Intel on Tuesday unveiled its latest artificial intelligence chip, called Gaudi 3, as chipmakers rush to produce semiconductors that can train and deploy big AI models, such as the one underpinning OpenAI’s ChatGPT.
Intel says the new Gaudi 3 chip is over twice as power-efficient as and can run AI models one-and-a-half times faster than Nvidia’s H100 GPU. It also comes in different configurations like a bundle of eight Gaudi 3 chips on one motherboard or a card that can slot into existing systems.
Intel tested the chip on models like Meta’s open-source Llama and the Abu Dhabi-backed Falcon. It said Gaudi 3 can help train or deploy 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 high-end chip of choice for AI builders over the past year.
Intel said that the new Gaudi 3 chips would be available to customers in the third quarter, and companies including Dell, HP and Supermicro will build systems with the chips. Intel didn’t provide a price range for Gaudi 3.
“We do expect it to be highly competitive” with Nvidia’s latest chips, said Das Kamhout, vice president of Xeon software at Intel, on a call with reporters. “From our competitive pricing, our distinctive 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 continues to make the vast majority of AI chips.
Running generative AI and buying Nvidia GPUs can be expensive, and companies are looking for additional suppliers to help bring costs down.
The AI boom has more than tripled Nvidia’s stock over the past year. Intel’s stock is only up 18% over the same time period.
AMD is also looking 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 performance gains. Those chips are expected to start shipping later this year.
Nvidia has been so successful thanks to a powerful suite of proprietary software called CUDA that enables AI scientists to access all the hardware features in a GPU. Intel is teaming up with other chip and software giants, including Google, Qualcomm and Arm to build open software that isn’t proprietary and could enable software companies to easily switch chip providers.
“We are working with the software ecosystem to build open reference software, as well as building blocks that allow you to stitch together a solution that you need, rather than be forced into buying a solution,” Sachin Katti, senior vice president of Intel’s networking group, said on a call with reporters.
Gaudi 3 is built on a five nanometer process, a relatively recent manufacturing technique, suggesting that the company is using an outside foundry to manufacture the chips. In addition to designing Gaudi 3, Intel also plans to manufacture AI chips, potentially for outside companies, at a new Ohio factory expected to open in 2027 or 2028, CEO Patrick Gelsinger told reporters last month.