Intel Arc Pro B-Series GPUs and Xeon 6 Shine in MLPerf Inference v5.1

Published Date : 10/09/2025 

In the latest MLPerf Inference v5.1 benchmarks, Intel's Arc Pro B60 GPUs and Xeon 6 processors demonstrate significant performance and cost advantages over NVIDIA's RTX Pro 6000 and L40S in Llama 8B models. 

MLPerf v5.1 Benchmarks Highlight Intel's AI Inference Capabilities

Today, MLCommons has released the latest MLPerf Inference v5.1 benchmarks, which showcase Intel's GPU Systems featuring Intel® Xeon® with P-cores and Intel® Arc™ Pro B60 graphics. These systems, code-named Project Battlematrix, have demonstrated impressive performance and cost efficiency in Llama 8B models. Specifically, the Intel Arc Pro B60 offers performance per dollar advantages of up to 1.25x and up to 4x compared to NVIDIA RTX Pro 6000 and L40S, respectively.

A Powerful Validation of Intel's AI Strategy

According to Lisa Pearce, Intel corporate vice president and general manager of Software, GPU and NPU IP Group, the MLPerf v5.1 results are a strong validation of Intel's GPU and AI strategy. She stated, “Our Arc Pro B-Series GPUs, combined with a new inference-optimized software stack, enable developers and enterprises to develop and deploy AI-powered applications with inference workstations that are powerful, easy to set up, affordably priced, and scalable.”

Addressing the Needs of Modern AI Inference

Until recently, professionals seeking high inference performance without compromising data privacy or incurring heavy subscription costs had limited options. Intel's new GPU Systems, designed to meet the needs of modern AI inference, provide an all-in-one platform that combines full-stack validated hardware and software. These systems aim to simplify adoption and ease of use with a containerized solution built for Linux environments, optimized for multi-GPU scaling and PCIe P2P data transfers, and designed with enterprise-class reliability and manageability features such as ECC, SRIOV, telemetry, and remote firmware updates.

The Role of CPUs in AI Systems

CPUs continue to play a vital role in AI systems, serving as the orchestration hub for preprocessing, transmission, and overall system coordination. Intel has made significant improvements in CPU-based AI performance over the past four years, establishing Intel Xeon as the preferred CPU for hosting and managing AI workloads in GPU-powered systems. Notably, Intel Xeon 6 with P-cores achieved a 1.9x performance improvement generation-over-generation in MLPerf Inference v5.1.

Intel's Commitment to AI Inference

Intel remains the only vendor submitting server CPU results to MLPerf, demonstrating its leadership and deep commitment to accelerating AI inference capabilities across both compute and accelerator architectures. The latest MLPerf Inference v5.1 results further underscore Intel's dedication to providing powerful, accessible, and scalable solutions for AI inference workloads.

Additional Resources

For more detailed information, you can refer to the MLCommons MLPerf Inference v5.1 Results and the Project Battlematrix introduction.

Notices & Disclaimers

Performance varies by use, configuration, and other factors. Learn more at www.Intel.com/PerformanceIndex. Performance results are based on testing as of dates shown in configurations and may not reflect all publicly available updates. Visit MLCommons for more details. No product or component can be absolutely secure. Cost estimates are based on internal estimates from public and partner sourcing for a configuration using Intel Xeon w7-2475x, 4x Arc Pro B60 Dual GPU cards, and 2 memory sticks of 64GB DDR5 5600MHz memory as of September 2025. Your costs and results may vary. Intel technologies may require enabled hardware, software, or service activation. 

Frequently Asked Questions (FAQS): 

Q: What are the key benchmarks in MLPerf Inference v5.1?

A: The key benchmarks in MLPerf Inference v5.1 include performance metrics for various AI models, such as Llama 8B, across different hardware configurations.

Q: How does the Intel Arc Pro B60 compare to NVIDIA's RTX Pro 6000 and L40S in terms of cost and performance?

A: The Intel Arc Pro B60 offers performance per dollar advantages of up to 1.25x and up to 4x compared to NVIDIA RTX Pro 6000 and L40S, respectively, in Llama 8B models.

Q: What is Project Battlematrix?

A: Project Battlematrix is the code name for Intel's new GPU Systems designed to meet the needs of modern AI inference, providing an all-in-one platform that combines full-stack validated hardware and software.

Q: What role do CPUs play in AI systems?

A: CPUs act as the orchestration hub, handling preprocessing, transmission, and overall system coordination in AI systems.

Q: What improvements did Intel Xeon 6 with P-cores achieve in MLPerf Inference v5.1?

A: Intel Xeon 6 with P-cores achieved a 1.9x performance improvement generation-over-generation in MLPerf Inference v5.1. 

More Related Topics :