Intel Launches Data Center CPUs and GPUs | Knowledge of the data center

Intel has announced an update for their next-generation server chip – the Xeon CPU Max series and the Data Center GPU Max – in the hopes to capture market share from its competitor Advanced Micro Devices (AMD).

This follows Intel’s poor 2022 earnings report, targeted layoffs, restrictions on Chinese access to US semiconductor technologyand the return of President and CEO Pat Gelsinger.

Under Gelsinger, Intel worked to restore its lleadership in semiconductor process technology, an effort that required the company to retool factories for the expansion of its GPU business. R&D and production issues delayed the delivery of the Xeon CPU Max series, formerly called sapphire rapidsand the GPU Data Center previously known as Ponte Vecchio.

The delays have hampered Intel’s data center business, giving AMD 30.3% of the server processor market by Q3 2022. It could be difficult for Intel to quickly regain the lead. AMD is expected to launch its next-generation Epyc server processors before Intel releases the competing Xeon CPU and Data Center GPU by January 2023.

On top of Intel’s woes, Samsung overtakes them in 2021 to become world’s largest semiconductor vendor by revenue, according to preliminary results from research firm Gartner. The report adds that Samsung held 13% market share, compared to 12.5% ​​for Intel. Last year, Intel had the lowest growth rate among the top 25 vendors, at just 0.5% according to Gartner.

Gelsinger addressed these issues by saying that Intel is “coming to a better balance point between supply and demand.” in a third quarter earnings call.

“We’ve been way behind demand and supply for many, many consecutive quarters. Obviously the last two quarters have seen adjustments in inventory levels and we think we’re going to be in a better balanced position. between supply and demand early next year,” Gelsinger said.

Intel’s road to recovery included positives third quarter results with revenue of $15.3 billionand shares of Intel stock rebounded just before November 9 chip ad. Intel has also gained 8% since hitting a 52 week low on October 13. But that’s still $3.9 billion below third-quarter 2021 earnings and well below their historical norms.

How Intel is surviving in a market no longer dominated by PCs

Both the Xeon CPU Max and the Data Center GPU have come under intense scrutiny in the HPC industry due to production delays.. However, Intel said in late September that it was shipping blades with Max-series GPUs to Argonne National Laboratory to power the Aurora supercomputer.

Jeff McVeigh, vice president of Intel, said he wants to ensure that no HPC workload is left behind and that the chips offer “a solution that maximizes bandwidth, maximizes compute, maximizes developer productivity and ultimately maximizes impact.”

Vivek Arya, Securities Analyst at BofA says Intel’s stated goals are hard to achieve do everything stacked against a turnover. Arya adds that “the company’s goal of achieving process leadership by 2025 assumes flawless execution on five nodes in four years, which is difficult to achieve.”

Intel’s major customers, Apple and Microsoft, are compounding the revenue loss as they have switched to chips they started making themselves.

Still, Intel’s announcement of its GPU strategy seems like a powerful ploy to gain ground against AMD. Gelsinger has offered no other viable long-term solutions beyond big buyouts, cost-cutting measures and the divestment of its profitable NAND memory business.

Recently, the Biden administration banned chipmakers that receive funding from the US Flea and Science Act (CHIP), approved by Congress in August, to build advanced technology facilities in China. He is part of the US response to US companies demanding more government support to reduce reliance on components produced in Chinese factories. The law aims to increase domestic semiconductor manufacturing with a $50 billion investment in companies that request it.

Intel can regain market share here as fabless (outsourced manufacturing) chip companies like AMD and Nvidia design chips but do not manufacture them. Rivals would see no direct benefit from factory building subsidies or tax breaks for production tools. Intel has previously said it hopes to receive up to $3 billion for each new factory it builds in the United States and noted that the $3 billion per factory is a cap in the legislation.

What We Know About Intel’s Data Center GPU Max Series and Xeon Max Processor

Max Series products will also power several other HPC systems critical to national security and basic research, including Crossroads at Los Alamos National Laboratory, CTS-2 systems at Lawrence Livermore National Laboratory and Sandia National Laboratory, and Camphor3 at Kyoto University.

The Xeon Max processor is the first x86-based chip with high-bandwidth memory (HBM), which the company claims accelerates many HPC workloads without requiring code changes. The Max Series GPU is Intel’s highest-density processor, packing more than 100 billion transistors into a 47-tile package with up to 128 gigabytes of high-bandwidth memory (HBM).

The Xeon Max processor offers up to 56 performance cores made up of four tiles and connected using Intel’s multi-die interconnect bridge (EMIB) technology, in a 350-watt envelope. It packs 64GB of high-bandwidth onboard memory, along with PCI Express 5.0 and CXL1.1 I/O. The Xeon Max processors will provide more than 1 GB of HBM capacity per core, enough to accommodate the most common HPC workloads, according to Intel.

Soni Brown is a freelance content writer for Data Center Knowledge. In her editorial and reporting roles, she creates compelling content for data center industry stakeholders.

Donald E. Patel