MLCommons just released new MLPerf Inference v5.0 benchmark results. These tests show Intel Xeon 6 chips with Performance Cores running six important AI tests. The numbers prove that these processors work almost twice as fast as older versions. Intel Xeon 6 stands out as a top choice for companies building modern AI systems.
Karin Eibschitz Segal from Intel says Xeon 6 hits the sweet spot between fast performance and power saving. She heads the Data Center and AI team as corporate vice president. Her team keeps making each new generation of Intel chips better at AI tasks. They focus on helping businesses run AI more efficiently.
Companies need powerful CPUs as they add more AI to their products. These central processors handle many critical jobs like preparing data and moving information between system parts. Intel remains the only company that sends server CPU results to the MLPerf testing group. They keep showing up with test results when others don't.
The latest tests reveal that Intel Xeon 6 with P-cores runs almost twice as fast as 5th-generation Intel Xeon chips. This speed boost appears across many tests, including ResNet50, RetinaNet, 3D-UNet, and the brand new GNN-RGAT. These results make Intel Xeon 6 attractive for AI tasks and smaller language models.
Intel has made huge progress in AI speed during recent years. Their first MLPerf submission used 3rd Gen Xeon chips back in 2021. Since then, they've achieved fifteen times faster performance on the ResNet50 test. Better software added another 22% speed increase for GPT-J and 11% faster results on 3D U-Net tests.
These benchmark results come from many computer makers using Intel chips. Four major partners—Cisco, Dell Technologies, Quanta, and Supermicro—sent in test results using Intel Xeon 6 with P-cores. They showed that these chips excel at different AI tasks across various system designs. As AI becomes part of everyday computing, more businesses want Xeon systems.
Karin Eibschitz Segal from Intel says Xeon 6 hits the sweet spot between fast performance and power saving. She heads the Data Center and AI team as corporate vice president. Her team keeps making each new generation of Intel chips better at AI tasks. They focus on helping businesses run AI more efficiently.
Companies need powerful CPUs as they add more AI to their products. These central processors handle many critical jobs like preparing data and moving information between system parts. Intel remains the only company that sends server CPU results to the MLPerf testing group. They keep showing up with test results when others don't.
The latest tests reveal that Intel Xeon 6 with P-cores runs almost twice as fast as 5th-generation Intel Xeon chips. This speed boost appears across many tests, including ResNet50, RetinaNet, 3D-UNet, and the brand new GNN-RGAT. These results make Intel Xeon 6 attractive for AI tasks and smaller language models.
Intel has made huge progress in AI speed during recent years. Their first MLPerf submission used 3rd Gen Xeon chips back in 2021. Since then, they've achieved fifteen times faster performance on the ResNet50 test. Better software added another 22% speed increase for GPT-J and 11% faster results on 3D U-Net tests.
These benchmark results come from many computer makers using Intel chips. Four major partners—Cisco, Dell Technologies, Quanta, and Supermicro—sent in test results using Intel Xeon 6 with P-cores. They showed that these chips excel at different AI tasks across various system designs. As AI becomes part of everyday computing, more businesses want Xeon systems.
