Intel has announced the release of its new Xeon 6 with Performance-cores (P-cores) and Gaudi 3 AI accelerators, offering double the performance for AI and HPC workloads. These innovations deliver significant improvements in performance per watt, with optimized total cost of ownership (TCO), enabling businesses to scale AI infrastructure efficiently.

Intel Expands AI Capabilities with Xeon 6 and Gaudi 3 AI Accelerators

In response to the growing demand for scalable and efficient AI infrastructure, Intel has introduced two powerful new products: Xeon 6 with Performance-cores (P-cores) and Gaudi 3 AI accelerators. These advancements underscore Intel's focus on delivering high-performance AI systems with reduced TCO, enabling enterprises to meet AI and high-performance computing (HPC) workloads with enhanced efficiency.

Intel's Xeon 6 processor is engineered for compute-intensive tasks, offering twice the performance of its predecessor and integrating AI acceleration capabilities in every core. Alongside this, the Gaudi 3 AI accelerator is optimized for large-scale generative AI, providing advanced networking capabilities and seamless integration with AI frameworks like PyTorch and Hugging Face.

Key Features of Intel Xeon 6 and Gaudi 3 AI Accelerators

• Intel® Xeon® 6 with P-cores: Xeon 6 delivers substantial improvements in AI processing, including an increased core count and double the memory bandwidth, making it suitable for workloads ranging from edge devices to cloud environments. Its embedded AI acceleration in each core ensures high efficiency, doubling performance over its predecessor.
• Intel® Gaudi® 3 AI Accelerator: Gaudi 3 is designed to handle the intensive demands of generative AI, featuring 64 Tensor processor cores (TPCs) and eight matrix multiplication engines (MMEs). With 128 GB of HBM2e memory and advanced networking features, Gaudi 3 accelerates deep learning processes, offering up to 20% more throughput and twice the price-performance compared to competing solutions.

Optimized AI Systems and Cost Efficiency

Intel’s latest innovations offer enterprises an optimized AI infrastructure with significant cost savings and performance benefits. The company has partnered with major OEMs such as Dell Technologies and Supermicro to co-engineer systems specifically tailored for AI deployments. Notably, Intel’s robust x86 architecture, used in 73% of GPU-accelerated servers, ensures flexibility and compatibility across AI workloads.

By enhancing AI infrastructure with TCO advantages and boosting performance per watt, Intel is helping businesses efficiently scale their AI capabilities from prototype to production environments.

Accelerating Enterprise AI Adoption with Co-Engineering and New Solutions

Intel's collaboration with partners enables seamless integration of generative AI solutions into production-ready systems. Through co-engineering efforts, Intel is addressing the challenges of real-time monitoring, error handling, and security, ensuring smoother transitions for enterprises deploying AI at scale.

The introduction of the Open Platform Enterprise AI (OPEA) platform integrates microservices optimized for Xeon 6 and Gaudi 3 systems. This platform allows for efficient deployment and scalability of retrieval-augmented generation (RAG) solutions, ensuring businesses can rapidly adopt cutting-edge AI applications.

Expanding Enterprise Access with Intel Tiber Portfolio and Developer Cloud

Intel's commitment to expanding access to AI technology is evident in its Tiber portfolio, which addresses the challenges enterprises face in deploying AI across cloud, edge, and data center environments. The Intel Tiber Developer Cloud provides early access to Xeon 6 and Gaudi 3 for testing and tech evaluation, with production-ready Gaudi 3 clusters rolling out next quarter.

Moreover, new service offerings such as SeekrFlow, an AI platform from Seekr, offer businesses an end-to-end solution for developing trusted AI applications. The platform is powered by Intel’s AI tools, including the latest Gaudi 3 software, enabling developers to create high-performance AI models with ease.

Intel’s release of Xeon 6 with P-cores and Gaudi 3 AI accelerators marks a significant step forward in the evolution of AI infrastructure. These new products offer unparalleled performance for AI and HPC workloads while optimizing cost efficiency, making them essential tools for enterprises looking to scale AI capabilities. Through its partnerships, co-engineering efforts, and expanded access to AI technologies, Intel continues to lead the way in transforming AI systems for the future. Read full articl from Intel's blog post here.