TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Groq, a Stealthy Startup Founded by Google’s TPU Team, Is Raising $60M

115 点作者 exotree超过 6 年前

14 条评论

writepub超过 6 年前
Hardware is hard. For Gorq to be successful -<p>- The boards using their chips need to fit into commodity interfaces (PCIE? DIMM? ) in Open Compute hardware<p>- Someone needs to buy hundreds of thousands of these to even minutely impact their bottom line<p>- Multiple such big volume wins need to consistently happen.<p>- Their IP needs to actually be defensible. Otherwise, an Intel&#x2F;Samsung whose manufacturing prowess &amp; channel reach is multiple times that of Gorq will undercut the pricing with <i>almost</i> the same performance&#x2F;watt. Oh, and they&#x27;ll happily play nice with Open Compute, other standards bodies.<p>- Most of all, their product needs to work as advertised, at scale, in a reliable fashion. This is easier said than done in semiconductors, especially for the kind of performance gains they&#x27;re marketing.<p><i>If</i> they&#x27;d gotten here with ~$30M capital, and demonstrated traction in the marketplace, I&#x27;d give them a chance, but expecting Google&#x27;s pay while working at an independent chip startup, with $0 in revenues portends financial doom. It&#x27;s not the founder&#x27;s fault though - hardware is capital intensive, not compatible with agile development &amp; an MVP just won&#x27;t cut it - it needs to be fully functional &amp; reliable right out of the bat. I sincerely hope I&#x27;m wrong, as I&#x27;d like to see the silicon put back in silicon valley - having been in the semiconductor industry for 20 years, it just seems unlikely.
评论 #17929422 未加载
评论 #17930153 未加载
评论 #17928737 未加载
评论 #17931258 未加载
评论 #17931471 未加载
评论 #17928600 未加载
nicodjimenez超过 6 年前
Delivering an incremental speed boost vs nvidia chips seems like a tough way to win. Even if they can be 2x faster, people will still stick to cuda. They either have to be 5-10x faster or find a new market. Definitely possible. Would be pretty embarrassing for nvidia if that happened.
评论 #17932172 未加载
评论 #17931988 未加载
评论 #17932616 未加载
sytelus超过 6 年前
The numbers on their website is quite stunning if they are true:<p>- 16X more power efficient than TitanX<p>- 3X more ops than TitanX<p>- 25K images&#x2F;sec vs 5K images&#x2F;sec inference on nVidia<p>I&#x27;m completely bewildered why NVidia hasn&#x27;t came up with deep learning specific chips yet that doesn&#x27;t have crud of massive rendering pipeline.<p><a href="https:&#x2F;&#x2F;blog.groq.com&#x2F;2017&#x2F;11&#x2F;09&#x2F;69&#x2F;" rel="nofollow">https:&#x2F;&#x2F;blog.groq.com&#x2F;2017&#x2F;11&#x2F;09&#x2F;69&#x2F;</a>
评论 #17932604 未加载
scottlegrand2超过 6 年前
If it&#x27;s FP32 or FP16&#x2F;32, it&#x27;s interesting. If it&#x27;s INT8&#x2F;32 it&#x27;s incrementally better than a 2080TI GPU, and if it&#x27;s INT4&#x2F;32, it&#x27;s stillborn.
评论 #17929434 未加载
评论 #17929134 未加载
syntaxing超过 6 年前
I&#x27;m kind of curious how many electrical engineers and talent that can design ASICs are out there? Most electrical engineers that I have met that designed ASICs at one point or another were for mainly military use or in the Semi industry. But the team that designed the chips were always super small compared to the mechanical or electrical team.
评论 #17929430 未加载
woodrowbarlow超过 6 年前
founded by _former members_ (albeit founding members) of Google&#x27;s TPU team.<p>i wonder how they negotiated this from a legal standpoint. every employment contract i&#x27;ve ever signed certainly would not allow for starting a project this similar to my employer&#x27;s core business.
评论 #17928184 未加载
评论 #17928202 未加载
Soundest超过 6 年前
This looks interesting. Whilst I agree with other commenters that it&#x27;s hard to compete in hardware I think there&#x27;s a good niche for this product. Google isn&#x27;t going to start selling TPUs so something off-the-shelf for machine learning might make some real traction. Best case, lot&#x27;s of sales to cloud providers (amazon, microsoft etc.) and lots of custom houses. Worst case would be acquisition by MS or Amazon.<p>Having said that, it&#x27;s certainly true that Nvidia are tough to beat. But right now we&#x27;re in a bubble, VC will throw millions at companies and big corporations will throw billions at acquisitions. So I think it&#x27;s probably a very profitably move in general.
评论 #17929627 未加载
jmunjr超过 6 年前
This is troubling:<p><a href="https:&#x2F;&#x2F;seekingalpha.com&#x2F;article&#x2F;4206948-nvidias-inference-problem-alarming-sell-side-ai-iq" rel="nofollow">https:&#x2F;&#x2F;seekingalpha.com&#x2F;article&#x2F;4206948-nvidias-inference-p...</a>
HNNewer超过 6 年前
I believe they got so much funding because they are coming from Google, not for the product itself. They could have sold even crap hardware.
lquist超过 6 年前
Is this competitive to NVIDIA&#x27;s chips? How worried should they be?
评论 #17928853 未加载
评论 #17932731 未加载
nobrains超过 6 年前
Are they planning to create cryptocurrency miners?
person_of_color超过 6 年前
Are they hiring?
评论 #17930005 未加载
css超过 6 年前
&gt; a company with a very spartan website<p>What does the word &quot;spartan&quot; mean in this context? &quot;Serious?&quot; &quot;Utilitarian?&quot; I do not know this usage of the word.<p>Edit (from Webster):<p>&gt; 2 b <i>often not capitalized</i>: marked by simplicity, frugality, or avoidance of luxury and comfort. &quot;A spartan room&quot;
评论 #17929975 未加载
评论 #17929930 未加载
anonymous5133超过 6 年前
Definetly seems interesting.