up to datePublished globallyStrongestAI ChipsWSE-3The chip uses an advanced 4 trillion transistor 5nm process and has amazing performance. A single machine can train a model with 24 trillion parameters, far exceeding NVIDIA H100 GPU.
The technical advantage of WSE-3 lies in its 900,000 AI cores and 44GB on-chip SRAM storage, which enable its peak performance to reach 125FP16PetaFLOPS, achieving single-chip cluster-level performance.
Cerebras' CS-3 supercomputer further improves the efficiency of large-scale AI model training, and can train next-generation cutting-edge large models that are ten times larger than GPT-4 and Gemini. It also greatly simplifies the training process and has excellent scalability. This innovation will greatly promote the development of the AI field and accelerate the pace of the global AI revolution.
The launch of the new generation AI chip WSE-3 marks a huge advancement in AI technology. Its advanced process and powerful performance make it a leader in the industry.
Moreover, the performance of a single machine has reached the cluster level, greatly improving the efficiency and speed of AI model training. Cerebras' CS-3 supercomputer has pushed AI model training to a new level, achieving unprecedented scale and speed. This technological advancement will have a profound impact on the entire AI industry and accelerate the application and development of AI technology in various fields.