PPU
Alibaba Cloud's in-house AI accelerator chip (Parallel Processing Unit), used for large model training and inference. It has been deployed at scale in autonomous driving and model training workloads.
Timeline (last 90 days)
Nearly 30 automakers and intelligent driving solution providers conduct intelligent driving R&D on Alibaba Cloud, actually using more than 100,000 PPU cards.
PPU is Alibaba Cloud's self-developed AI accelerator chip, used for large model training and inference.
Alibaba's subsidiary T-Head released a new generation of AI chips, the PPU series, with computing power approximately three times that of the previous generation, scalable to 500,000 cards in a single cluster, with mass production planned for Q1 of next year.
The NeSLE project implements the NES's 6502 CPU, PPU, and bus entirely as CUDA kernels, using a parallelism model of one thread per environment.
SNES的PPU芯片能同时处理128个精灵对象，每帧最多显示32个
All Facts (5)
Nearly 30 automakers and intelligent driving solution providers conduct intelligent driving R&D on Alibaba Cloud, actually using more than 100,000 PPU cards.
50%UnverifiedPPU is Alibaba Cloud's self-developed AI accelerator chip, used for large model training and inference.
50%UnverifiedThe NeSLE project implements the NES's 6502 CPU, PPU, and bus entirely as CUDA kernels, using a parallelism model of one thread per environment.
50%UnverifiedSNES的PPU芯片能同时处理128个精灵对象，每帧最多显示32个
50%UnverifiedAlibaba's subsidiary T-Head released a new generation of AI chips, the PPU series, with computing power approximately three times that of the previous generation, scalable to 500,000 cards in a single cluster, with mass production planned for Q1 of next year.
50%