Setting aside potential setbacks caused by U.S. trade policies, the steady cadence of AMD’s revival in HPC and the datacenter continued last week with AWS expanding availability of its AMD Epyc-based instances. Recall just a month ago, AMD announced it had physically verified its largest 7nm chip design – the massive 13.2-billion transistor Radeon Instinct Vega2 GPU. The latter effort was an important step given uncertainty over rival Intel’s manufacturing advancement.
Both Intel and AMD have been caught up in U.S. national security worries over China which resulted in blacklisting a number of important Chinese systems makers from purchasing advanced U.S. technology (See HPCwire article, Chinese Company Sugon Placed on US ‘Entity List’ After Strong Showing at International Supercomputing Conference).
Adoption by big cloud providers, always important to success for chipmakers, is even more important now and AMD has been successful in convincing big hyperscalers to take a chance on Epyc. AWS went so far as to note its AMD-based instances may deliver a 10 percent cost saving for some workflows. Intel, of course, remains the dominant player.
Here are the specifics of the new AWS availability regions:
- M5a and R5a– AWS Europe (Paris), US West (San Francisco), Canada (Montreal) and AWS GovCloud (US) Regions
- M5ad and R5ad– AWS Asia Pacific (Sydney), Canada (Montreal) and AWS GovCloud (US-West) Regions
- T3a– AWS Asia Pacific (Seoul, Sydney, Tokyo), Europe (Paris, London, Frankfurt), US West (San Francisco), Canada (Montreal) and AWS GovCloud (US) Regions
AWS provided a brief characterization of workflows well-suited for these instances in its announcement.
It reported M5a and M5ad instances as “ideal for business-critical applications, web and application servers, back-end servers for enterprise applications, gaming servers, caching fleets, and app development environments.”
The R5a and R5ad instances target high performance databases, distributed web scale in-memory caches, mid-size in-memory databases, real time big data analytics, and other enterprise applications.
AWS described T3a instances as offering a “balance of compute, memory, and network resources for a broad spectrum of general purpose workloads including micro-services, low-latency interactive applications, small and medium databases, virtual desktops, development environments, code repositories, and business-critical applications.”
Taken together, the AWS expansion and physical verification of its 7nm design, help maintain forward AMD momentum. Interestingly, the AMD Radeon Instinct Vega20 was tested using a TSMC-certified Calibre nmDRC software platform from Mentor and actually run on an Azure cloud platform, powered by AMD’s own Epyc processors.
Said Daniel Bounds, senior director of AMD Datacenter Products, “AMD demands speed and quality of execution in our cutting-edge semiconductor design work, so achieving two verification passes in one day in the cloud is critical to getting future designs to market. AMD is pleased to see that Mentor’s Calibre nmDRC scales on cloud-based AMD Epyc-powered servers not just in traditional use models, but also on the Azure public cloud.” (See HPCwire coverage, AMD Verifies Its Largest 7nm Chip Design in Ten Hours)