At Nvidia’s GTC 2021 technology conference, its CEO Jen-Hsun Huang announced the launch of Nvidia’s first data center CPU called Grace. As a processor based on the Arm architecture, it will be oriented towards AI and high-performance computing applications, which is of epoch-making significance for NVIDIA. The market also reacted quickly, with shares of Nvidia rising, while shares of Intel and AMD both fell.
Intel still has a huge market share in the server processor field, more than 90%, and AMD holds the rest. Recently, AMD has also shown very strong competitiveness. The debut of the new generation of EPYC series server processors codenamed Milan has brought enormous pressure to Intel and is expected to further win the market. With the release of Nvidia’s Grace CPU, a new competitor has entered the field, and it already has two supercomputing customers, the Swiss National Supercomputing Center (CSCS) for building a supercomputer called Alps, and the other One is the Los Alamos National Laboratory in the United States, and the new system is expected to debut in 2023. For Intel, there are chasing troops in the front and interception in the back, and the rookie Grace CPU is coming.
For this product that will compete with x86 architecture server processors, as one of the competitors, Intel CEO Pat Gelsinger (Pat Gelsinger) quickly responded. During an interview with Yahoo Finance Channel, Pat Kissinger was asked about the future competition between the two parties when Nvidia launched a new CPU that will focus on accelerating AI applications. He said:
“Last week, we announced Ice Lake-SP, and the response was very positive. And with Ice Lake-SP, we have an extraordinary expansion in AI capabilities. It’s Nvidia responding to us, not us responding to them. Clearly Intel has taken the lead when it comes to the AI-enhanced CPU idea. We also have a product line from Habana Labs (a specialist AI chip maker that Intel acquired in 2019), which definitely offers a very aggressive path, Our cloud partnership with AWS is a testament to that. The CPU concept was born out of Intel, and we’re now gradually incorporating AI into it, and I hope this will be an area of our offense in the future, not an area of defense.”
Pat Kissinger then said that even if Nvidia released the Grace CPU, it was Nvidia that defended when Intel was on the offensive. Pat Kissinger also mentioned Intel’s third-generation Xeon Scalable processors as chips that can provide extraordinary AI capabilities, arguing that Intel is still a leader in this area.
According to Bloomberg, Hans Mosesmann, an analyst at Rosenblatt Securities, said in a research note that Nvidia attaches great importance to CPUs and will not be limited by Intel and AMD having x86, and believes that Nvidia’s level of platform innovation is incredible. . Rosenblatt Securities reiterated some time ago that it maintained a “buy” rating on Nvidia’s stock price with a price target of $650.
As Nvidia joins the battlefield, Intel should gradually increase its efforts, and AMD also plans to add more AI functions to the next-generation EPYC server processor code-named Genoa, which will make the competition in this field even more intense.
The Links: 1MBI600NN-060 LM64P101 GET PARTS