AMD Unveils EPYC 'Milan' 7003 CPUs, Zen 3 Comes to 64-Core Server Chips
AMD unveiled its EPYC 7003 'Milan' processors today, claiming that the chips, which bring the company's powerful Zen 3 architecture to the server market for the first time, take the lead as the world's fastest server processor with its flagship 64-core 128-thread EPYC 7763. Like the rest of the Milan lineup, this chip comes fabbed on the 7nm process and is drop-in compatible with existing servers. AMD claims it brings up to twice the performance of Intel's competing Xeon Cascade Lake Refresh chips in HPC, Cloud, and enterprise workloads, all while offering a vastly better price-to-performance ratio.
Milan's agility lies in the Zen 3 architecture and its chiplet-based design. This microarchitecture brings many of the same benefits that we've seen with AMD's Ryzen 5000 series chips that dominate the desktop PC market, like a 19% increase in IPC and a larger unified L3 cache. Those attributes, among others, help improve AMD's standing against Intel's venerable Xeon lineup in key areas, like single-threaded work, and offer a more refined performance profile across a broader spate of applications.
One interesting new SKU is the EPYC 7663, a 56-core, 112-thread CPU with 7 working cores on each of the 8-core chiplets. There is also a 28-core EPYC 7453.
Next up, Zen 4 "Genoa".
Also at AnandTech, The Next Platform, Phoronix, and Ars Technica.
See also: The Tour of Italy with EPYC Milan: Interview with AMD's Forrest Norrod
AMD video announcement (51m4s) and recap (10m43s)
(Score: 3, Interesting) by bzipitidoo on Tuesday March 16 2021, @05:13PM (4 children)
Dang, I'm still running 2 core, 2 threads per core hardware with 8G or even just 4G RAM. Has the gap between high end server stuff and low end consumer laptop/tablet grown larger? My experience is that a 5 year old server is about equivalent in performance to a mid-range new consumer grade personal computer, while using twice as much or more power.
(Score: 2) by takyon on Tuesday March 16 2021, @06:10PM (3 children)
Workstations and server owners can actually benefit from extremely high core counts and extra memory channels, or at least, they can benefit before consumers will.
Broadly speaking, there's single-threaded and multi-threaded performance. A consumer-grade 5950X is going to clock up to around 4.9 GHz, higher than most if not all of these Epyc CPUs, and at 16 cores, it will handle everything most consumers throw at it. Compare it to a 64-core Epyc with lower clock speeds, and the user experience would actually worsen with the 64-core system.
With Zen 4, we should see 16 or 24 cores for consumers and 96 cores for workstations/servers.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 0) by Anonymous Coward on Wednesday March 17 2021, @01:52AM (2 children)
Why do servers have so many cores, memory channels and pci lanes, but lower clock speeds? IOW, why are server CPUs and boards designed the way they are in comparison to desktop boards and CPUs?
(Score: 2) by takyon on Wednesday March 17 2021, @05:56AM (1 child)
Here are some potential reasons:
There are CPUs in the lineup that clock higher, like the 32-core EPYC 75F3, or the EPYC 72F3, an expensive 8-core with only 1 working core per chiplet.
[SIG] 10/28/2017: Soylent Upgrade v14 [soylentnews.org]
(Score: 0) by Anonymous Coward on Wednesday March 17 2021, @09:07PM
cool, thanks