Chip maker AMD released MI300x, its most advanced AI chip, on Tuesday, upping the ante in the super chip turf war with Nvidia. According to reports, the new super chip unveiled by AMD will power the artificial intelligence race as it can be applied to large language models (LLMs).
The new super chip unveiled by AMD is poised to grab market share from super chip rival Nvidia, which has rolled out its super chip called H100.
'Most Complex Chip'
"There's an incredible amount of technology on here — we're talking about the most complex chip we've ever built ... The whole purpose is to make AI much much more accessible. So everybody who wants to use AI needs more GPUs, and we have a GPU that is incredibly powerful; very, very efficient; and we believe will be a significant winner in the AI market," said AMD chairman and CEO Lisa Su, according to Yahoo Finance.
AMD's latest pitch in the artificial intelligence super chip arena is interesting as rival Nvidia is a giant that controls more than 90 percent of the market share.
Nvidia Surge
Nvidia has been on a roll ever since OpenAI released ChatGPT late last year, with the market capitalization zooming past the sensational $1 trillion mark. Nvidia's market capitalization is closing in on the $1 trillion mark, at which point the company will join the likes of Apple, Microsoft, Alphabet and Amazon in the market cap hall of fame. The company's stock has surged more than 170 percent so far this year, driven by the AI frenzy.
Right now the super chip scene is clearly dominated by Nvidia, though Intel Corp, along with startups like Cerebras Systems and SambaNova Systems, are looking to pose a threat. Also in the race lately are the internal chip development programs from Google and Amazon.
AMD Shares
In anticipation that the chip gambit might become an uphill climb, AMD's shares sank on Tuesday following the CEO's latest announcement.
Late last month Nvidia Corp said it would manufacture an array of highly advanced AI products including a new supercomputer. "We're now at the tipping point of a new computing era with accelerated computing and AI that's been embraced by almost every computing and cloud company in the world ... It will give nonplayable characters conversational skills so they can respond to questions with lifelike personalities that evolve," said Nvidia CEO Jensen Huang.
Nvidia expects tech giants Alphabet, Facebook parent Meta Platforms and Microsoft will deploy the DGX GH200. This superchip eliminates the need for a traditional CPU-to-GPU PCIe connection by combining an Arm-based NVIDIA Grace CPU with an NVIDIA H100 Tensor Core GPU in the same package, using NVIDIA NVLink-C2C chip interconnects. "This increases the bandwidth between GPU and CPU by 7x compared with the latest PCIe technology, slashes interconnect power consumption by more than 5x, and provides a 600GB Hopper architecture GPU building block for DGX GH200 supercomputers," the company says.
