 23; ➁ The event may announce or tease the next-generation of Radeon GPUs; ➂ ASUS has announced a new limited-edition Warhammer 40,000: Space Marine 2-themed graphics card, hinting at an AMD Gamescom Event.
08/05/2024, 01:56 PM UTC
Steam数据显示GeForce RTX 4060在PC游戏玩家中越来越受欢迎Steam data shows the GeForce RTX 4060 is becoming really popular with PC gamers
➀ GeForce RTX 4060在最新的Steam硬件和软件调查中进入了前五名GPU。➁ 它的人气超过了GeForce GTX 1060和RTX 2060。➂ RTX 4060被定位为游戏玩家的高性价比选择，在游戏GPU市场中挑战AMD和Intel。➀ The GeForce RTX 4060 has entered the Top 5 GPUs in the latest Steam Hardware & Software Survey for July 2024. ➁ It has surpassed the GeForce GTX 1060 and RTX 2060 in popularity. ➂ The RTX 4060 is positioned as a cost-effective choice for gamers, challenging AMD and Intel in the gaming GPU market.
08/05/2024, 01:28 PM UTC
NVIDIA或将停产其最受欢迎的游戏GPU——GeForce RTX 3060NVIDIA is reportedly discontinuing its most popular gaming GPU, the GeForce RTX 3060
➀ NVIDIA计划停产其最受欢迎的游戏GPU——GeForce RTX 3060。➁ 此举将推动玩家转向新一代的RTX 4060。➂ 根据Steam硬件调查，RTX 3060一直是PC游戏玩家中最常用的GPU。➀ NVIDIA plans to discontinue the GeForce RTX 3060, its most popular gaming GPU. ➁ The decision is pushing gamers towards the newer RTX 4060. ➂ The RTX 3060 has been the most used GPU among PC gamers according to the Steam Hardware Survey.
08/05/2024, 03:19 AM UTC
AMD 增强版 Strix Halo APU 泄露：升级版 RDNA 3.5 GPU 提供 RTX 4070 80W 游戏性能AMD's beefed-up Strix Halo APU leaked: upgraded RDNA 3.5 GPU offers RTX 4070 80W gaming perf
➀ AMD 的 Strix Halo APU 配备 16 核 32 线程的 Zen 5 CPU 和升级版 RDNA 3.5 GPU。 ➁ APU 的 GPU 部分占 307mm²，CPU 部分占 66mm²，总计 439mm²。 ➂ 它提供三种热设计配置：55W、80W 和 120W，内存为 LPDDR5X-8533，接口为 256 位。➀ AMD's Strix Halo APU features 16 cores and 32 threads of Zen 5 CPU and an upgraded RDNA 3.5 GPU. ➁ The APU has a die size of 307mm² for the GPU and 66mm² for the CPU, totaling 439mm². ➂ It offers three thermal configurations: 55W, 80W, and 120W, with LPDDR5X-8533 memory on a 256-bit interface.
08/05/2024, 02:56 AM UTC
NVIDIA B200 Blackwell Ultra AI GPU配备288GB HBM3E，全新设计的B200A Blackwell Ultra配备144GBNVIDIA B200 Blackwell Ultra AI GPU has 288GB HBM3E, redesigned B200A Blackwell Ultra with 144GB
➀ NVIDIA的B200 Blackwell Ultra配备高达288GB的HBM3E内存。➁ 全新设计的B200A Blackwell Ultra配备144GB的HBM3E内存。➂ 这两款GPU都属于NVIDIA的中代更新，重点在于内存和性能的提升。➀ NVIDIA's B200 Blackwell Ultra features up to 288GB of HBM3E memory. ➁ The redesigned B200A Blackwell Ultra has 144GB of HBM3E memory. ➂ Both GPUs are part of NVIDIA's mid-generation refresh, focusing on memory and performance enhancements.
08/05/2024, 02:17 AM UTC
NVIDIA新款Blackwell AI GPU遭遇重大问题，需重新设计导致大幅延迟NVIDIA's new Blackwell AI GPUs 'encountering major issues' which require redesign, big delays
➀ NVIDIA的新款Blackwell AI GPU系列面临重大问题，将导致生产目标推迟至2025年上半年。➁ 问题涉及GPU架构和台积电的新CoWoS-L芯片封装。➂ NVIDIA据报道正在延长Hopper AI GPU系列，并引入新的Blackwell系列GPU，即B200A，以解决这些问题。➀ NVIDIA's new Blackwell AI GPU family is facing major issues that will delay production targets into the first half of 2025. ➁ The issues are related to the GPU architecture and TSMC's new CoWoS-L chip packaging. ➂ NVIDIA is reportedly extending the Hopper AI GPU series and introducing a new Blackwell family GPU, the B200A, to address the issues.
08/05/2024, 12:19 AM UTC
AMD新款Strix Point Radeon 890M GPU在《赛博朋克2077》中比英特尔Meteor Lake快40-50%AMD's new Strix Point Radeon 890M GPU is 40-50% faster than Intel Meteor Lake in Cyberpunk 2077
➀ AMD的新款Radeon 890M集成GPU在Strix Point APU中在游戏测试中超越了英特尔的Meteor Lake。➁ 基于RDNA 3.5的Radeon 890M在17W下运行，而英特尔的则是28W+。➂ 在《赛博朋克2077》中，Radeon 890M比英特尔的Arc GPU性能高出40-50%。➀ AMD's new Radeon 890M integrated GPU in Strix Point APUs outperforms Intel's Meteor Lake in gaming tests. ➁ The Radeon 890M, based on RDNA 3.5, operates at 17W, compared to Intel's 28W+. ➂ In Cyberpunk 2077, the Radeon 890M shows a 40-50% performance advantage over Intel's Arc GPU.
07/30/2024, 02:02 AM UTC
GDDR6X内存短缺将导致热门GeForce RTX 40系列GPU难以寻觅GDDR6X memory shortage to make this popular GeForce RTX 40 Series GPU hard to find
1、由于GDDR6X内存问题，GeForce RTX 4070预计将面临供应短缺；2、短缺问题还影响RTX 4070 SUPER、RTX 4070 Ti SUPER、RTX 4080 SUPER和GeForce RTX 4090；3、使用标准GDDR6内存的GeForce RTX 4060和RTX 4060 Ti不受影响。1. GeForce RTX 4070 expected to be in short supply due to GDDR6X memory issues; 2. The shortage also affects RTX 4070 SUPER, RTX 4070 Ti SUPER, RTX 4080 SUPER, and GeForce RTX 4090; 3. GeForce RTX 4060 and RTX 4060 Ti unaffected as they use standard GDDR6 memory.
07/30/2024, 01:39 AM UTC
AMD新款基于RDNA 3.5架构的Radeon 890M、880M集成GPU在Strix Point APU中测试AMD's new RDNA 3.5-based Radeon 890M, 880M integrated GPU inside Strix Point APUs tested
1、AMD的新款Ryzen AI 300系列'Strix Point' APUs搭载了新的RDNA 3.5 GPU，即Radeon 890M和880M。2、Radeon 890M和880M相比上一代RDNA 3 GPU提供了高达41%的性能提升。3、这些新的APU包含了Zen 5 CPU核心、RDNA 3.5 GPU核心以及一个新的XDNA 2 NPU用于AI工作负载。1. AMD's new Ryzen AI 300 series 'Strix Point' APUs feature a new RDNA 3.5 GPU, the Radeon 890M and 880M. 2. The Radeon 890M and 880M offer up to 41% more performance than previous-gen RDNA 3 GPUs. 3. The new APUs include Zen 5 CPU cores, RDNA 3.5 GPU cores, and a new XDNA 2 NPU for AI workloads.
07/29/2024, 11:00 AM UTC
NVIDIA高端RTX 40系列GPU产量削减50%，为RTX 50系列做准备NVIDIA's high-end RTX 40 series GPU production slashed by 50% preparing for RTX 50 series GPUs
1、NVIDIA为了准备推出RTX 50系列，已将高端RTX 40系列GPU的产量削减了50%。2、RTX 40系列，特别是RTX 4070及以上型号，预计将面临供应短缺和价格上涨。3、由于AI服务器市场的增长，像ASUS、GIGABYTE、MSI和ASRock这样的AIB制造商可能会从价格上涨中受益。1. NVIDIA has reduced the production of high-end RTX 40 series GPUs by up to 50% in preparation for the launch of the RTX 50 series. 2. Supply shortages and price hikes are expected for the RTX 40 series, especially for models like the RTX 4070 and above. 3. AIB manufacturers like ASUS, GIGABYTE, MSI, and ASRock may benefit from the price increases due to the AI server boom.
07/26/2024, 08:00 PM UTC
华擎推出适用于服务器的被动散热版 Radeon RX 7900 XTX 与 XT 显卡ASRock Launches Passively Cooled Radeon RX 7900 XTX & XT Cards for Servers
1、华擎推出适用于服务器的被动散热版 Radeon RX 7900 XTX 和 XT 显卡，以解决官方服务器显卡的高成本问题。2、这些显卡采用无 风扇的双槽散热器，设计用于高气流强制风冷，并配备12V-2×6电源连接器。3、分别配备24GB和20GB内存，这些显卡面向内存密集型的AI应用，提供相比高端专业显卡更具成本效益的解决方案。1. ASRock introduces passively cooled Radeon RX 7900 XTX and XT cards for server use, addressing the high cost of official server cards. 2. These cards feature a dual-slot heatsink without fans, designed for high-airflow forced-air cooling, and are equipped with a 12V-2×6 power connector. 3. With 24GB and 20GB of memory respectively, these cards are targeted at memory-intensive AI applications, offering a cost-effective solution compared to high-end professional cards.
07/26/2024, 04:07 AM UTC
NVIDIA推出新款GeForce RTX 3050，基于Ada Lovelace架构NVIDIA has a new GeForce RTX 3050, and it's built on Ada Lovelace
1、NVIDIA计划发布新款GeForce RTX 3050笔记本GPU。2、新GPU将采用Ada Lovelace架构，并使用比GeForce RTX 4050笔记本GPU更大的芯片。3、NVIDIA最新驱动程序中的AD106M GPU条目证实了这一新GPU的即将发布。1. NVIDIA plans to release a new version of the GeForce RTX 3050 Laptop GPU. 2. The new GPU will feature Ada Lovelace architecture and use a larger chip than the GeForce RTX 4050 Laptop GPU. 3. The AD106M GPU, listed in NVIDIA's latest driver release, confirms the upcoming launch of this new GPU.
07/26/2024, 02:11 AM UTC
NVIDIA GeForce RTX 5090传闻：比RTX 4090快48%，TITAN AI比RTX 4090快63%NVIDIA GeForce RTX 5090 rumor: 48% faster than RTX 4090, TITAN AI is 63% faster than RTX 4090
1、NVIDIA的GeForce RTX 5090传闻比RTX 4090快48%。2、TITAN AI版本据报道比RTX 4090快63%。3、两款显卡均基于新的GB202架构，并配备下一代GDDR7内存。1. NVIDIA's GeForce RTX 5090 is rumored to be 48% faster than the RTX 4090. 2. The TITAN AI variant is reported to be 63% faster than the RTX 4090. 3. Both cards are based on the new GB202 architecture and feature next-gen GDDR7 memory.
07/22/2024, 03:26 AM UTC
SCALE工具让NVIDIA CUDA应用程序无需修改代码即可运行在AMD Radeon GPU上SCALE tool lets NVIDIA CUDA applications run on AMD Radeon GPUs without modifying code
1、Spectral Compute推出SCALE工具，允许NVIDIA CUDA应用程序无需修改代码即可在AMD GPU上运行。2、SCALE模仿NVIDIA CUDA Toolkit，使工具和脚本无缝运行。3、该工具已通过Blender和Llama-cpp等应用程序在AMD GPU上的测试。1. Spectral Compute introduces SCALE, a toolkit allowing NVIDIA CUDA applications to run on AMD GPUs without code modification. 2. SCALE mimics the NVIDIA CUDA Toolkit, enabling tools and scripts to function seamlessly. 3. The toolkit has been tested with various applications, including Blender and Llama-cpp, running on AMD GPUs.
07/18/2024, 03:49 AM UTC
AMD新款基于RDNA 3.5架构的Radeon 880M集成GPU在Strix Point APU中：ASUS称性能提升15%AMD's new RDNA 3.5-based Radeon 880M integrated GPU in Strix Point APUs: 15% faster, says ASUS
1、AMD的新款Strix Point APU搭载了Radeon 880M集成GPU，性能比前代提升15%。2、Radeon 880M旨在达到40W GeForce RTX 3050笔记本GPU的性能水平。3、ASUS在其Vivobook笔记本中展示了这款新GPU，强调了在MOBA和FPS游戏中的性能提升。1. AMD's new Strix Point APUs feature the Radeon 880M integrated GPU, which is 15% faster than its predecessor. 2. The Radeon 880M is designed to match the performance of a 40W GeForce RTX 3050 Laptop GPU. 3. ASUS showcases the new GPU in its Vivobook laptop, highlighting improved gaming performance in MOBA and FPS titles.
07/17/2024, 05:32 AM UTC
英特尔下一代Panther Lake CPU细节曝光：最高16核，Xe3 'Celestial' GPU，5 Tiles设计Intel Panther Lake-H, Panther Lake-U CPU details: up to 16 cores, Xe3 'Celestial' GPU, 5 Tiles
1、英特尔下一代Panther Lake CPU将配备最多16个核心和12个Xe3 'Celestial' GPU核心。2、这些CPU将分为Panther Lake-H和Panther Lake-U两个系列。3、Panther Lake CPU将在2025年上半年采用英特尔的18A工艺节点进入生产，预计在2025年下半年上市。1. Intel's next-gen Panther Lake CPUs will feature up to 16 cores and 12 Xe3 'Celestial' GPU cores. 2. The CPUs will be available in two series: Panther Lake-H and Panther Lake-U. 3. The Panther Lake CPUs will enter production on Intel's 18A process node in the first half of 2025, with expected availability in the second half of 2025.
07/16/2024, 03:59 AM UTC
GeForce RTX 4070 原型机曝光：幸好最终版有12GB显存GeForce RTX 4070 prototype with 10GB of VRAM emerges, thankfully we got 12GB
1、NVIDIA未发布的GeForce RTX 4070原型机曝光，配备10GB显存；2、实际发布的版本配备12GB显存，内存带宽快20%；3、原型机拥有更多的CUDA核心和更快的时钟速度。1. NVIDIA's unreleased GeForce RTX 4070 prototype with 10GB VRAM is revealed; 2. The actual released version features 12GB VRAM with a 20% faster memory bandwidth; 3. The prototype has more CUDA Cores and faster clock speeds than the retail version.
07/16/2024, 01:55 AM UTC
NVIDIA即将在中国发布下一代旗舰GeForce RTX 5090D GPUNVIDIA set to release a next-gen flagship GeForce RTX 5090D GPU in China
1、NVIDIA计划于2025年1月发布专为中国市场定制的GeForce RTX 5090D，因美国对高性能GPU的出口限制。2、RTX 5090D预计将成为PC游戏强有力的选择，但AI能力受限。3、有猜测称GeForce RTX 50系列（包括RTX 5090和RTX 5080）将于2025年1月全球 发布。1. NVIDIA plans to release the GeForce RTX 5090D in January 2025, tailored for the Chinese market due to U.S. export restrictions on high-performance GPUs. 2. The RTX 5090D is expected to be a powerhouse for PC gaming but with limited AI capabilities. 3. There is speculation about a global launch of the GeForce RTX 50 Series in January 2025, including the RTX 5090 and RTX 5080.
07/15/2024, 02:59 AM UTC
英伟达为下一代Blackwell AI GPU向台积电增加25%订单，应对强劲AI需求NVIDIA raises orders 25% for TSMC for its next-gen Blackwell AI GPUs amid strong AI demand
1、由于强劲的AI需求，英伟达为下一代Blackwell AI GPU向台积电增加了25%的订单。2、Blackwell AI GPU系列，包括B100、B200和GB200，预计将在2025年主导AI市场。3、亚马逊、戴尔、谷歌、Meta和微软等主要科技公司计划在其新的AI服务器中使用这些GPU。1. NVIDIA increases orders for TSMC by 25% for its next-gen Blackwell AI GPUs due to strong AI demand. 2. The Blackwell AI GPU family, including B100, B200, and GB200, is expected to dominate the AI market in 2025. 3. Major tech companies like Amazon, Dell, Google, Meta, and Microsoft plan to use these GPUs in their new AI servers.
07/15/2024, 12:54 AM UTC
NVIDIA下一代GeForce RTX 5090：500W TDP，RTX 5080 350W TDP由Seasonic泄露NVIDIA's next-gen GeForce RTX 5090: 500W TDP, RTX 5080 with 350W TDP leaked by Seasonic
1、NVIDIA下一代GeForce RTX 50系列TDP由Seasonic泄露，RTX 5090为500W，RTX 5080为350W。2、RTX 5090预计比当前的RTX 4090高出11%的TDP。3、NVIDIA据报道正在开发下一代散热解决方案和PCB设计，以应对增加的功耗。1. NVIDIA's next-gen GeForce RTX 50 series TDPs leaked by Seasonic, with RTX 5090 at 500W and RTX 5080 at 350W. 2. The RTX 5090 is expected to have an 11% higher TDP than the current RTX 4090. 3. NVIDIA is reportedly working on next-gen cooling solutions and PCB designs to handle the increased power consumption.
07/05/2024, 01:37 AM UTC
AMD透露有客户准备斥资数十亿打造搭载120万块GPU的AI超级计算机AMD teases 'sober people' ready to spend billions on '1.2 million GPU' AI supercomputer
1、AMD收到来自未知客户的咨询，要求建造一台搭载120万块AI GPU的超级计算机。2、AMD数据中心解决方案集团执行副总裁兼总经理Forrest Norrod在接受采访时证实了这一消息。3、该项目规模预计是当前最大超级计算机Frontier的30倍。1. AMD has received inquiries from unknown clients to build a supercomputer with 1.2 million AI GPUs. 2. Forrest Norrod, EVP and GP of AMD's Datacenter Solutions Group, confirmed the news in an interview. 3. The scale of the project is estimated to be 30 times larger than the current largest supercomputer, Frontier.
07/05/2024, 01:01 AM UTC
华擎新款Radeon RX 7900 Creator系列GPU：配备12V-2x6电源接口与吹风式散热器ASRock's new Radeon RX 7900 Creator Series GPU features 12V-2x6 power connector, blower cooler
1、华擎将其Radeon 7900 WS系列更名为Creator系列。2、新系列配备吹风式散热器和新型12V-2x6电源接口。3、设计用于多GPU集群、AI推理、生成式AI和专业工作站使用。1. ASRock rebrands its Radeon 7900 WS series to the Creator Series. 2. Features include a blower-style cooler and a newer 12V-2x6 power connector. 3. Designed for multi-GPU clusters, AI inference, generative AI, and professional workstation use.
07/03/2024, 03:00 AM UTC
第三波推出搭载Ryzen Threadripper PRO和四块RTX 6000 Ada世代GPU的AI专用电脑Thirdwave Launches AI-Focused PC with Ryzen Threadripper PRO and Four RTX 6000 Ada Generation GPUs
1. 第三波推出新的AI开发工作站，配备强大的硬件；2. raytrek Workstation N8630可搭载多达4块RTX 6000 Ada GPU和高性能CPU；3. 公司还宣布正在开发自己的AI模型raytrek AI，用于企业数据分析。1. Thirdwave introduces new workstations for AI development with powerful hardware; 2. The raytrek Workstation N8630 features up to 4 RTX 6000 Ada GPUs and a high-performance CPU; 3. The company also announces the development of its own AI model, raytrek AI, for business data analysis.
07/01/2024, 01:10 AM UTC
英特尔下一代Arc Battlemage X2和X3 GPU曝光：最高448个EU和256位内存总线Intel Arc Battlemage X2 and X3 GPUs spotted: up to 448 EUs on a 256-bit memory bus
1、英特尔下一代Arc Battlemage X2和X3 GPU被发现，最高配备448个EU和256位内存总线。2、Battlemage 'X3'采用BMG-G10 GPU芯片。3、X3型号具有448个EU或28个Xe2-GPU核心，16GB GDDR6X内存和一个8层PCB。1. Intel's next-generation Arc Battlemage X2 and X3 GPUs have been spotted with up to 448 EUs and a 256-bit memory bus. 2. The Battlemage 'X3' uses the BMG-G10 GPU die. 3. The X3 model features 448 EUs or 28 Xe2-GPU cores, 16GB of GDDR6X memory, and an 8-layer PCB.
06/29/2024, 03:51 AM UTC
第40集TT秀 - AI音乐之战，VR游戏衰退，Kingpin显卡回归TT Show Episode 40 - The war on AI music, VR's decline, and Kingpin returns
1、音乐行业对AI音乐生成器提起诉讼；2、VR游戏正在衰退，索尼停止了PlayStation VR 2；3、高端GeForce 'Kingpin Edition'显卡将通过与PNY的合作回归。1. The music industry is suing AI music generators; 2. VR gaming is declining, with Sony halting PlayStation VR 2; 3. High-end GeForce 'Kingpin Edition' GPUs are set to return through a partnership with PNY.
06/21/2024, 04:09 AM UTC
AMD降低Ryzen 8000G 'Hawk Point' APU系列价格AMD drops Ryzen 8000G 'Hawk Point' APU pricing: Ryzen 7 8700G APU now $299, 8500G for $159
1、AMD降低了其Ryzen 8000G系列'Hawk Point' APU的价格；2、Ryzen 7 8700G APU的价格现在是299美元，从329美元降低；3、Ryzen 5 8600G和8500G APU也分别降价至199美元和159美元。1. AMD has reduced the prices of its Ryzen 8000G series 'Hawk Point' APUs; 2. The Ryzen 7 8700G APU is now priced at $299, down from $329; 3. The Ryzen 5 8600G and 8500G APUs have also seen price reductions, to $199 and $159 respectively.
06/21/2024, 01:01 AM UTC
MAINGEAR发布新款ML-16游戏笔记本：搭载酷睿超能CPU和RTX 40系列GPU，起价1869美元MAINGEAR unleashes ML-16 gaming laptop: Core Ultra CPU, RTX 40 series GPU starts from $1869
1. MAINGEAR发布新款ML-16游戏笔记本，搭载英特尔酷睿超能100系列'Meteor Lake' CPU和NVIDIA GeForce RTX 40系列'Ada Lovelace' GPU。2. ML-16旨在提供高生产力和游戏性能，配备16英寸2560 x 1600显示屏，240Hz刷新率，80Wh电池和多种电源模式。3. 功能包括RGB键盘，Wi-Fi 6E，蓝牙5.3，可定制的SSD和RAM选项，以及限量首发版，带有独家额外赠品。1. MAINGEAR introduces the new ML-16 gaming laptop with Intel's Core Ultra 100 series 'Meteor Lake' CPUs and NVIDIA's GeForce RTX 40 series 'Ada Lovelace' GPUs. 2. The ML-16 aims for high productivity and gaming performance with a 16-inch 2560 x 1600 display at 240Hz, an 80Wh battery, and multiple power modes. 3. Features include an RGB keyboard, Wi-Fi 6E, Bluetooth 5.3, customizable SSD and RAM options, and a limited launch edition with exclusive extras.
06/18/2024, 01:01 AM UTC
NVIDIA CEO黄仁勋：公司1000亿美元收入大多在台湾制造NVIDIA CEO Jensen Huang says company's $100 billion revenue is mostly manufactured in Taiwan
1、NVIDIA CEO黄仁勋访问台湾参加2024年Computex并会见科技合作伙伴。2、他提到NVIDIA的年增长超过1000亿美元，大部分在台湾制造。3、AI已经使台湾及其公司受益，NVIDIA在AI GPU市场处于领先地位。4、黄仁勋鼓励年轻人专注于学习并尽力而为。5、他希望NVIDIA愿意接受挑战，从失败中学习，而不仅仅是追求财务增长。1. NVIDIA CEO Jensen Huang visited Taiwan for Computex 2024 and met tech partners. 2. He mentioned NVIDIA's annual growth exceeds $100 billion, mostly manufactured in Taiwan. 3. AI has benefited Taiwan and its companies, and NVIDIA is leading the AI GPU market. 4. Jensen encourages young people to focus on learning and doing their best. 5. He aims for NVIDIA to embrace challenges and learn from failures, not just pursue financial growth.
06/16/2024, 03:29 AM UTC
AMD下一代RDNA 5可能采用独特的“多芯片”GPU设计AMD's next-gen RDNA 5 could feature unique 'multi-chiplet' GPU, with three dedicated dies
1. AMD为未来可能采用的'多芯片'设计的GPU架构提交了一项专利，这可能在RDNA 5中得到应用。2. 该专利描述了三种不同的芯片利用模式，包括'单GPU'模式、'独立性模式'和'混合模式'。3. 多芯片方法允许 灵活且成本有效地配置GPU资源，并可用于组装具有不同数量芯片的不同GPU。1. AMD has filed a patent for a future GPU architecture that uses a 'multi-chiplet' design, which could be featured in RDNA 5. 2. The patent describes three different modes of chiplet utilization, including 'single GPU' mode, 'independency mode', and 'hybrid mode'. 3. The multi-chiplet approach allows for flexible and cost-effective configuration of GPU resources, and can be used to assemble multiple different GPUs with varying numbers of chiplets.
06/13/2024, 12:06 AM UTC
NVIDIA下一代RTX 5090传闻：512位GDDR7显存，RTX 5080将配备256位显存NVIDIA's next-gen RTX 5090 rumors: GDDR7 on 512-bit bus, RTX 5080 will have 256-bit bus
1、NVIDIA下一代GeForce RTX 5090传闻配备512位内存总线和GDDR7显存。 2、预计RTX 5090在4K游戏性能上至少提升50%。 3、传闻RTX 5080配备GDDR7显存，性能提升20%+。 4、三星的GDDR7显存准备就绪，速度为28Gbps和32Gbps。 5、预计RTX 5090能够实现4K 240FPS游戏。 6、新GPU在光线追踪性能上预计获得2-3倍提升。 7、DLSS 4将在RTX 50系列上提供增强的图像质量和性能。 8、与当前RTX 4090和RTX 4080 SUPER相比，能效提升。1. NVIDIA's next-gen GeForce RTX 5090 rumored with 512-bit memory bus and GDDR7 memory. 2. RTX 5090 expected to deliver at least a 50% improvement in 4K gaming performance. 3. RTX 5080 rumored to have 20%+ performance improvements with GDDR7 memory. 4. Samsung's GDDR7 memory ready with speeds of 28Gbps and 32Gbps. 5. RTX 5090 anticipated to achieve 4K 240FPS gaming. 6. Ray tracing performance expected to see 2-3x gains with the new GPUs. 7. DLSS 4 to offer enhanced image quality and performance on the RTX 50 series. 8. Improved power efficiency compared to the current RTX 4090 and RTX 4080 SUPER.
06/12/2024, 12:18 AM UTC
NVIDIA据报道在2023年拥有98%的AI GPU市场份额，出货量接近400万NVIDIA reportedly has 98% of the AI GPU market, ships almost 4 million AI GPUs in 2023
1. NVIDIA据报道在2023年出货量达到376万台数据中心GPU，拥有98%的AI GPU市场份额。 2. 该公司从AI GPU销售中获得的收入达到了362亿美元，是2022年收入的三倍。 3. TechInsights分析师James Sanders讨论了谷歌TPU和其他AI处理器对市场的影响，指出尽管谷歌已成为一个重要的参与者，但云平台将继续对英 特尔、AMD和NVIDIA的芯片有需求。1. NVIDIA reportedly shipped 3.76 million data center GPUs in 2023, with a 98% market share in AI GPU shipments. 2. The company's revenue from AI GPU sales reached $36.2 billion, three times the revenue of 2022. 3. TechInsights analyst James Sanders discussed the impact of Google's TPUs and other AI processors on the market, noting that while Google has become a significant player, cloud platforms will continue to have demand for chips from Intel, AMD, and NVIDIA.
06/08/2024, 03:31 PM UTC
NVIDIA的RTX 5090和5080以及AMD的RDNA 4 GPU可能会推迟到2025年初发布NVIDIA's RTX 5090 and 5080 could be delayed to early 2025 - as well as AMD's RDNA 4 GPUs?
1、报道显示，AMD和NVIDIA的下一代GPU可能会面临延迟，潜在的发布时间可能要到2025年。 2、AMD的RDNA 4和Intel的Battlemage GPU也在讨论之中，最有可能的情况是在2025年的CES上首次亮相，尽管可能会在2024年底宣布。 3、NVIDIA的Blackwell GPU，包括RTX 5090和5080，也声称将推迟到2025年初，2024年底可能会有“纸上发布”。 4、这些潜在延迟背后的原因尚属猜测，但专注于更有利可图的AI GPU和高端市场缺乏竞争可能是NVIDIA考虑的因素。1. Reports suggest that next-gen GPUs from both AMD and NVIDIA could face delays, with potential launches not happening until 2025. 2. AMD's RDNA 4 and Intel's Battlemage GPUs are also mentioned, with a CES 2025 debut being the most likely scenario, although an announcement could come late 2024. 3. NVIDIA's Blackwell GPUs, including the RTX 5090 and 5080, are claimed to be delayed to early 2025 as well, with possible 'paper launches' happening later in 2024. 4. The reasons behind these potential delays are speculative, but the focus on more profitable AI graphics cards and lack of competition in the high-end market could be factors for NVIDIA.
05/24/2024, 09:19 AM UTC
Nvidia 第一季度营收飙升 262% 因数据中心需求Nvidia sees Q1 revenue jump 262% from data center demand
1、Nvidia 第一季度营收达到 26 亿美元，同比增长 262%。 2、数据中心业务单元贡献了大部分收入，达到 22.6 亿美元。 3、Hopper GPU Chips 的销售是收入增长的主要原因，包括 H100 和 B200。
04/19/2024, 12:25 PM UTC
Galax launches single-slot RTX 4060 Ti — uses a very slim blower design
We reported on the sighting of a single-slot Galax GeForce RTX 4060 Ti 16GB blower graphics card last October, and now one has appeared in a Chinese tech site review.
04/19/2024, 12:00 PM UTC
Razer Blade 14 (2024) review: Small frame, full power, high price
The Razer Blade 14 is the most powerful 14-inch gaming laptop we've tested, and it has a bright display and lots of ports. But it's far more expensive than the competition.
04/19/2024, 12:00 AM UTC
这个小型GPU是如何创造未来的**How this tiny GPU invented the Future**
A tiny GPU with a big secret: how AMD's Fiji invented the future as the first HBM + silicon interposer GPU for the mass market.
