Login
Intel-Xeon-6

Intel Xeon 6 Boosts AI Performance in MLPerf Benchmarks

Intel has made a significant leap in artificial intelligence (AI) performance with the latest results from the MLPerf Inference v4.1 benchmark. This industry-standard test suite, developed by MLCommons, measures the performance of AI solutions across various applications. In this edition, Intel’s 5th generation Intel® Xeon® processors and, for the first time, the new Intel Xeon 6 processors with performance cores (P-cores) have demonstrated substantial improvements in AI performance.

The Intel Xeon 6 processors have achieved up to 1.9 times better AI performance compared to the previous generation, and up to 17 times better over a four-year period. These impressive results reinforce Intel’s position as a leader in delivering CPU solutions for high-performance AI applications.

Intel Xeon 6 Technological Advances: A Game Changer in AI

Intel showcased six performance tests in MLPerf, covering models such as ResNet50, RetinaNet, 3DUNet, BERT, DLRM v2, and GPT-J. These benchmarks highlight the enhanced capabilities of the Xeon 6 processors, now offering unprecedented performance across various AI tasks, from natural language processing to computer vision.

Pallavi Mahajan, Intel’s Corporate Vice President and General Manager of Data Center and AI Software, stated: “The latest MLPerf results demonstrate that continuous investment and resourcing are key to improving AI performance. Over the past four years, we’ve raised the bar on AI performance in Intel Xeon processors up to 17 times, as evidenced by MLPerf.”

The success of the Xeon 6 lies in innovations like Intel® Advanced Matrix Extensions (AMX), which have significantly improved AI processing efficiency by introducing new data types. These enhancements have enabled the Xeon 6 to deliver up to 17 times better AI performance in natural language processing, as seen in BERT, and up to 15 times better in computer vision tasks, such as ResNet50.

Significance of MLPerf Results and Strategic Collaborations

The MLPerf results not only underscore Intel’s leadership in the sector but also highlight the importance of robust CPUs in effectively deploying AI solutions. Intel is currently the only server processor provider presenting CPU results in MLPerf, underscoring its commitment to continuous AI performance improvement.

Furthermore, Intel has closely collaborated with its OEM partners, including Cisco, Dell Technologies, HPE, Quanta, and Supermicro, to ensure that its Xeon processors are compatible with a wide range of AI workloads. This strategic alliance allows customers to run AI applications alongside their traditional business operations, optimizing the use of existing infrastructure.

Looking ahead, Intel has announced that it will reveal more details about the Xeon 6 processors with P-cores at an upcoming launch event. The company expects these processors to become a benchmark in the industry, providing businesses with powerful tools to enhance their AI infrastructure.

In conclusion, the MLPerf Inference v4.1 results not only highlight the evolution of Intel Xeon processors towards more efficient and powerful AI solutions but also reaffirm Intel’s commitment to continuous innovation and support for its OEM partners. With 1.9 times better AI performance in the latest tests and significant improvements over the past four years, the Intel Xeon 6 processors are positioning themselves as the preferred choice for companies looking to maximize their AI capabilities. Will Intel maintain its leadership in the fast-paced race of AI performance in the coming years?

REQUEST FOR A QUOTE