SANTA CLARA, Calif.- May 17, 2024 – Intel has released its new Xeon 6 P-Core CPUs to maximize GPU performance in AI systems. The proclamation, made in Santa Clara, California, unveils the new processing units designed to optimize advanced artificial intelligence. Highlighting enhancements in memory, core count, and bandwidth, Intel’s new processors aim to significantly advance processing capabilities overall. Experts anticipate the release to drive increased efficiency and performance across AI workloads.
Intel Launches Xeon 6 CPUs to Maximize GPU Performance in AI Systems
Santa Clara, Calif. – Intel has introduced its new line of Xeon 6 P-Core central processing units (CPUs), engineered to optimize the performance of advanced, GPU-driven artificial intelligence (AI) systems. These processors are immediatly available and are integrated into nvidia’s DGX B300 AI systems.
Key Features of the xeon 6 P-Core CPUs
- Performance-cores: Equipped with Intel’s Performance-cores for enhanced processing capabilities.
- Intel priority Core turbo (PCT): Allows dynamic prioritization of high-priority cores,enabling higher turbo frequencies.
- Intel Speed Select Technology – Turbo Frequency (SST-TF): Delivers customizable CPU core frequencies to improve GPU performance for demanding AI workloads.
PCT and SST-TF: A Leap Forward in AI Performance
Intel emphasizes that the combination of PCT and Intel SST-TF represents a significant leap forward in AI system performance.
PCT enables the dynamic prioritization of high-priority cores,allowing them to achieve higher turbo frequencies,while lower-priority cores operate at base frequency to optimize resource distribution. According to Intel, PCT can run up to eight high-priority cores at elevated turbo frequencies.
Enhanced Specifications and Capabilities
- Core Count: Up to 128 P-cores per CPU.
- pcie Lanes: 20% more PCIe lanes than previous-generation Xeon processors,with up to 192 PCIe lanes per 2S server.
- Memory Speed: 30% faster memory speeds compared to competing AMD EPYC processors, thanks to Multiplexed Rank DIMMs (MRDIMMs) and Compute Express Link.
- Memory Bandwidth: Up to 2.3x higher memory bandwidth compared to the previous generation.
Memory and Cache Enhancements
Intel’s Xeon 6 CPUs feature 2 DIMMs per channel (2DPC), supporting up to 8TB of system memory. The CPUs also include up to 504 MB L3 cache for faster data retrieval.
Intel AMX and FP16 Precision
The Xeon 6 processors incorporate Intel AMX, which offloads specific tasks to the CPU. Intel has confirmed that AMX now supports FP16 precision arithmetic, enabling efficient data pre-processing and critical CPU tasks in AI workloads. Along with the three new P-Core processors, Intel has also introduced a B-variant, the 6716P-B, which features 40 cores and consumes less power.
Processor Specifications
| Product Name | Total Cores | Max turbo Frequency | Processor Base Frequency | Cache | TDP |
|---|---|---|---|---|---|
| Intel Xeon 6732P Processor | 32 | 4.3 GHz | 3.8 GHz | 144 MB | 350 W |
| Intel Xeon 6774P Processor | 64 | 3.90 GHz | 2.50 GHz | 336 MB | 350 W |
| Intel Xeon 6776P Processor | 64 | 3.90 GHz | 2.30 GHz | 336 MB | 350 W |
| Intel Xeon 6716P-B Processor | 40 | 3.5 GHz | 2.5 GHz | 160 MB | 235 W |
Intel asserts that its Xeon 6 processors with P-cores provide the ideal combination of performance and energy efficiency
to meet the growing demands of AI computing.
Additional Announcements at Computex 2025
At Computex 2025,Intel also unveiled its new $299 Arc Pro B50 with 16GB of memory,along with ‘Project Battlematrix’ workstations featuring 24GB Arc Pro 60 GPUs.
FAQ Section
- What are the main benefits of the Intel Xeon 6 P-Core CPUs?
- Enhanced performance for GPU-driven AI systems, customizable core frequencies, and improved memory bandwidth.
- What is Intel Priority Core Turbo (PCT)?
- A technology that allows dynamic prioritization of high-priority cores, enabling higher turbo frequencies.
- Which Nvidia systems are these processors integrated into?
- Nvidia DGX B300 AI systems.
