At this week’s Computex in Taipei, Intel unveiled technologies and architectures to accelerate the AI ecosystem from the data centre, cloud and network to the edge and PC.

“AI is driving one of the most consequential eras of innovation the industry has ever seen,” said Intel CEO Pat Gelsinger in his keynote speech. “The magic of silicon is once again enabling exponential advancements in computing that will push the boundaries of human potential and power the global economy for years to come.”

He said Intel was one of the only companies in the world innovating across the full spectrum of the AI market opportunity from semiconductor manufacturing to PC, network, edge and data centre.

“Our latest Xeon, Gaudi and Core Ultra platforms, combined with the power of our hardware and software ecosystem, are delivering the flexible, secure, sustainable and cost-effective options our customers need to maximise the immense opportunities ahead,” Gelsinger said, highlighting the benefits of open standards and Intel’s ecosystem helping to accelerate the AI opportunity. He was joined on stage by Acer CEO Jason Chen, Asus Chairman Jonney Shih, Microsoft CEO Satya Nadella and Inventec president Jack Tsai.

In six months, Intel went from launching fifth-generation Xeon processors to introducing the inaugural member of the Xeon 6 family; from previewing Gaudi AI accelerators to offering enterprises a cost-effective generative AI (genAI) training and inference system; and from ushering in the AI PC era with Core Ultra processors in more than eight million devices to unveiling the forthcoming client architecture slated for release later this year.

The entire Xeon 6 (download.intel.com/newsroom/2024/client-computing/Xeon-6-Fact-Sheet.pdf) platform and family of processors can address a broad array of use cases and workloads, from AI and other high-performance compute needs to scalable cloud-native applications. Both E-cores and P-cores are built on a compatible architecture with a shared software stack and an open ecosystem of hardware and software vendors.

The first of the Xeon 6 processors to debut is the Xeon 6 E-core (code-named Sierra Forest), which is available now; Xeon 6 P-cores (code-named Granite Rapids) are expected to launch next quarter.

Additionally, Xeon 6 E-core has density advantages, enabling rack-level consolidation of three-to-one, providing a rack-level performance gain of up to 4.2x and performance per watt gain of up to 2.6x when compared with second-generation Xeon processors on media transcode workloads.

Harnessing the power of genAI is becoming faster and less expensive. As the dominant infrastructure choice, x86 operates at scale in nearly all data centre environments, serving as the foundation for integrating the power of AI while ensuring cost-effective interoperability and the benefits of an open ecosystem of developers and users.

Xeon processors are a suitable CPU head node for AI workloads and operate with Gaudi AI accelerators, which are purposely designed for AI workloads. Together, these two can seamlessly integrate into existing infrastructure.

As the only MLPerf-benchmarked alternative to Nvidia H100 for training and inference of large language models (LLM), the Gaudi architecture gives users genAI performance with a price-performance advantage that provides choice and fast deployment time at lower operating cost.

Gaudi 3 in an 8192-accelerator cluster is projected to offer up to 40% faster time-to-train versus the equivalent size Nvidia H100 GPU cluster and up to 15% faster training throughput for a 64-accelerator cluster versus Nvidia H100 on the Llama2-70B model. In addition, Gaudi 3 is projected to offer an average of up to two times faster inferencing versus Nvidia H100, running popular LLMs such as Llama-70B and Mistral-7B.

To make these AI systems broadly available, Intel is collaborating with at least ten system providers, including six new providers who announced they’re bringing Gaudi 3 to market. Collaborators include Asus, Foxconn, Gigabyte, Inventec, Quanta and Wistron, expanding the production offerings from Dell, Hewlett Packard Enterprise, Lenovo and Supermicro.

Beyond the data centre, Intel is scaling its AI footprint at the edge and in the PC. With more than 90,000 edge deployments and 200 million CPUs delivered to the ecosystem, Intel has enabled enterprise choice for decades.

AI PCs are projected to make up 80% of the PC market by 2028, according to Boston Consulting Group. In response, Intel has moved quickly to create a hardware and software platform for the AI PC, enabling more than 100 independent software vendors (ISVs), 300 features (www.intel.com/content/www/us/en/newsroom/news/intel-launches-ai-pc-acceleration-program.html) and support of 500 AI models (www.intel.com/content/www/us/en/newsroom/news/more-than-500-ai-models-run-optimized-on-intel-core-ultra-processors.html) across its Core Ultra platform.

Building on these, the company also revealed the architectural details of Lunar Lake (download.intel.com/newsroom/2024/client-computing/Lunar-Lake-Architecture-Fact-Sheet.pdf), its flagship processor for AI PCs. Lunar Lake should deliver up to 40% lower SoC power and more than three times the AI compute. It’s expected to ship in the third quarter of 2024.