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.

Correcting Intel's Deep Learning Benchmark Mistakes

141 pointsby Smerityalmost 9 years ago

11 comments

ChuckMcMalmost 9 years ago
Ok, its officially the new new thing when corporate communications types are sniping at each other :-)<p>Once we start seeing press releases of products that are not available yet being shown to kill the existing competition we&#x27;ll know that the hype train has officially gone super-product-sonic (that is a hype wave travelling faster than the product releases can support it).
评论 #12301383 未加载
评论 #12301285 未加载
modelessalmost 9 years ago
&gt; Titan uses four-year-old GPUs<p>... as does nearly every public cloud provider. I agree with most of the article, but you can&#x27;t fault Intel for benchmarking the hardware that cloud providers are actually offering.<p>I&#x27;m not sure what exactly NVIDIA is doing with their Tesla product line but whatever it is, it&#x27;s really restricting the availability of recent GPU hardware. Even Azure&#x27;s GPU instances released this month are using the Kepler architecture from 2012. It&#x27;s fully two generations out of date now, and that&#x27;s sad.
评论 #12301263 未加载
评论 #12303606 未加载
评论 #12302004 未加载
cynodalmost 9 years ago
@imaleppert Agree getting the details on the testing would be helpful. I think NV was more pointing out that Intel were putting their latest against NV&#x27;s oldest. It&#x27;d be like testing a RX480 against a GTX660. What&#x27;s the use of that?<p>@modeless the new Azure instances have M60&#x27;s or you can purchase a 1080 or new TitanX which are both available (although stock has been tight).<p><a href="https:&#x2F;&#x2F;azure.microsoft.com&#x2F;en-us&#x2F;blog&#x2F;azure-n-series-preview-availability&#x2F;" rel="nofollow">https:&#x2F;&#x2F;azure.microsoft.com&#x2F;en-us&#x2F;blog&#x2F;azure-n-series-previe...</a>
评论 #12301439 未加载
iamleppertalmost 9 years ago
Why don&#x27;t they provide a link to their testing methodology? They need to back up their claims (on both sides) with the actual configuration, all versions, and sample datasets for people to independently verify.<p>A docker container that runs their performance suite would be ideal.
评论 #12301462 未加载
评论 #12303616 未加载
Gladdyualmost 9 years ago
`each with one or two sockets of less-capable processors, like Xeon Phi.`<p>Corporate smacktalk...
StreamBrightalmost 9 years ago
I do not usually believe in any vendor provided performance benchmark unless I fabricated it myself. On a more serious note, benchmarking is pretty hard and people usually discount things that seem minor until you found it the bottleneck, like the interconnect in this case for example. Another problem with synthetic benchmarks is that you can always optimize for your exact use case and it usually yields to pretty good improvements, comparable to buying faster equipment. The ultimate quiestion which is more cost efficient, buying faster CPU&#x2F;GPU or hiring a performance expert.
ldarginalmost 9 years ago
I&#x27;m a bit surprised that Nvidia mentioned nothing about performance per watt in their reply.
评论 #12301329 未加载
评论 #12301454 未加载
nzjrsalmost 9 years ago
I actually came away from the article thinking, heh Phi looks pretty good.
dmoyalmost 9 years ago
Well that was a pretty concise explanation of mistakes. Benchmarking is indeed difficult to get right.
评论 #12302097 未加载
sidkashyapover 8 years ago
They do not mention the version of Caffe used to test the Intel systems, Intel claims its numbers based on an optimized branch of Caffe and not the public (BVLC) version
NotQuantumalmost 9 years ago
Shots fired, wow Nvidia is pissed that Intel is trying to edge into the market with their Phi series.