TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

AMD's Future in Servers: New 7000-Series CPUs Launched and EPYC Analysis

365 pointsby sataialmost 8 years ago

16 comments

DuskStaralmost 8 years ago
4 dies per package is a pretty interesting way of doing things - probably helps yields immensely, but I can't imagine it does anything good for intra-processor latency. 142 ns to ping a thread on a different CCX within a die isn't too horrible, but I really want to know what sort of penalty you'll have from going to a different die within a package.
评论 #14598628 未加载
评论 #14599427 未加载
评论 #14599016 未加载
评论 #14601565 未加载
评论 #14598970 未加载
评论 #14598398 未加载
评论 #14600683 未加载
mastazialmost 8 years ago
For anyone looking for info about the socket:<p>* Epyc uses socket SP3 <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Socket_SP3" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Socket_SP3</a><p>* Threadripper uses socket TR4 <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Socket_TR4" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Socket_TR4</a><p>* Sockets SP3 and TR4 have the same number of pins (4094 pins) and they have the same cooler bracket mount (see <a href="https:&#x2F;&#x2F;www.overclock3d.net&#x2F;news&#x2F;cases_cooling&#x2F;noctua_showcase_epyc_threadripper_ready_tr4_sp3_ready_cpu_coolers&#x2F;1" rel="nofollow">https:&#x2F;&#x2F;www.overclock3d.net&#x2F;news&#x2F;cases_cooling&#x2F;noctua_showca...</a> )<p>* However they are still two separate sockets so you shouldn&#x27;t expect to be able to use Epyc on TR4 or Threadripper on SP3
评论 #14600216 未加载
myrandomcommentalmost 8 years ago
I would really love it if there was a benchmark around running VMs and containers for something like this. Our dev&#x2F;test system is all docker containers so that is what we would care about.<p>I guess it would be hard as there are to many ways to scale out what you run - how many VMs, how many containers, what are you running in them? It would be an interesting benchmark matrix to sort for.<p>It would be interesting just to see how many containers you could start, run lighttpd and each server a static web page? Maybe 1&#x2F;2 with the page and 1&#x2F;2 with an application that builds the page? Who knows...to many variables.<p>I think we will just by a system when we can and try our workload on it. Oh, well.
评论 #14598815 未加载
评论 #14599643 未加载
hyperbovinealmost 8 years ago
Soooo the Linux kernel now compiles in 15.6 seconds. Jeebus I feel old...
评论 #14598406 未加载
评论 #14598502 未加载
评论 #14598798 未加载
评论 #14598938 未加载
评论 #14599736 未加载
dbcooperalmost 8 years ago
Baidu and Microsoft will be customers:<p><a href="https:&#x2F;&#x2F;www.bloomberg.com&#x2F;news&#x2F;articles&#x2F;2017-06-20&#x2F;amd-server-chip-revival-effort-enlists-some-big-friends" rel="nofollow">https:&#x2F;&#x2F;www.bloomberg.com&#x2F;news&#x2F;articles&#x2F;2017-06-20&#x2F;amd-serve...</a>
评论 #14604342 未加载
sataialmost 8 years ago
1 socket 16 &#x2F; 32 @ 2.9GHz max for $700+... it looks like 16 core Threadripper with reasonable frequencies for less then $999 looks in reach...
评论 #14599673 未加载
评论 #14598841 未加载
girstalmost 8 years ago
intel had a monopoly on high-end chipsets for _far_ too long. I&#x27;m glad, there is some competition.
gbrown_almost 8 years ago
Those TDPs look pretty high, what are vendors willing to put into 1U high 0.5U wide style servers with 2 sockets these days? Last I looked I seem to recall it was around up to 145W.
评论 #14598634 未加载
评论 #14598691 未加载
评论 #14598977 未加载
评论 #14598539 未加载
评论 #14598508 未加载
bsaulalmost 8 years ago
A bit off topic, but does anyone knows if AI ( aka modern neural networks) plays a role in cpu design nowadays ?
评论 #14599038 未加载
评论 #14603062 未加载
评论 #14599124 未加载
评论 #14599181 未加载
评论 #14599079 未加载
评论 #14599210 未加载
nik736almost 8 years ago
Why does AMD compare their single socket CPUs to Intels E5-2XXX line? Intel has E5-1XXX single socket CPUs.
评论 #14598349 未加载
评论 #14598658 未加载
评论 #14598341 未加载
评论 #14598339 未加载
jnordwickalmost 8 years ago
I didn&#x27;t see any info on the cpu cache architecture which governs performance for many applications now.<p>Anybody have any info on things like L0 to L2 size, type, latencies, etc?
评论 #14599797 未加载
评论 #14599032 未加载
dangalmost 8 years ago
Related: <a href="https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=14598660" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=14598660</a>
Keyframealmost 8 years ago
What&#x27;s the SSEs and AVXs performance like on Ryzen&#x2F;EPYC compared to intel?
评论 #14602549 未加载
irishjohnniealmost 8 years ago
Wow! AMD EPYC + Xilinx FPGA!
评论 #14600651 未加载
greptomaniaalmost 8 years ago
While I&#x27;m excited to see AMD&#x27;s offering, as a scientific-HPC user I can&#x27;t help but wonder how much marketshare AMD will be able to gain without more information on supporting software - specifically good compilers + math libraries (cf. Intel compilers + MKL).<p>Strangely, I&#x27;ve not seen much on HN, or elsewhere, make mention of AMD&#x27;s software support. Is this because it doesn&#x27;t exist, or because compilers are less &quot;sexy&quot; than shiny new hardware?
评论 #14607273 未加载
评论 #14604652 未加载
garaetjjtealmost 8 years ago
&gt;In this case, an EPYC 7281 in single socket mode is listed as having +63% performance (in SPECint) over a dual socket E5-2609v4 system.<p>So, quad-CPU is faster than dual-CPU? Not surprising.
评论 #14598514 未加载
评论 #14598411 未加载