At the end of 2024, Intel introduced the latest Xeon processor as part of the Xeon 6 processor range, offering an array of models, tailored to meet diverse workload requirements. With AI continuing to play a critical role across industries, organisations are increasingly in need of advanced, high-performance CPUs to keep pace with rapid technological evolution. Addressing these demands, Intel has positioned the Xeon 6 series as a comprehensive solution, core to this is the flagship Performance-cores (P-cores) variant, specifically designed to tackle the most demanding AI workloads with exceptional efficiency and power.
At Boston, we understand the evolving needs of our customers, especially in performance-critical environments. In response, we recently conducted in-depth benchmark testing on the Intel Xeon 6 6980P in our lab. These tests provide valuable insights and detailed performance figures, offering a clearer picture of how this processor performs under demanding workloads.
Intel Xeon 6 processors with Performance-cores (P-cores)
The Intel Xeon 6900P (Granite Rapids-AP) supports 1 or 2 socket configurations and especially fit for the demanding high-performance computing (HPC) and AI workloads. The Xeon 6 6900P series features an all-P-core design with up to 128 cores per socket. It supports 12-channels of DDR5-6400 or 8800MT/s MRDIMM/ MCR DIMM memory. There are 96 lanes of PCIe Gen5 per CPU for 192 lanes total, and there is CXL 2.0 support, all while enabling a full 6 UPI lanes for socket-to-socket bandwidth. Intel Xeon 6980P has 504MB of L3 cache.
Features highlight for Xeon 6 6900P series CPU
Intel provides a wide range of Xeon 6 processor configurations with different core counts and capabilities thanks to its highly modular design approach. This flexibility is enabled by advanced packaging technologies such as EMIB, a high-speed interconnect for linking multiple dies; a modular die fabric that allows seamless communication between components; and a tile-based, multi-die architecture that lets Intel combine various functional blocks as needed.
Xeon 6 processors are designed with separate compute and I/O dies, connected by a high-speed fabric that links all dies within the package. The compute dies, built on the Intel 3 process, contain all the CPU cores—either Performance (P) or Efficiency (E) cores—while the I/O dies, manufactured on the Intel 7 process, handle key connectivity features such as UPI, PCIe and CXL, as well as integrated Intel Accelerator engines.
The CPU used for the Benchmark Test: Intel Xeon 6980P
The CPU we used in the benchmark test is the flagship model - Xeon 6980P. The Intel Xeon 6980P features 128 cores / 256 threads, and a 2.0GHz base clock with 3.2GHz all-core turbo frequency and 3.9GHz maximum turbo frequency. The Xeon 6980P features 504MB of L3 cache and a 500 Watt TDP. As with the rest of the Xeon 6 P-core SKUs, there is support for 12 channel DDR5-6400 memory or MRDIMM 8800MT/s.
Xeon 6 6900P series SKU summary
Our Test Server
The test server was as follows:
- Supermicro Hyper Server SYS-222HA-TN (dual socket)
- 2x Intel 6980P CPUs (128 Cores / 256 Threads and 500W TDP)
- 24x 32GB DDR5-6400 ECC RDIMMs memory (Total: 768GB)
- 1.92TB U.2 NVMe disk
- OS: Rocky Linux 9.4
In the benchmark tests below, we used 5th Generation Xeon CPUs (Emerald Rapids) for comparison with the Xeon 6 6980P CPU. The 5th Generation Xeon CPUs included in this evaluation are the 8592+ (64 cores and 128 threads, 350W TDP), the 6530 (32 cores and 64 threads, 270W TDP), and the 6544Y (16 cores and 32 threads, 270W TDP). As shown, the maximum number of cores supported by the 5th Generation CPUs is 64, which is half the core count of the Xeon 6 6980P with 128 cores. Based on this, it is expected that the Xeon 6 6980P would outperform the 5th Generation CPUs. With that in mind, let’s take a closer look at the benchmark results below.
Chart: 7-Zip in Compression results
Chart: 7-Zip in Decompression results
In the 7-Zip test, which benchmarks data compression and decompression performance in MIPS (Million Instructions Per Second), the Xeon 6 6980P scored 853627 points, while the Xeon 8592+ achieved 596244. This reflects over a 30% performance improvement for the Xeon 6 6980P in these workloads, largely due to its significantly higher core count. The results also show nearly a 3x performance gain for the Xeon 6 6530 and 6544Y CPUs, further demonstrating the substantial efficiency and throughput improvements offered by the Xeon 6 architecture.
Chart: Blender (BMW27) test results
The Xeon 6 6980P, with its 128 cores and 256 threads per socket, delivers leading 3D rendering performance in Blender tests. By supporting branched path tracing, the Xeon 6 6980P achieves outstanding results in both Blender benchmark tests. It completes the BMW 27 rendering in just 7.63 seconds and the complex interior rendering in the Barbershop test in 80.02 seconds. These results demonstrate an impressive performance gain of over 40% in both the BMW 27 and Barbershop tests compared to 8592+ CPU.
Chart: PassMark CPU test results
The PassMark CPU benchmark performs a wide variety of calculations. The benchmark score is based on how quickly the CPU can execute a complex series of instructions. The faster the processor completes the tasks, the higher the benchmark score. The tests cover several tasks, including mathematical operations, compression, encryption, physics and both multithreaded and single-threaded workloads. In these tests, the Xeon 6 6980P scored 172758, while the Xeon 8592+ scored 143181, showing a notable performance gain of around 20% compared to the 8592+. The 6980P continues to demonstrate impressive results when compared with the 5th generation SKUs, clearly proving to be superior for handling a wide range of computational tasks and workloads.
Summary
The Xeon 6980P does manage to land well in front of 5th gen Intel processors. Generationally around 1.2x the performance with Granite Rapids compared to the prior Emerald Rapids flagships granted there is twice the number of cores.
Boston Labs
These benchmark tests were conducted at our Boston Labs facility. Boston Labs is all about enabling our customers to make informed decisions in selecting the right hardware, software and overall solution for their specific challenges.
If you’d like to request a test drive of Xeon 6 Processors, please get in contact by emailing sales@boston.co.uk or call us on 01727 876100 and one of our experienced sales engineers will gladly guide you through building the perfect solution just for you.
Author
Alvin Leung
Field Application Engineer
Boston Limited