NVIDIA's High-end AI Chip, H100, Priced Over $40,000 Due to Soaring Demand for AI Software Chips
According to CNBC, the demand for chips required for training and deploying artificial intelligence software is soaring, resulting in the price of NVIDIA's high-end chip H100 exceeding $40,000.
John Carmack, a 3D gaming pioneer and former Meta consulting CTO, mentioned the price of NVIDIA's H100 processor on Twitter. At least eight H100s were listed on eBay last Friday, with prices ranging from $39,995 to just under $46,000. Some retailers have previously offered the chip at around $36,000.
Released last year, the H100 is NVIDIA's latest flagship artificial intelligence chip, succeeding the A100. The A100, which costs around $10,000, is known as the "workhorse" for artificial intelligence applications.
Developers are using the H100 to build so-called Large Language Models (LLMs), which are the core of artificial intelligence applications such as OpenAI's ChatGPT. Running these systems is expensive and requires powerful computers to process thousands of bytes of data over several days or weeks. They also depend on immense computing power so that AI models can generate text, images, or predictions.
Training artificial intelligence models, particularly large models like GPT, requires hundreds of high-end NVIDIA GPUs to work together.
Microsoft spent hundreds of millions of dollars to purchase tens of thousands of NVIDIA A100 chips to help build ChatGPT. NVIDIA controls the vast majority of the market for artificial intelligence chips.
NVIDIA also offers a supercomputer called DGX that works with eight GPUs, which it calls a "transformer," optimized for a specific artificial intelligence architecture that supports many recent artificial intelligence advancements. Industry experts say that to build models larger and more data-hungry than current ones, more powerful chips are needed.
Earlier this year, the company announced new services that allow companies to rent DGX computers for $37,000 per month. According to this price, the system will use NVIDIA's earlier A100 processor.
