Author page description
July 18
- $6 million Bitcoin theft lands cop five and a half years in prison — officer stole 50 Bitcoins from Silk Road 2.0 founder during investigation, inside job unravelled by the man he was trying to put away
➀ A former UK National Crime Agency officer stole 50 bitcoins (now worth $6M) during Silk Road 2.0 investigation;
➁ The theft remained undetected for 4 years until the case's prime suspect exposed the missing funds;
➂ Digital evidence from hidden notebooks and iPhone led to the officer's conviction and 66-month sentence.
- TSMC's quarterly sales hit a record $30 billion — contract chipmaker plans over 15 new fabs to meet still-growing AI demand
➀ TSMC reported a record $30.07 billion revenue in Q2 2025, driven by AI processor demand;
➁ Plans to build 15+ new fabs worldwide, including 11 in Taiwan and overseas expansions;
➂ N2 node enters mass production in late 2025, with AI-focused A16 node on track for 2026.
- Input latency is the all-too-frequently missing piece of framegen-enhanced gaming performance analysis
➀ Input latency remains critical in framegen-enhanced gaming, tied to base frame rates despite boosted visual smoothness;
➁ Introducing MALT (Maximum Acceptable Latency Threshold) as a key metric for evaluating framegen experiences, showing 45-60 ms as the usability range;
➂ Testing reveals higher-end GPUs enable better image quality at equivalent latency, challenging Nvidia’s marketing claims about low-end cards surpassing previous-gen flagships.
- Donkey Kong Bananza among Switch 2 games with no DLSS support — reviewers balk at Nintendo's aversion to technology
① Nintendo Switch 2's DLSS support, powered by NVIDIA, is absent in major titles like Donkey Kong Bananza and Mario Kart World, raising questions about developer adoption;
② Performance issues in Bananza are mitigated by AMD's FSR 1 and SMAA, but reviewers criticize the lack of HDR and modern upscaling tech;
③ Future DLSS implementation remains uncertain, with Metroid Prime 4's potential use of the tech still unconfirmed.
- Japanese chipmaker Rapidus begins test production of 2nm circuits — company commits to single-wafer processing ahead of 2027 mass production target
➀ Rapidus initiates 2nm GAA transistor test wafer production with successful electrical characteristics validation;
➁ Pioneers full single-wafer processing for all frontend steps to enhance precision and defect control using AI-driven optimization;
➂ Process Development Kit (PDK) launch scheduled for Q1 2026, supporting client chip design validation at IIM-1 facility.
July 17
- Synopsys acquires simulation specialist Ansys for $35 billion following Chinese regulator approval — Merger to power end-to-end design platform
➀ Synopsys obtained Chinese regulatory approval for its $35 billion acquisition of Ansys, creating a comprehensive design platform spanning chip and system simulation;
➁ The merged entity integrates Synopsys' EDA tools with Ansys' multiphysics simulation, enabling AI-driven co-optimization for chips, AI processors, and automotive systems;
➂ Regulatory scrutiny persists with China mandating license flexibility and interoperability, potentially reshaping the global EDA competitive landscape.
- This RTX 5090 is cheaper than anything we saw on Prime Day and isn't even discounted — grab Zotac's triple fan beast for just $2,499
➀ Zotac's GeForce RTX 5090 hits a new low price of $2,499, undercutting recent Prime Day deals;
➁ The card boasts 32GB GDDR7, 21,760 CUDA cores, and 25% performance gains over RTX 4090;
➂ Despite the price drop, it still exceeds Nvidia's unattained $2,000 MSRP target for Founders Edition.
- Analogue 3D Nintendo 64 remake delayed again — tariffs blamed for shipment pushback to ‘late August’
➀ The Analogue 3D Nintendo 64 remake faces another delay, now pushed to late August due to recent tariff changes, though pre-order prices remain unchanged;
➀ The FPGA-powered console promises 100% cartridge compatibility and 4K HDMI output while retaining original controller ports;
➂ Limited production strategy and high modding costs for original N64 hardware drive interest in this modern retro revival.
- Save £192 off this Herman Miller Vantum gaming chair — your back and your wallet will thank you
➀ The Herman Miller Vantum Gaming Chair is discounted by 20%, now priced at £765 with £192 savings;
➁ Praised for ergonomic design and lightweight build, it offers exceptional comfort and posture support for gamers;
➂ Originally co-designed with Logitech, this Editor's Choice award-winning chair merges office ergonomics with gaming aesthetics.
- AMD quietly reveals cheapest Ryzen AI yet — AI 5 330 is a quad-core budget processor with a 50 TOPS NPU
➀ AMD expands Ryzen AI 300-series with the budget-friendly Ryzen AI 5 330 featuring a 50 TOPS NPU for Copilot+ PCs;
➁ The quad-core CPU operates at 2.0–4.5 GHz with a Radeon 820M iGPU, prioritizing AI compliance over raw performance;
➂ Targets cost-sensitive markets to boost AMD's AI PC footprint using scaled-down Zen 5/c cores and lower-tier GPU configurations.
July 16
- Tariff fears caused a surge in entry-level 3D printer sales — Chinese companies accounted for 95% of entry-level machines shipped globally
➀ Entry-level 3D printer shipments surged 15% YoY in Q1 2025, with Chinese companies dominating 95% of the global market.
➁ Bambu Lab's sales grew 64% YoY, while Creality retained a 39% market share despite a slight decline.
➂ 3D printing is increasingly used for finished product manufacturing in medical and automotive industries, beyond just prototyping.
- Intel releases new tool to measure gaming image quality — AI tool measures impact of upscalers, frame gen, others; Computer Graphics Video Quality Metric now available on GitHub
➀ Intel released CGVQM, an AI-driven metric for objectively evaluating real-time gaming image quality, addressing distortions from upscalers and frame generation;
➁ The tool combines a new human-rated dataset (CGVQD) and a 3D-CNN model, outperforming existing metrics in tests;
➂ It generalizes well to new videos and could standardize image quality analysis in game development.
- HP OmniBook X Flip 14 review: Speedy and spicy
➀ The HP OmniBook X Flip 14 delivers strong performance with AMD Ryzen AI 7 350 and 32GB RAM, outperforming Intel-based rivals in benchmarks.
➁ Its 14-inch OLED display offers vivid colors but has moderate brightness, while the lattice-free keyboard may require adaptation for typing accuracy.
➂ Despite ample ports and a premium build, the laptop suffers from short 7.5-hour battery life and noticeable heat under heavy workloads.
- Firefox dev says Intel Raptor Lake crashes are increasing with rising temperatures in record European heat wave — Mozilla staff's tracking overwhelmed by Intel crash reports, team disables the function
➀ Mozilla's Firefox crash reports surged due to Intel Raptor Lake CPU instability worsened by European heatwaves;
➀ Intel's microcode updates only mitigate, not fix, the physical degradation issue, with latest patch 0x12F causing renewed instability;
➁ Intel extended warranties for affected CPUs to 5 years, but AMD gains market share with competitive Ryzen chips.
July 15
- Noctua NH-P1 Review: For those who demand silence
➀ Noctua NH-P1 is a fanless CPU cooler offering silent operation and robust thermal performance for non-overclocking users;
➁ Compatible with Intel and AMD systems but faces GPU slot limitations on AMD platforms;
➂ Includes premium accessories like NT-H2 thermal paste and supports optional fan upgrades for enhanced cooling.
- Meta is building AI data centers in tents and isn't slowing down — Zuckerberg reveals plans for 'several multi-GW clusters,' including one called Hyperion that's almost as big as Manhattan
➀ Meta accelerates AI infrastructure with 'tented' data centers to reduce costs and deployment time, aiming for gigawatt-scale clusters like Prometheus (1GW) and Hyperion (5GW);
➁ Plans include eliminating backup power redundancy and leveraging on-site natural gas plants, despite risks of extreme weather;
➂ Zuckerberg positions Meta as a future AI compute leader, potentially pivoting to infrastructure services amid competition from NVIDIA and xAI.
- UK Nintendo Switch 2 owners can grab PNY's 256GB microSD Express card for £59, finally back in stock
➀ PNY's 256GB microSD Express card for Nintendo Switch 2 is back in stock at Amazon UK for £59;
➁ Delivering SSD-level speeds up to 890MB/s read with PCIe Gen3x1 NVMe tech;
➂ Doubles the console's internal storage amid ongoing supply shortages for Switch 2 accessories.
- Nintendo Switch 2 account bans continue — Content creator with over a million subs issues warning after buying an old copy of Bayo 3 on eBay
➀ Nintendo Switch 2's aggressive anti-piracy measures led to a console ban for YouTuber Quinn Nelson after using a second-hand Bayonetta 3 cartridge flagged as pirated;
➁ The ban restricts all online services access while retaining offline gameplay functionality, highlighting Nintendo's hardware-level DRM enforcement;
➂ Controversy arises as critics question consumer rights violations, with some pointing to Nelson's past use of MIG Flash cart for legitimate game backups as possible contributing factor.
- Nvidia to resume H20 sales in China — says U.S. government has promised to grant licenses, deliveries to start soon
➀ The U.S. will allow NVIDIA to resume H20 AI GPU sales to China after a $5.5B export ban loss, with licenses pending approval;
➁ NVIDIA’s China-specific H20 faced prior restrictions under both Biden and Trump administrations, despite CEO Jensen Huang's lobbying against export controls;
➂ The move follows U.S.-China trade talks and rare earth/EDA concessions, though H20 sales remain conditional, pushing NVIDIA to develop compliant alternatives like RTX Pro GPUs.
July 14
- Google Gemini crumbles in the face of Atari Chess challenge — admits it would 'struggle immensely' against 1.19 MHz machine, says canceling the match most sensible course of action
➀ Google Gemini backed out of a chess match with a 1.19 MHz Atari 2600 after realizing its overconfidence, citing potential struggles despite previous boasts of evaluating millions of moves.
➀ The AI’s hesitation mirrors similar behavior from ChatGPT and Microsoft Copilot, raising questions about LLMs’ limitations in strategic games despite advanced language capabilities.
➀ Experts highlight that LLMs excel in linguistic tasks but lack specialized chess algorithms, as demonstrated by the Atari’s 128-byte RAM outclassing complex AI models in this niche challenge.
- TensorWave just deployed the largest AMD GPU training cluster in North America — features 8,192 MI325X AI accelerators tamed by direct liquid-cooling
➀ TensorWave deployed North America's largest AMD GPU cluster with 8,192 MI325X accelerators using direct liquid cooling;
➁ MI325X features 256GB HBM3e memory, competes with NVIDIA H200 at lower cost but faces scaling limitations;
➂ TensorWave plans MI350X integration, signaling AMD's growing foothold in AI infrastructure.
- Snot-filtering tech could be the answer to a dust-free PC — Korean scientists turn to nature to improve air filtration
➀ Korean scientists developed a bio-inspired dust filter mimicking nasal mucus-coated hairs, with 2-3X longer effectiveness than traditional filters, capturing more particulate matter (PM);
➁ The ‘PRO filter’ uses biocompatible silicone oil as a mucus substitute, minimizing particle redispersion and allowing reusable washing/reapplication;
➂ Target applications extend beyond PCs to air conditioners, data centers, and hospitals, offering cost-effective industrial air purification.
- Amazon's $8 billion Anthropic investment rumors suggest it would rather sell AI infrastructure than compete with ChatGPT and Gemini
➀ Amazon reportedly plans an additional $8 billion investment in Anthropic, aiming to dominate AI infrastructure through AWS rather than consumer-facing AI products;
➁ AWS generates $30 billion in Q1 2025, leveraging its cloud infrastructure to power AI development while rivals like Google and Meta focus on consumer chatbots;
➂ Amazon develops custom AI chips for inference workloads, positioning itself as an "AI shovel seller" in the Nvidia-dominated market.
July 13
- Nvidia's new driver update finally brings Smooth Motion to RTX 40-series GPUs, works like AMD's Fluid Motion Frames and claims to double your FPS with a single click in any game
① NVIDIA's 590.26 Preview Driver enables Smooth Motion on RTX 40-series GPUs, mimicking AMD's Fluid Motion Frames to double FPS via AI-generated interpolated frames;
② The driver-level solution works universally across games without developer integration, achieving 82→164 FPS boosts in titles like World of Warcraft with minimal latency;
③ Currently a developer preview requiring Nvidia Profile Inspector tweaks, the feature shows promise despite potential artifacts, offering a free alternative to third-party upscaling tools.
- Holographic ribbon aims to oust magnetic tape with 50-year life span and 200TB capacity per cartridge — HoloMem says optical ribbon-based carts work with some components of existing systems, reducing fricition
➀ HoloMem introduces holographic ribbon storage with 200TB capacity and 50-year lifespan, surpassing LTO-10 magnetic tape's 18TB and 5-year benchmarks;
➀❷ The system uses low-cost components like $5 laser diodes and polymer ribbons, enabling seamless integration with legacy tape libraries;
➂ Competitors like Cerabyte and Microsoft Project Silica face higher costs and technical barriers, while HoloMem secures Intel Ignite backing and prepares real-world testing in UK data centers.
- McDonald's McHire bot exposed personal information of 64M people by using '123456' as a password in 2025
➀ Security researchers exposed vulnerabilities in McDonald's McHire chatbot using password '123456', compromising 64 million applicants' data;
➀ The breach revealed names, contact details, work preferences, and authentication tokens through API flaws;
➂ Paradox resolved the vulnerabilities within a day after disclosure, highlighting systemic security failures in enterprise systems.
- Security vulnerability on U.S. trains that let anyone activate the brakes on the rear car, was known for 13 years — operators refused to fix the issue until now
➀ A wireless security vulnerability in U.S. trains, allowing unauthorized brake activation via $500 hardware, has persisted since 2012 despite researcher warnings;
➁ The American Association of Railways (AAR) dismissed the issue as 'theoretical' for over a decade until CISA's 2025 public advisory forced action;
➂ Implementation of a patch is delayed until 2027, raising concerns about rail safety and systemic negligence in critical infrastructure.
July 12
- Prime Day is officially over, but here are some great deals you can still get at Amazon, Newegg, and others
➀ Prime Day saturation sales have ended but notable deals persist across Amazon/Newegg with up to 50% discounts;
➁ Top-tier hardware bargains span RTX 5070 graphics cards at $530, 4TB Gen5 SSDs below $0.06/GB, and 49-inch 240Hz OLED monitors under $1,100;
➂ Flagship components like AMD's 128MB V-Cache Ryzen 9 9950X3D hit record lows while essential tools/small form factor solutions showcase broad price cuts.
- New leaked Bambu Lab 3D printer diagrams reveal H2S model in the works — diagrams accidentally included in Wiki page
➀ A leaked diagram in Bambu Lab's official Wiki revealed the upcoming H2S 3D printer, featuring a single-nozzle design and similarities to the A1 tool head.
➁ The H2S may offer a larger build height (340mm vs. 325mm) compared to the H2D, positioned as a cost-effective alternative to the laser-focused $3,649 H2D model.
➂ Industry observers speculate the H2S follows Bambu Lab's strategy of releasing simplified, affordable versions after flagship models (similar to P1P/P1S after X1 Carbon).
- AMD's 96-core Ryzen Threadripper Pro 9995WX benchmark results listed in Geekbench database
➀ AMD's 96-core Threadripper Pro 9995WX shows a 14% single-core performance gain over its predecessor in Geekbench, but still lags behind Apple's M4 Max;
➁ In multi-core tests, the CPU outperforms its Zen 4-based predecessor by 15% but faces close competition with Apple's M4 Max due to Geekbench's workload limitations;
➂ Geekbench's short-duration tests fail to fully utilize the 96-core design, favoring consumer CPUs with higher clocks for burst performance.