➀ TPU is a specialized computing core for matrix multiplication (TensorCore) connected to a stack of fast memory (HBM);

➁ TensorCore consists of MXU (Matrix Multiplication Unit), VPU (Vector Processing Unit), and VMEM (Vector Memory);

➂ TPU is very fast in matrix multiplication and has high FLOPs/s;

➃ TPU has various network configurations like ICI and DCN for efficient communication.