Intel launches Gaudi 3 AI chip as Nvidia competition intensifies | Wilnesh News

Intel On Tuesday, Intel launched its latest artificial intelligence chip, Gaudi 3, as chipmakers rush to produce semiconductors that can train and deploy large-scale artificial intelligence models, such as models that support OpenAI’s ChatGPT.

Intel said the new Gaudi 3 chips are more than twice as energy efficient as conventional chips and can run artificial intelligence models one and a half times faster. NVIDIA H100 GPU. It also comes in different configurations, such as eight Gaudi 3 chips bundled on a motherboard or a card that can be plugged into an existing system.

Intel tested the chip on the following models Yuan Open source Llama and Abu Dhabi-powered Falcon. It is said that Gaudi 3 can assist in training or deploying models, including Stable Diffusion for speech recognition or OpenAI’s Whisper model.

Intel says its chips consume less power than Nvidia’s chips.

NVIDIA has one estimated Graphics processing units (GPUs) account for 80% of the artificial intelligence chip market. In the past year, GPUs have been the high-end chip of choice for artificial intelligence builders.

Intel said the new Gaudi 3 chips will be available to customers in the third quarter, including Dell, life value and Super Micro The system will be built using chips. Intel did not provide a price range for Gaudi 3.

Das Kamhout, Intel’s vice president of Xeon software, said in an interview with reporters, “We do expect it to be very competitive with Nvidia’s latest chips.” “We are using industry-standard Ethernet based on our competitively priced, unique open integrated network-on-chip. We believe this is a strong product.”

The data center AI market is also expected to grow as cloud providers and enterprises build the infrastructure to deploy AI software, suggesting that even if Nvidia continues to produce the vast majority of AI chips, there is room for other competitors.

The cost of running generative AI and purchasing Nvidia GPUs can be high, and the company is looking for more suppliers to help reduce costs.

The artificial intelligence boom has more than tripled Nvidia’s stock price over the past year. Intel’s stock price rose only 18% during the same period.

AMD is also looking to expand and sell more artificial intelligence chips for use in servers. Last year, it launched a new data center GPU called the MI300X, which already counts Meta and Microsoft as customers.

Earlier this year, Nvidia released the B100 and B200 GPUs, successors to the H100 that also promise performance improvements. The chips are expected to begin shipping later this year.

Nvidia owes its success to a powerful suite of proprietary software called CUDA, which gives AI scientists access to all the hardware capabilities in GPUs.Intel is working with other chip and software giants, including Google, Qualcomm and arm Create non-proprietary open software that allows software companies to easily switch chip providers.

“We are working with the software ecosystem to create open reference software and build modules so you can piece together the solutions you need rather than being forced to buy solutions,” said Sachin Katti, senior vice president of networking at Intel. “The group said in an interview with reporters.

Gaudi 3 is manufactured using a 5nm process, a relatively new manufacturing technology that suggests the company is using external foundries to manufacture the wafers. Intel CEO Patrick Gelsinger told reporters last month that in addition to designing Gaudi 3, Intel plans to produce artificial intelligence chips at a new Ohio factory expected to open in 2027 or 2028, possibly for external use. Made by the company.