Intel tested the chip on models like Meta’s open-source Lama and Abu Dhabi-backed Falcon. It says Gaudi can help train or deploy 3 models, including Stable Diffusion or OpenAI’s Whisper model for speech recognition.
Intel says its chips use less power than Nvidia’s.
Nvidia has an estimated 80% of the AI chip market with its graphics processors, known as GPUs, which have been the top-tier chips for AI makers over the past year.
Intel said the new Gaudi 3 chips will be available to consumers in the third quarter, and companies including Dell, HP and Supermicro will build systems with the chips. Intel did not provide a price range for the Gaudi 3.
“We expect it to be very competitive with Nvidia’s latest chips,” Das Kamhot, vice president of Xeon software at Intel, said on a call with reporters. “From our competitive pricing, to our dedicated open integrated network on chip, we’re using industry-standard Ethernet. We believe it’s a strong offering.”
The data center AI market is also expected to grow as cloud providers and businesses build infrastructure to deploy AI software, suggesting there is room for other competitors even if Nvidia’s AI chips Continue to form the vast majority.
Running generative AI and buying Nvidia GPUs can be expensive, and companies are looking for additional suppliers to help keep costs down.
The AI boom has more than tripled Nvidia’s stock over the past year. Intel’s stock is up just 18% over the same period.
AMD is trying to expand and sell more AI chips for servers. Last year, it introduced a new data center GPU called the MI300X, which already counts Meta and Microsoft as customers.
Earlier this year, Nvidia revealed its B100 and B200 GPUs, which are the successors to the H100 and also promise a performance boost. These chips are expected to start shipping later this year.
Nvidia has been very successful thanks to a powerful suite of proprietary software called CUDA that enables AI scientists to access all the hardware features in the GPU. Intel is working with other chip and software companies, including Google, Qualcomm and Arm, to develop open software that is not proprietary and can enable software companies to easily switch chip suppliers.
“We’re working with the software ecosystem to create open reference software, as well as building blocks that allow you to assemble the solution you need, rather than forcing you to buy a solution,” Sachin Kutty, senior vice president of networking at Intel. The group said on a call with reporters.
The Gaudi 3 is built on a five-nanometer process, a relatively recent manufacturing technique, suggesting the company is using an outside foundry to make the chips. CEO Patrick Gelsinger told reporters last month that in addition to designing the Gaudi 3, Intel also plans to manufacture AI chips, possibly for outside companies, at a new factory in Ohio that could open in 2027 or 2028. I hope to open.