cancel
Showing results for 
Search instead for 
Did you mean: 

5th Gen AMD EPYC™ Processors Lead Enterprise and Cloud Workloads Forward

raghu_nambiar
5 0 11K

When AMD introduced the 1st generation of the AMD EPYC™ processor, we redefined what was possible. Previous, monolithic approaches to processor design were shuffled off the stage in favor of innovative approaches to meet the ever-increasing demands of modern datacenters.

Each successive generation of AMD EPYC processors has delivered both incremental and step-function levels of improvement in performance, scalability and energy efficiency. With the launch of the highly anticipated 5th generation based on AMD “Zen5” and “Zen5c” processor cores, I can confidently state that the AMD legacy of innovation and leadership is secure.

Servers equipped with the newly unveiled 5th Gen AMD EPYC processors deliver remarkable generational improvements, including up to twice the workload performance in many scenarios. Notable features include:

  • New “Zen 5” Cores: Providing up to 17% higher instructions per clock (IPC) for single-threaded tasks, resulting in substantial performance gains.[1]
  • Increased Core Density: With a core count rising to 192—an increase of 50% from previous generations—5th Gen AMD EPYC processors elevate performance and scalability to unprecedented levels.
  • Enhanced Memory: Achieving speeds of up to 6400 MT/s and supporting 8 TB of capacity per processor to efficiently manage data-intensive workloads.
  • Enhanced Confidential Computing: Allowing confidential guest VMs to securely interact with PCIe TDISP devices.

These advancements underscore the ongoing AMD commitment to developing leadership server processor technologies that deliver new levels of performance, efficiency, and advanced capabilities to the data center.

The Ecosystem Is More Prepared Than Ever

Our partners drive our success! Software is the key to any business solution, and it takes a seamless alignment between hardware and software optimization to deliver strong performance and efficiency.  Today’s businesses rely on tried-and-true server and business application vendors to enable technology-driven innovation.  The AMD portfolio of trusted partners integrates AMD EPYC processors into their advanced solutions. AMD thrives on our close collaboration with a diverse and expanding network of ecosystem partners across both open-source and commercial software stacks. This partnership fuels the introduction of next-generation engineering innovations that enhance AMD EPYC 9005 Series Processors and provide immediate value to enterprises, cloud environments, high-performance computing (HPC), and AI applications.

We extend our deepest appreciation to our extensive ecosystem of partners who work alongside our engineers to deliver a broad array of datacenter solutions, including (but by no means limited to):

Alibaba, Altair, Anjuna, Ansys, AspenTech, Astera, Autodesk, Azure, Amazon, Broadcom, ByteDance, Cadence, Canonical, Casa Systems, CGG, Chaos Group, Cisco, Citrix, Cloudera, Cloudian, Cohesity, Convergent Science, Cornelis, Couchbase, Dassault Systèmes, Databricks, DataStax, Dell, Elastic, Emerson, Epic, Ericsson, ESI, Exasol, F5 Networks, Flow Science, Google, Halliburton, Hexagon, HPE, HuggingFace, Juniper Networks, Keysight, KIOXIA, Lenovo, MarkLogic, Marvell, Mavenir, MaxLinear, Microchip, Micron, Microsoft, MongoDB, Montage, NetScout, Neural Magic, Nokia, Nutanix, Nvidia, Oracle, Pivot3, Pixar, Quobyte, Radisys, Red Hat, RedisLabs, Rescale,  Samsung, SAP, Sentieon, Shearwater, Siemens, SK Hynix, SLB, SnowFlake, Solidigm, Splunk, StorMagic, Supermicro, SUSE, Synamedia, Synopsys, Tencent,  TigerGraph, Transwarp, VAST Data, Vertica, Viavi Solutions, VMware by Broadcom, Weka, Wind River, Western Digital,  XConn and others. We also collaborate with Apache Cassandra, Apache Spark, Apache Hadoop, Confidential Computing Consortium, CXL Consortium, Cloud Native Computing Foundation, FreeBSD Foundation, Linux Foundation, OpenAnolis Community, OpenCloudOS community, OpenEuler Community, and others.

Please visit the AMD Data Center Partner Ecosystem page to see the full list of AMD EPYC ecosystem partners.

Benchmarks to Best Practices

AMD is committed to industry-standard benchmarks that are fully transparent, peer-reviewed or independently audited, and verifiable. We provide full disclosure reports to enable others to reproduce our results. We take pride in holding over 500 world records across a diverse range of important, customer-relevant workloads. Additionally, we have thoroughly tested a wide array of commonly used workloads that evaluate the real-world performance of customer-relevant tasks.

Understanding that 'time to value' is vital for our customers, we provide seamless, out-of-the-box experiences when deploying applications on AMD EPYC processors. Our focus on optimizing both industry-standard benchmarks and commonly used workloads has led us to create a comprehensive library of tuning guides. These resources, available in the AMD Documentation Hub, encapsulate the insights gained from our optimization efforts, enabling customers to maximize performance and efficiency with ease.

Performance Proof Points

Let’s take a closer look at how 5th Gen AMD EPYC processors excel across critical workloads essential for enterprise and cloud deployments. Please remember that our SKU selection is based on best practice recommendations; your specific needs may vary. We encourage you to reach out to your AMD representatives for personalized advice and support on selecting the most suitable CPUs.

In this blog, I am comparing 5th Gen and 4th Gen AMD EPYC processors to highlight our commitment to continuously improving and driving innovation from generation to generation. The 5th Gen AMD EPYC lineup presents a wide array of options, featuring processors with 8 to 192 cores and TDP ratings ranging from 155 to 500 watts. I am focusing primarily on the 5th Gen AMD EPYC 9755, which offers 128 cores at 500W TDP, the 4th Gen AMD EPYC 9654, a 96-core processor from our previous generation, and the Intel® Xeon® Platinum 8592+, the flagship of the 5th Gen Intel Xeon lineup. I am also highlighting the 64-core 5th Gen AMD EPYC 9575F processor that stands out as an industry leader by surpassing the 5 GHz barrier, making it an excellent choice for enterprise virtual environments and applications with per-core software licensing by offering an excellent total cost of ownership (TCO).

General Purpose Computing

The SPEC CPU® 2017 benchmark is a highly regarded, regulated, and audited industry standard for assessing compute-intensive workloads. It rigorously tests processors, memory subsystems, and compilers across a variety of systems, providing valuable insights into performance and efficiency. This benchmark serves as a critical tool for evaluating the capabilities of modern computing architectures. It includes 43 benchmarks divided into four suites. This blog focuses on SPECrate® 2017 Integer and SPECrate® 2017 Floating Point—perhaps two of the most foundational points of comparison that customers request.

Figure 1 shows the 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 SPECrate® 2017_int_base uplifts of ~2.41x and ~1.60x versus the 5th Gen Intel Xeon 8592+, respectively. It also shows the 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 SPECrate® 2017_fp_base uplifts of ~1.84x and ~1.17x versus the same Intel system, respectively.[2][3]

raghu_nambiar_0-1728515260894.png

 Figure 1: General purpose computing

Server Side Java

Java® is a versatile and powerful programming language known for its portability, allowing developers to create applications that run seamlessly across different platforms. The SPECjbb® 2015 benchmark assesses the performance of server-side Java applications by simulating a corporate IT environment that handles a mix of point-of-sale requests, online transactions, and data-mining tasks. Its relevance extends to Java Virtual Machine (JVM) vendors, hardware manufacturers, Java application developers, researchers, commercial and end users, and academia.

The SPECjbb® 2015 benchmark includes two key metrics: max-JOPS and critical-JOPS. Max-JOPS measures the highest throughput for Java e-commerce transactions, reflecting the system's maximum capacity. In contrast, critical-JOPS indicates the throughput that meets the p99 response time criteria, representing the typical operational range for latency-sensitive Java e-commerce applications.

Figure 2 shows both:

  • 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 SPECjbb® 2015-MultiJVM max-jOPS uplifts of ~2.10x and ~1.48x versus the 5th Gen Intel Xeon 8592+, respectively.[4]
  • 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 SPECjbb® ® 2015-MultiJVM critical-jOPS uplifts of ~2.02x and ~1.47x versus the 5th Gen Intel Xeon 8592+, respectively.[5]

raghu_nambiar_1-1728515260895.png

 Figure 2:  SPECjbb® 2015 performance (high core-count processors)

We discussed the performance of the top-tier SKUs. Now, let's examine the competitive performance at an identical core count. This will help you understand how much additional performance can be achieved while maintaining the same number of cores, a crucial consideration for many enterprises who have software licenses that are based on core count.

Figure 3 shows both:

  • 5th Gen AMD EPYC 9555 and 4th Gen AMD EPYC 9554 SPECjbb® 2015-MultiJVM max-jOPS uplifts of ~1.29x and ~1.06x versus the 5th Gen Intel Xeon 8592+, respectively.[6]
  • 5th Gen AMD EPYC 9555 and 4th Gen AMD EPYC 9554 SPECjbb® ® 2015-MultiJVM critical-jOPS uplifts of ~1.30x and ~1.04x versus the 5th Gen Intel Xeon 8592+, respectively.[7]

raghu_nambiar_2-1728515260896.png

 Figure 3: SPECjbb® 2015 performance (64-core processors)

Business Applications

SAP Sales and Distribution (SAP SD) is a crucial logistics module within SAP Enterprise Resource Planning (ERP) software. The SAP-SD 2-Tier benchmark evaluates hardware performance by measuring database efficiency in SAP Application Performance Standard units (SAPS). SAPS is a hardware-independent metric that assesses system performance within the SAP environment based on the Sales and Distribution (SD) benchmark.

Figure 4 shows the 5th Gen AMD EPYC 9965 and 4th Gen AMD EPYC 9654 SAP SD uplifts of ~2.08x and ~1.24x versus the 5th Gen Intel Xeon 8592+, respectively.[8]

raghu_nambiar_3-1728515260897.png

 Figure 4: SAP SD 2-Tier Benchmark Users

Relational Database Management Systems (RDBMS)

MySQL™ is one of the most widely used open-source database management systems globally. It is commonly employed in enterprises and cloud-native environments for both decision support and transaction processing systems. AMD used TPROC-H to evaluate the performance of decision support systems and TPROC-C for transaction processing systems. TPROC-H is based on the well-known TPC-H benchmark standard, while TPROC-C is derived from the TPC-C benchmark.

Figure 5 shows both:

  • 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 MySQL TPROC-C uplifts of ~3.45x and ~1.91x versus the 5th Gen Intel Xeon 8592+, respectively.[9]
  • 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 MySQL TPROC-H uplifts of ~2.79x and ~1.44x versus the 5th Gen Intel Xeon 8592+, respectively.[10]

raghu_nambiar_4-1728515260898.png

 Figure 5: MySQL TPROC-C (transactions per minute) and TPROC-H (queries per hour)

Web Server 

NGINX™ is a popular web server known for its flexibility to act as a reverse proxy, load balancer, mail proxy, and HTTP cache. It's designed to efficiently handle client requests and deliver web content. NGINX can operate as a standalone web server or boost performance and security by serving as a reverse proxy for other servers. AMD used the popular WRK tool to evaluate performance by generating significant HTTP loads during benchmark testing.

Figure 6 shows the 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 NGINX uplifts of ~4.17x and ~1.80x versus the 5th Gen Intel Xeon 8592+, respectively.[11]

raghu_nambiar_5-1728515260899.png

 Figure 6: NGINX WRK Benchmark

In-Memory Analytics

Redis™ is a popular in-memory data store that works as a key-value database, cache, and message broker, with optional durability features. It's great for cloud environments and supports functions like streaming, microservices, and data analytics, helping users to be more productive. Redis can handle various data types, including strings, lists, maps, and more. Its in-memory design provides excellent performance.

AMD used a tool called redis-benchmark to measure how well Redis performs. This tool simulates multiple clients connecting to the server and measures how long it takes to complete requests. The results show the average number of requests your Redis server can handle per second.

Figure 7 shows the 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 Redis uplifts of ~3.20x and ~2.57x versus the 5th Gen Intel Xeon 8592+, respectively.[12]

raghu_nambiar_6-1728515260900.png

 Figure 7: Redis-Bench

Media Processing

FFmpeg is a powerful, free, open-source software project that includes a comprehensive suite of libraries, codecs, and tools for handling video, audio, and other multimedia files and streams. The core FFmpeg program excels at command-line processing of multimedia content, making it a popular choice for tasks such as encoding, transcoding, editing, video scaling, post-production, and ensuring standards compliance. AMD testing focused on transcoding a raw 4K resolution input file using the VP9 codec, resulting in an MKV output file. This demonstrates FFmpeg's capabilities in managing high-quality video content efficiently.

Figure 8 shows the 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 FFmpeg uplifts of ~3.99x and ~1.90x versus the 5th Gen Intel Xeon 8592+, respectively.[13]

raghu_nambiar_7-1728515260900.png

 Figure 8: FFmpeg VP9 transcoding to MKV

Big Data Analytics

Big Data Analytics is crucial in today’s enterprise environments. The Apache® Cassandra® database stands out as an excellent option for achieving scalability and continuous availability without compromising performance. It simplifies data distribution in multi-tenant setups and is particularly well-suited for cloud applications. With its predictable scalability and proven fault tolerance, Cassandra is perfect for managing mission-critical data. AMD utilized the popular Cassandra-Stress benchmark to evaluate the performance of our 5th Gen EPYC processors.

Figure 9 shows the 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 Cassandra uplifts of ~3.75x and ~2.25x versus the 5th Gen Intel Xeon 8592+, respectively, tested using Cassandra-stress workload with a 70/30 RW ratio.[14]

raghu_nambiar_8-1728515260901.png

 Figure 9: Cassandra-Stress

High Performance Data Caching

Memcached™ is a high-performance, distributed, in-memory caching system that stores key-value pairs for small pieces of arbitrary data, such as strings and objects, sourced from database calls, APIs, or rendered pages. It is widely used in cloud environments for its ability to quickly serve cached items while enabling easy scaling and cost-effectiveness under heavy loads. Memcached is often employed to cache database query results, session data, web pages, APIs, and objects like images, files, and metadata. Built for simplicity and efficiency, it supports rapid development and deployment to tackle the challenges of large data caches. To assess its performance, AMD utilized the Memtier Benchmark, which measures throughput and latency across different configurations.

Figure 10 shows the 5th Gen AMD EPYC 9755 and 4th Gen AMD EPYC 9654 Memcached uplifts of ~3.55x and ~1.75x versus the 5th Gen Intel Xeon 8592+, respectively.[15]

raghu_nambiar_9-1728515260902.png

 Figure 10: Memtier Benchmark

Virtualization and Consolidation

VMmark 4.0 is a benchmarking tool created by VMware to assess the performance of virtualization platforms. It evaluates how effectively a system can run multiple virtual machines (VMs) simultaneously under workloads that mimic real-world applications. This benchmark specifically tests server hardware and hypervisors through a variety of tasks, including database transactions, web server requests, and file services. The results offer valuable insights into a virtualization environment's ability to manage multiple VMs, aiding organizations in making informed infrastructure decisions.

Figure 11 shows the 5th Gen AMD EPYC 9965 and 9975 processors delivering generational VMmark 4 uplifts of ~1.45x and ~1.25x relative to the 4th Gen AMD EPYC 9654 processor, respectively.[16]

raghu_nambiar_10-1728515260903.png

 Figure 11: VMmark 4 score (high core count CPUs)

Now let’s compare the performance of 64-core 5th Gen AMD EPYC processor-based servers against the 4th Gen AMD EPYC and the 5th Gen Intel Xeon. All results are normalized to the performance of the Intel Xeon Platinum 8592+, offering a clear view of how these advanced processors compare. This comparison is crucial, as many software stacks are licensed per core, meaning optimizing core performance can significantly reduce licensing costs. The 5th Gen AMD EPYC 9575F processor stands out as an industry leader by surpassing the 5 GHz barrier, making it an excellent choice for enterprise virtual environments.

Figure 12 shows the 5th Gen AMD EPYC 9575F and 4th Gen AMD EPYC 9554 uplifts of ~1.61x and ~1.28x versus the 5th Gen Intel Xeon 8592+, respectively for VMmark 4.0.[17]

fig12.png

 Figure 12:  VMmark 4 score (64-core CPUs)

Power Efficiency

Efficient power consumption is crucial for today's data centers, as it helps lower cooling needs, cut utility costs, and support sustainability objectives. The SPEC CPU®2017 Integer Rate Energy metric builds on the well-known SPEC CPU®2017 Integer Rate benchmark by adding energy efficiency measures. This metric assesses both performance and power usage, positioning itself as the industry standard for evaluating efficiency.

Figure 13 shows the 5th Gen AMD EPYC 9965 and 4th Gen AMD EPYC 9654 SPECrate® 2017_int_base_energy uplifts of ~2.19x and ~1.62x versus the 5th Gen Intel Xeon 8592+, respectively.[18]

raghu_nambiar_12-1728515260904.png

 Figure 13: SPEC CPU®2017 Integer Rate Energy

Wrapping Up

The 5th Generation of AMD EPYC processors represent a significant leap forward in server processors, extending recognized AMD x86 leadership in high core counts, advanced multi-threading, strong performance, and impressive energy efficiency. Designed to meet fast growing demands for performance, capacity and efficiency for modern data centers, the AMD EPYC architecture delivers exceptional value across a variety of applications, including critical enterprise and cloud-native workloads. 5th Gen AMD EPYC processors boast a cutting-edge design with an increased core count, improved instructions per cycle (IPC), and enhanced memory and I/O bandwidth, all while incorporating innovative confidential computing technologies. This groundbreaking processor is setting a new standard in server technology, thanks to the invaluable support of our ecosystem partners throughout this remarkable journey. In my next blog, I'll explore the innovations and differentiators that the 5th Gen EPYC processors bring to HPC and AI. Stay tuned!

Raghu Nambiar is a Corporate Vice President of Data Center Ecosystems and Solutions for AMD. His postings are his own opinions and may not represent AMD’s positions, strategies or opinions. Links to third party sites are provided for convenience and unless explicitly stated, AMD is not responsible for the contents of such linked sites and no endorsement is implied.

References

  1. 9xx5-001: Based on AMD internal testing as of 9/10/2024, geomean performance improvement (IPC) at fixed-frequency. 5th Gen EPYC CPU Enterprise and Cloud Server Workloads generational IPC Uplift of 1.170x  (geomean) using a select set of 36 workloads and is the geomean of estimated scores for total and all subsets of SPECrate®2017_int_base (geomean ), estimated scores for total and all subsets of SPECrate®2017_fp_base (geomean), scores for Server Side Java multi instance max ops/sec,   representative Cloud Server workloads (geomean), and representative Enterprise server workloads (geomean). “Genoa” Config (all NPS1): EPYC 9654 BIOS TQZ1005D 12c12t (1c1t/CCD in 12+1), FF 3GHz, 12x DDR5-4800 (2Rx4 64GB), 32Gbps xGMI; “Turin” config (all NPS1): EPYC 9V45    BIOS RVOT1000F 12c12t (1c1t/CCD in 12+1), FF 3GHz, 12x DDR5-6000 (2Rx4 64GB), 32Gbps xGMI. Utilizing Performance Determinism and the Performance governor on Ubuntu® 22.04 w/ 6.8.0-40-generic kernel OS for all workloads. 5th Gen EPYC generational ML/HPC Server Workloads IPC Uplift of 1.369x (geomean) using a select set of 24 workloads and is the geomean of representative ML Server Workloads (geomean), and representative HPC Server Workloads (geomean). Genoa Config (all NPS1) “Genoa” config: EPYC 9654 BIOS TQZ1005D 12c12t (1c1t/CCD in 12+1), FF 3GHz, 12x DDR5-4800 (2Rx4 64GB), 32Gbps xGMI; Turin” config (all NPS1):   EPYC 9V45 BIOS RVOT1000F 12c12t (1c1t/CCD in 12+1), FF 3GHz, 12x DDR5-6000 (2Rx4 64GB), 32Gbps xGMI. Utilizing Performance Determinism and the Performance governor on Ubuntu 22.04 w/ 6.8.0-40-generic kernel OS for all workloads except LAMMPS, HPCG, NAMD, OpenFOAM, Gromacs, which utilize 24.04 w/ 6.8.0-40-generic kernel. SPEC® and SPECrate® are registered trademarks for Standard Performance Evaluation Corporation. Learn more at www.spec.org.
  2. 9xx5-018B: SPECrate®2017_fp_base comparison based on published scores from www.spec.org as of 10/10/2024. 2P AMD EPYC 9965 (2350 SPECrate®2017_fp_base, 384 Total Cores, 500W TDP, $14,813 CPU $), 4.700 SPECrate®2017_fp_base/CPU W, 0.159 SPECrate®2017_fp_base/CPU $, https://www.spec.org/cpu2017/results/res2024q4/cpu2017-20240923-44822.html). 2P AMD EPYC 9755 (2320 SPECrate®2017_fp_base, 256 Total Cores, 500W TDP, $12,984 CPU $), 4.640 SPECrate®2017_fp_base/CPU W, 0.179 SPECrate®2017_fp_base/CPU $, https://www.spec.org/cpu2017/results/res2024q4/cpu2017-20240923-44823.html). 2P AMD EPYC 9654 (1480 SPECrate®2017_fp_base, 192 Total Cores, 360W TDP, $11,805 CPU $), 4.111 SPECrate®2017_fp_base/CPU W, 0.125 SPECrate®2017_fp_base/CPU $, https://www.spec.org/cpu2017/results/res2022q4/cpu2017-20221024-32605.html). 2P Intel Xeon Platinum 8592+ (1260 SPECrate®2017_fp_base, 128 Total Cores, 350W TDP, $11,600 CPU $) 3.600 SPECrate®2017_int_base/CPU W, 0.109 SPECrate®2017_int_base/CPU $, https://www.spec.org/cpu2017/results/res2024q3/cpu2017-20240701-43949.html). SPEC®, SPEC CPU®, and SPECrate® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information. Intel CPU TDP at https://ark.intel.com/.
  3. 9xx5-002C: SPECrate®2017_int_base comparison based on published scores from www.spec.org as of 10/10/2024. 2P AMD EPYC 9965 (3000 SPECrate®2017_int_base, 384 Total Cores, 500W TDP, $14,813 CPU $), 6.060 SPECrate®2017_int_base/CPU W, 0.205 SPECrate®2017_int_base/CPU $, https://www.spec.org/cpu2017/results/res2024q3/cpu2017-20240923-44833.html). 2P AMD EPYC 9755 (2720 SPECrate®2017_int_base, 256 Total Cores, 500W TDP, $12,984 CPU $), 5.440 SPECrate®2017_int_base/CPU W, 0.209 SPECrate®2017_int_base/CPU $, https://www.spec.org/cpu2017/results/res2024q4/cpu2017-20240923-44837.pdf). 2P AMD EPYC 9754 (1950 SPECrate®2017_int_base, 256 Total Cores, 360W TDP, $11,900 CPU $), 5.417 SPECrate®2017_int_base/CPU W, 0.164 SPECrate®2017_int_base/CPU $, https://www.spec.org/cpu2017/results/res2023q2/cpu2017-20230522-36617.html). 2P AMD EPYC 9654 (1810 SPECrate®2017_int_base, 192 Total Cores, 360W TDP, $11,805 CPU $), 5.028 SPECrate®2017_int_base/CPU W, 0.153 SPECrate®2017_int_base/CPU $, https://www.spec.org/cpu2017/results/res2024q1/cpu2017-20240129-40896.html). 2P Intel Xeon Platinum 8592+ (1130 SPECrate®2017_int_base, 128 Total Cores, 350W TDP, $11,600 CPU $) 3.229 SPECrate®2017_int_base/CPU W, 0.097 SPECrate®2017_int_base/CPU $, http://spec.org/cpu2017/results/res2023q4/cpu2017-20231127-40064.html). 2P Intel Xeon 6780E (1410 SPECrate®2017_int_base, 288 Total Cores, 330W TDP) 4.273 SPECrate®2017_int_base/CPU W, 0.124 SPECrate®2017_int_base/CPU $, https://spec.org/cpu2017/results/res2024q3/cpu2017-20240811-44406.html). SPEC®, SPEC CPU®, and SPECrate® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information. Intel CPU TDP at https://ark.intel.com/.
  4. 9xx5-004A: SPECjbb®2015-MultiJVM max-jOPS comparison based on published results from www.spec.org as of 10/10/2024. 2P EPYC 9965 (1263805 SPECjbb®2015-MultiJVM max-jOPS, 921700 SPECjbb®2015 MultiJVM critical-jOPS), 384 total cores, https://spec.org/jbb2015/results/res2024q4/jbb2015-20240918-01420.html). 2P EPYC 9755 (1171060 SPECjbb®2015-MultiJVM max-jOPS, 851468 SPECjbb®2015 MultiJVM critical-jOPS), 256 total cores, https://spec.org/jbb2015/results/res2024q4/jbb2015-20240904-01327.html). 2P EPYC 9654 (828952 SPECjbb®2015-MultiJVM max-jOPS, 373695 SPECjbb®2015 MultiJVM critical-jOPS), 192 total cores, https://spec.org/jbb2015/results/res2024q2/jbb2015-20240515-01274.html). Versus: 2P Intel Xeon Platinum 8592+ (558,626 SPECjbb®2015 MultiJVM max-jOPS, 292175 SPECjbb®2015 MultiJVM critical-jOPS, 128 total cores, https://spec.org/jbb2015/results/res2024q2/jbb2015-20240430-01260.html). SPEC® and SPECjbb® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information.
  5. 9xx5-060: SPECjbb®2015-MultiJVM critical-jOPS comparison based on published results from www.spec.org as of 10/10/2024. 2P EPYC 9965 (921700 SPECjbb®2015-MultiJVM critical-jOPS, 1263805 SPECjbb®2015-MultiJVM max-jOPS), 384 total cores, https://spec.org/jbb2015/results/res2024q4/jbb2015-20240918-01407.html. 2P EPYC 9755 (851468 SPECjbb®2015-MultiJVM critical-jOPS, 1171060 SPECjbb®2015-MultiJVM max-jOPS), 256 total cores, https://spec.org/jbb2015/results/res2024q4/jbb2015-20240918-01395.html. 2P EPYC 9654 (622315 SPECjbb®2015-MultiJVM critical-jOPS, 664375 SPECjbb®2015-MultiJVM max-jOPS), 192 total cores, https://spec.org/jbb2015/results/res2022q4/jbb2015-20221019-00860.html. Versus: 2P Intel Xeon Platinum 8592+ (421207 SPECjbb®2015 MultiJVM critical-jOPS, 480007 SPECjbb2015-MultiJVM max-jOPS), 128 total cores, https://spec.org/jbb2015/results/res2024q1/jbb2015-20231224-01210.html. SPEC® and SPECjbb® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information.
  6. 9xx5-090: SPECjbb®2015-MultiJVM max-jOPS comparison based on compliant results and published results from www.spec.org as of 10/10/2024. 2P EPYC 9555 (718135 SPECjbb®2015-MultiJVM max-jOPS, 357222 SPECjbb®2015 MultiJVM critical-jOPS), 128 total cores, https://spec.org/jbb2015/results/res2024q4/jbb2015-20240916-01341.html). 2P EPYC 9554 (592109 SPECjbb®2015-MultiJVM max-jOPS, 255136 SPECjbb®2015 MultiJVM critical-jOPS), 128 total cores, 153GB 24x64GB 2Rx4 PC5-38400R, SUSE Linux Enterprise Server 15 SP6, 1x480GB NVMe), compliant run. Versus: 2P Intel Xeon Platinum 8592+ (558,626 SPECjbb®2015 MultiJVM max-jOPS, 292175 SPECjbb®2015 MultiJVM critical-jOPS, 128 total cores, https://spec.org/jbb2015/results/res2024q2/jbb2015-20240430-01260.html). SPEC® and SPECjbb® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information.
  7. 9xx5-091: SPECjbb®2015-MultiJVM critical-jOPS comparison based on compliant run and published results from www.spec.org as of 10/10/2024. 2P EPYC 9555 (546462 SPECjbb®2015-MultiJVM critical-jOPS, 565438 SPECjbb®2015-MultiJVM max-jOPS), 128 total cores, https://spec.org/jbb2015/results/res2024q4/jbb2015-20240916-01340.html. 2P EPYC 9554 (439679 SPECjbb®2015-MultiJVM critical-jOPS, 463251 SPECjbb®2015-MultiJVM max-jOPS, 128 total cores, 1536GB 24x64GB 2Rx4 PC5-38400R, SUSE Linux Enterprise Server 15 SP6, 1x480GB NVMe), compliant run). Versus: 2P Intel Xeon Platinum 8592+ (421207 SPECjbb®2015 MultiJVM critical-jOPS, 480007 SPECjbb2015-MultiJVM max-jOPS), 128 total cores, https://spec.org/jbb2015/results/res2024q1/jbb2015-20231224-01210.html. SPEC® and SPECjbb® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information.
  8. 9xx5-076: SAP Sales and Distribution benchmark comparison based on published results as of 10/10/2024. 2P AMD EPYC 9965 powered server (384 total cores, 768 threads), 2.35TB Memory, 201,000 benchmark users, 1,129,470 SAPS, https://www.sap.com/dmc/benchmark/2024/Cert24071.pdf. 2P AMD EPYC 9654 powered server (192 total cores, 384 threads), 1.5TB Memory, 148,000 benchmark users, 809,570 SAPS, https://www.sap.com/dmc/benchmark/2022/Cert22029.pdf. 2P Intel Xeon 8592+ powered server (128 total cores, 256 threads), 1.5TB Memory, 96,740 benchmark users, 528,560 SAPS, https://www.sap.com/dmc/benchmark/2023/Cert23077.pdf. For 2.137x the SAPS with the EPYC 9965 system. For 1.532x the SAPS with the EPYC 9654 system. Or (normalized to 9654): EPYC 9965 system has 1.395x the SAPS. Xeon 8592+ system has 0.654x the SAPS. For more details see https://www.sap.com/benchmark. SAP and SAP logo are the trademarks or registered trademarks of the SAP SE (or an SAP affiliate company) in Germany and in several other countries.
  9. 9xx5-005A: MySQL TPROC-C workload (SQL Server OLTP Brokerage) estimate based on internal AMD measurements as of 09/15/2024. The HammerDB TPROC-C workload is an open-source workload derived from TPC-BenchmarkTM Standard, and as such is not comparable to published TPC-C TM results, as the results do not comply with the TPC-C Benchmark Standard. Workload configs: MySQL 8.0.39, 8 core nodes (Multi-SUT), HammerDB-4.4, duration 5min, 32 v users, warehouses 128, aggregate New Orders Per Minute (NOPM); 2P AMD EPYC 9965 powered server (384 total cores), 2.35TB Memory, BIOS RVC100DB, OS VMWare ESXi 8.0.3 build 70965425, 1x1.6TB and 10x3.84TB storage. VM Configurations: 8 cores/VM, 48 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro; 2P AMD EPYC 9755 powered server (256 total cores), 2.35TB Memory, BIOS RVOT1000C, OS VMWare ESXi 8.0.3 build 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 32 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro; 2P AMD EPYC 9654 powered server (192 total cores), 1.5TB Memory, BIOS TVC100BD_2, OS VMWare ESXi 8.0.3 build 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 24 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off; 2P Intel Xeon 8592+ powered server (128 total cores), 1TB Memory, BIOS ESE124B, OS VMWare ESXi 8.0.3 build 24022510, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 16 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off. CPU Score (TPM) Relative_8592+ Relative_9654: Intel 8592+ (64c) 9431248 1 0.523; AMD EPYC 9654 (96c) 18037794 1.913 1; AMD EPYC 9755 (128c) 32598005 3.456 1.807; AMD EPYC 9965 (192c) 36863796 3.909 2.043. Results may vary based on factors including but not limited to system configurations, software versions, and BIOS settings. TPC, TPC Benchmark, and TPC-C are trademarks of the Transaction Processing Performance Council.
  10. 9xx5-054: MySQL TPROC-H workload (SQL Server OLTP Brokerage) estimate based on internal AMD measurements as of 09/15/2024. The HammerDB TPROC-H workload is an open-source workload derived from TPC-BenchmarkTM Standard, and as such is not comparable to published TPC-C TM results, as the results do not comply with the TPC-C Benchmark Standard. Workload configs: MySQL 8.0.39, 8 core nodes (Multi-SUT), SF30, HammerDB-4.4, 4 vusers. 2P AMD EPYC 9965 powered server (384 total cores), 2.35TB Memory, BIOS RVC100DB, OS VMWare ESXi 8.0.3 build 70965425, 1x1.6TB and 10x3.84TB storage. VM Configurations: 8 cores/VM, 48 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro. 2P AMD EPYC 9755 powered server (256 total cores), 2.35TB Memory, BIOS RVOT1000C, OS VMWare ESXi 8.0.3 build0 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 32 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro. 2P AMD EPYC 9654 powered server (192 total cores), 1.5TB Memory, BIOS TVC100BD_2, OS VMWare ESXii 8.0.3 build 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 24 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off. 2P Intel Xeon 8592+ powered server (128 total cores), 1TB Memory, BIOS ESE124B, OS VMWare ESXi 8.0.3 build 24022510, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 16 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off. CPU Score Relative_8592+ Relative_9654, Intel 8592+ (64c) 68718.4 1 0.693, AMD EPYC 9654 (96c) 99114.0 1.442 1, AMD EPYC 9755 (128c) 191726.7 2.790 1.934, AMD EPYC 9965 (192c) 247945.2 3.608 2.502. Results may vary based on factors including but not limited to system configurations, software versions, and BIOS settings. TPC, TPC Benchmark, and TPC-H are trademarks of the Transaction Processing Performance Council.
  11. 9xx5-055: NGINX workload based on internal AMD measurements as of 09/15/2024. Workload configs: NGINX 1.24.0-2ubuntu7, wrk workload, duration=60s, 400 connections, 8 core nodes, WRK workload, results in rps. 2P AMD EPYC 9965 powered server (384 total cores), 2.35TB Memory, BIOS RVC100DB, OS VMWare ESXi 8.0.3 build 70965425, 1x1.6TB and 10x3.84TB storage. VM Configurations: 8 cores/VM, 48 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro. 2P AMD EPYC 9755 powered server (256 total cores), 2.35TB Memory, BIOS RVOT1000C, OS VMWare ESXi 8.0.3 build0 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 32 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 6.8.0-39-generic BOOT_IMAGE=/vmlinuz-6.8.0-39-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro. 2P AMD EPYC 9654 powered server (192 total cores), 1.5TB Memory, BIOS TVC100BD_2, OS VMWare ESXii 8.0.3 build 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 24 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 6.8.0-39-generic BOOT_IMAGE=/vmlinuz-6.8.0-39-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off. 2P Intel Xeon 8592+ powered server (128 total cores), 1TB Memory, BIOS ESE124B, OS VMWare ESXi 8.0.3 build 24022510, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 16 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 6.8.0-39-generic BOOT_IMAGE=/vmlinuz-6.8.0-39-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off. CPU Score Relative_8592+ Relative_9654. Intel 8592+ (64c) 10664602.2 1 0.555. AMD EPYC 9654 (96c) 19232188.0 1.803 1. AMD EPYC 9755 (128c) 44492096.1 4.172 2.313. AMD EPYC 9965 (192c) 46429262.2 4.354 2.414. Results may vary based on factors including but not limited to system configurations, software versions, and BIOS settings.
  12. 9xx5-057: 9xx5-057: Redis workload based on internal AMD measurements as of 09/15/2024. Workload configs: Redis 7.4.0, data size 10000, tests GET SET, pipelinerequests=16. 2P AMD EPYC 9965 powered server (384 total cores), 2.35TB Memory, BIOS RVC100DB, OS VMWare ESXi 8.0.3 build 70965425, 1x1.6TB and 10x3.84TB storage. VM Configurations: 8 cores/VM, 48 VMs, 48GB memory, Ubuntu® 22.04.4 LTS, Linux® 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro. 2P AMD EPYC 9755 powered server (256 total cores), 2.35TB Memory, BIOS RVOT1000C, OS VMWare ESXi 8.0.3 build0 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 32 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 6.8.0-39-generic BOOT_IMAGE=/vmlinuz-6.8.0-39-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro. 2P AMD EPYC 9654 powered server (192 total cores), 1.5TB Memory, BIOS TVC100BD_2, OS VMWare ESXi 8.0.3 build 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 24 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 6.8.0-39-generic BOOT_IMAGE=/vmlinuz-6.8.0-39-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off. 2P Intel Xeon 8592+ powered server (128 total cores), 1TB Memory, BIOS ESE124B, OS VMWare ESXi 8.0.3 build 24022510, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 16 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 6.8.0-39-generic BOOT_IMAGE=/vmlinuz-6.8.0-39-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off. CPU GET SET Total Relative_8592+ Relative_9654, Intel 8592+ (64c) 26503956.6 21767203.7 48271160.3 1 0.390, AMD EPYC 9654 (96c) 72403055.2 51485426.6 123888481.8 2.567 1, AMD EPYC 9755 (128c) 84839653.7 69563201.3 154402855.1 3.199 1.246, AMD EPYC 9965 (192c) 128996111.9 110135933.7 239132045.6 4.954 1.930. Results may vary based on factors including but not limited to system configurations, software versions, and BIOS settings.
  13. 9xx5-006: AMD internal testing as of 09/01/2024, on FFMPEG (Raw to VP9, 1080P, 302 Frames, 1 instance/thread, video source: https://media.xiph.org/video/derf/y4m/ducks_take_off_1080p50.y4m). System Configurations: 2P AMD EPYC™ 9965 reference system (2 x 192C) 1.5TB 24x64GB DDR5-6400 running at 6000MT/s,   SAMSUNG MZWLO3T8HCLS-00A07, NPS=4, Ubuntu 22.04.3 LTS, Kernel Linux 5.15.0-119-generic, BIOS RVOT1000C (determinism enable=power),  10825484.25 Frames/Hour Median. 2P AMD EPYC™ 9654 production system (2 x 96C) 1.5TB 24x64GB DDR5-5600, , SAMSUNG MO003200KYDNC, NPS=4, Ubuntu 22.04.3 LTS, Kernel Linux 5.15.0-119-generic, BIOS 1.56 (determinism enable=power) , 5154133.333 Frames/Hour Median. 2P Intel Xeon Platinum 8592+ production system (2 x 64C) 1TB 16x64GB DDR5-5600, 3.2 TB NVME, Ubuntu 22.04.3 LTS, Kernel Linux 6.5.0-35-generic), BIOS ESE122V-3.10, 2712701.754 Frames/Hour Median. For 3.99x the performance with the AMD EPYC 9965 vs Intel Xeon Platinum 8592+ systems. For 1.90x the performance with the AMD EPYC 9654 vs Intel Xeon Platinum 8592+ systems. Results may vary based on factors including but not limited to BIOS and OS settings and versions, software versions and data used. Intel 8592+, (64c) 3784640.5563 1 0.492, AMD EPYC 9654 (96c) 7690521.2 2.032 1, AMD EPYC 9755 (128c) 10122712.1 2.675 1.316, AMD EPYC 9965 (192c) 16422386.0 4.339 2.135. Results may vary based on factors including but not limited to system configurations, software versions, and BIOS settings.
  14. 9xx5-053: Cassandra results based on AMD internal testing as of 09/15/2024. Workload configs: Cassandra 4.1.6, 8-core nodes, results in aggregate operations/second stress workload, (3 read, 1 write), 10,000,000 records. 2P AMD EPYC 9755 powered reference server (256 total cores), 2.35TB Memory, BIOS RVOT1000C, OS VMWare ESXi 8.0.3 build0 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 32 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro. 2P AMD EPYC 9654 powered server (192 total cores), 1.5TB Memory, BIOS TVC100BD_2, OS VMWare ESXii 8.0.3 build 70965425, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 24 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic, BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off. 2P Intel Xeon 8592+ powered server (128 total cores), 1TB Memory, BIOS ESE124B, OS VMWare ESXi 8.0.3 build 24022510, 1x1.6TB and 8x3.84TB storage. VM Configurations: 8 cores/VM, 16 VMs, 48GB memory, Ubuntu 22.04.4 LTS, Linux 5.15.0-119-generic BOOT_IMAGE=/vmlinuz-5.15.0-119-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro spec_rstack_overflow=off. CPU Score Relative_8592+ Relative_9654, Intel 8592+ (64c) 1475403 1 0.445, AMD EPYC 9654 (96c) 3316359 2.25 1, AMD EPYC 9755 (128c) 5525390 3.75 1.667, Results may vary based on factors including but not limited to system configurations, software versions, and BIOS settings.
  15. 9xx5-062: Memcached™ workload based on internal AMD measurements as of 09/26/2024. Workload configs: Memcached 1.6.24, (Client command line: memtier_benchmark --hide-histogram -t 8 -P memcache_text -c 10 -p $port --pipeline=8 --ratio=1:10 --test-time=60), (Server command line: memcached -p $port -c 4096 -t 8 -u memcache). 2P AMD EPYC 9965 powered server (384 total cores), 1.5TB 24x64GB DDR5-6400 (running at 6000 MT/s), BIOS RVOT1000C, OS Ubuntu 24.04.1 LTS Linux 6.8.0-35-generic, 1x1.6TB SAMSUNG MZVL2512HCJQ-00B00 and 10x3.84TB storage SAMSUNG MZWLO3T8HCLS-00A07, BOOT_IMAGE=/boot/vmlinuz-6.8.0-35-generic root=UUID=bb4e178b-796c-4388-a452-5a311c97b97c ro. 2P AMD EPYC 9755 powered server (256 total cores), 1.5TB 24x64GB DDR5-6400 (running at 6000 MT/s), BIOS RVOT1001A, OS Ubuntu 24.04.1 LTS Linux 6.8.0-35-generic, 1x1.6TB SAMSUNG MZVL2512HCJQ-00B00 and 2x3.84TB storage SAMSUNG MZWLO3T8HCLS-00A07, BOOT_IMAGE=/boot/vmlinuz-6.8.0-35-generic root=UUID=5142bd33-94fb-4496-bbe3-2614a2004ba7 ro. 2P AMD EPYC 9654 powered server (192 total cores), 1.5TB 24x64GB DDR5-4800, BIOS RTI100DB, OS Ubuntu 24.04.1 LTS Linux 6.10.0+, 1x1.6TB KXG60ZNV256G TOSHIBA and 8x3.84TB storage SAMSUNG MZWLO3T8HCLS-00A07, BOOT_IMAGE=/boot/vmlinuz-6.10.0+ root=UUID=69d9b65b-2fab-4731-8496-d6c8147a5e7a ro spec_rstack_overflow=off. 2P Intel Xeon 8592+ powered server (128 total cores), 1TB 16x64GB DDR5-5600, BIOS ESE122V, OS Ubuntu 24.04.1 LTS Linux 6.8.0-45-generic, 4x3.84TB SAMSUNG MZWLO3T8HCLS-00A07 storage, BOOT_IMAGE=/vmlinuz-6.8.0-45-generic root=/dev/mapper/ubuntu--vg-ubuntu--lv ro. CPU Score Relative_8592+ Relative_9654, Intel 8592+ (64c) 28685109.51 1 0.571, AMD EPYC 9654 (96c) 50217042 1.751 1, AMD EPYC 9755 (128c) 101861435.92 3.551 2.028, AMD EPYC 9965 (192c) 122462349.57 4.269 2.439. Results may vary based on factors including but not limited to system configurations, software versions, and BIOS settings.
  16. 9xx5-092: VMmark® 4.0.1 host/node FC SAN comparison based on published results as of 10/10/2024. Configurations: 2 node, 2P AMD EPYC 9965 (384 total cores) powered server running VMware ESXi8.0 U3, 5.17 @ 5.8 tiles, https://www.vmware.com/docs/2024-10-10-Dell-PowerEdge-R7725-5-17. 2 node, 2P AMD EPYC 9755 (256 total cores) powered server running VMware ESXi 8.0 U3, 4.53 @ 5 tiles, https://www.vmware.com/docs/2024-10-10-Dell-PowerEdge-R7725-4-53. 2 node, 2P AMD EPYC 9654 (192 total cores) powered server running VMware ESXi 8.0 U3, 3.43 @ 4 tiles, https://www.vmware.com/docs/2024-10-03-Supermicro-AS-2125HS-TNR. VMmark is a registered trademark of VMware in the US or other countries.
  17. 9xx5-071: VMmark® 4.0.1 host/node FC SAN comparison based on “independently published” results as of 10/10/2024. Configurations: 2 node, 2P AMD EPYC 9575F (128 total cores) powered server running VMware ESXi8.0 U3, 3.31 @ 4 tiles, https://www.infobellit.com/BlueBookSeries/VMmark4-FDR-1003. 2 node, 2P AMD EPYC 9554 (128 total cores) powered server running VMware ESXi 8.0 U3, 2.64 @ 3 tiles,  https://www.infobellit.com/BlueBookSeries/VMmark4-FDR-1002. 2 node, 2P Intel Xeon Platinum 8592+ (128 total cores) powered server running VMware ESXi 8.0 U3, 2.06 @ 2.4 Tiles, https://www.infobellit.com/BlueBookSeries/VMmark4-FDR-1001. VMmark is a registered trademark of VMware in the US or other countries.
  18. 9xx5-085: SPECrate®2017_int_energy_base comparison based on independently published scores from www.spec.org as of 10/10/2024. 2P AMD EPYC 9965 (2430 SPECrate®2017_int_energy_base, 384 Total Cores,  https://www.infobellit.com/BlueBookSeries/SPECINT-Energy-FDR-1003). 2P AMD EPYC 9654 (1800 SPECrate®2017_int_energy_base, 192 Total Cores, https://www.infobellit.com/BlueBookSeries/SPECINT-Energy-FDR-1002). 2P Intel Xeon Platinum 8592+ (1110 SPECrate®2017_int_energy_base, 128 Total Cores,  https://www.infobellit.com/BlueBookSeries/SPECINT-Energy-FDR-1001). SPEC®, SPEC CPU®, and SPECrate® are registered trademarks of the Standard Performance Evaluation Corporation. See www.spec.org for more information. Intel CPU TDP and pricing at https://ark.intel.com/ as of 10/3/2024.
About the Author
Raghu Nambiar currently holds the position of Corporate Vice President at AMD, where he leads a global engineering team dedicated to shaping the software and solutions strategy for the company's datacenter business. Before joining AMD, Raghu served as the Chief Technology Officer at Cisco UCS, instrumental in driving its transformation into a leading datacenter compute platform. During his tenure at Hewlett Packard, Raghu made significant contributions as an architect, pioneering several groundbreaking solutions. He is the holder of ten patents, with several more pending approval, and has made extensive academic contributions, including publishing over 75 peer-reviewed papers and 20 books in the LNCS series. Additionally, Raghu has taken on leadership roles in various industry standards committees. Raghu holds dual Master's degrees from the University of Massachusetts and Goa University, complemented by completing an advanced management program at Stanford University.