AMD EPYC processors have been steadily improving over the past seven years, offering increased performance and energy efficiency with each new generation. The latest innovation, 3D V-Cache technology, has significantly boosted performance for demanding workloads.

How 3D V-Cache Works

• Increases L3 cache size: By stacking additional cache layers, 3D V-Cache triples the L3 cache compared to standard EPYC processors.
• Benefits demanding workloads: Particularly effective for design and simulation tasks like computational fluid dynamics (CFD), finite element analysis (FEA), and electronic design automation (EDA).

Real-World Impact

• STMicroelectronics: Achieved a 12% performance boost in EDA and CAD workloads using 3rd Gen EPYC with 3D V-Cache compared to other CPUs.
• Performance benchmarks: Showed substantial improvements in various EDA tools and CFD simulations with the latest 4th Gen EPYC processors featuring 3D V-Cache.

Overall, AMD EPYC processors with 3D V-Cache are a powerful solution for organizations handling technical computing workloads.

At this year's Red Hat Summit, we at AMD had our work cut out for us. That's because for the better part of its existence, Red Hat's OpenShift has been largely associated with Intel^® Xeon^® processors. And that's fair enough -- when OpenShift was first released in the early 2010s, Intel had little real competition in the server market. But as anyone in the tech industry knows, a lot can change in 10+ years. And a lot has!  The Summit was our opportunity to showcase just how much as evolved.

Among the favorite subjects here on the AMD EPYC TechTalk podcast series is artificial intelligence. We especially enjoy hearing about how enterprise companies are faring on their AI journeys. I recently had the pleasure of speaking with Robert Daigle, director & global AI business leader at Lenovo. Robert spoke about the state of AI in the enterprise and how his group helps companies find AI solutions for their particular business. Highlights from our talk are posted below.

I always enjoy learning from people who understand how essential collaboration is to innovation. Alan Benjamin, CEO of GigaIO, is one of those people--and he was kind enough to spend some time speaking with us on our AMD TechTalk podcast. GigaIO, a company that provides workload-defined infrastructure for AI and technical computing, made news last year by configuring 32 AMD Instinct MI210 accelerators to a single-node server called the SuperNODE. Up to then, accessing 32 GPUs required four servers equipped with eight GPUs apiece — as well as all the cost and latency associated with linking up all of that extra hardware. I've included some of the interview highlights in the blog post below.

As much as I love advocating for AMD EPYC technologies, I have a special fondness for sharing what our partners and customers have to say about working with AMD and the EPYC processor portfolio.  It is one of my priorities when selecting topics and guests for the AMD TechTalk podcast.  Paul Sun, of IronYun, recently joined the podcast and spoke with me about the growing opportunities in visual analytics — thanks to advancements in AI-vision applications, Edge computing and new hardware. I've included highlights from our chat in this brief blog. 

Rebecca Weekly, vice president of infrastructure at Geico, is part of the leadership team overseeing digital transformation for the country's second-largest car insurer.  Rebecca kindly shared some of her experiences with the AMD EPYC TechTalk podcast series. The full interview can be heard here, it is a great listen. You can also find highlights from the interview below.

When facing the renewal of software licenses, the financial impact can be daunting, prompting considerations on cost-effective strategies. AMD offers a unique approach: when evaluating licensing requirements, refreshing infrastructure with AMD EPYC CPUs can prove to be extremely cost effective. This is particularly advantageous for software licensed per-core, such as VMware solutions, certified to run seamlessly on AMD infrastructure. AMD EPYC processors, including high-frequency variants, can reduce core count requirements, thereby lowering software license expenses significantly. Migration tools like VMware Architecture Migration Tool (VAMT) facilitate moving virtual machines from Intel to AMD clusters, ensuring a smooth transition with features like Change Window Support and Process Throttling. Bottom line - migrating to AMD EPYC based solutions can not only enhances performance but also improve total cost of ownership (TCO), security, and overall value.

Chaos develops 3D visualization technology for architecture, engineering, construction, product design, manufacturing, and media and entertainment. The company’s software solutions run on and are optimized for AMD EPYC processor-based systems. I’m happy to be able to share some highlights from our discussion in this brief blog post.  You can listen to the full podcast here.   

Look beyond the hype. AMD EPYC Processors outperform the NVIDIA Grace CPU Superchip across key workloads without compromises and expensive architectural transitions.

The confluence of high performance, highly efficient hardware and market leading, run-the-business software means productivity and innovation.  Ashish Ray of Oracle recently joined the AMD TechTalk podcast to share his thoughts on technology and business.  This article provides a summary of some of his key points.