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。

Ask HN: Why do all cloud providers have Tesla K80 vs. cheaper Titan X or 1080ti?

5 点作者 anilshanbhag大约 8 年前
For Tesla: K80 seems to have higher double precision flops<p>For Titan X&#x2F;1080ti: &gt; In terms of raw performance the titan x and 1080 ti have more flops per GPU. &gt; They are significantly cheaper (~1000$ vs 4000$)<p>There are numbers which show that the common nets like AlexNet are faster on Titan X (https:&#x2F;&#x2F;plot.ly&#x2F;~JianminSun&#x2F;4&#x2F;nvidia-titan-x-pascal-vs-nvidia-tesla-k80&#x2F;). However, most cloud providers like Amazon, Azure and Google recently introduced gpu instances with K80&#x27;s. Wouldn&#x27;t it be better to just have cloud instances with 1080ti s&#x2F;Titan X s ?

2 条评论

webmaven大约 8 年前
Consider that the K80 gives you more FLOPS-per-$ (just the chip cost, ie. CapEx), and also more FLOPS-per-watt (ie. OpEx).
foo80大约 8 年前
K80 has ECC memory, while enthusiast card like 1080 do not. K80 is dual gpu so it takes less space (more gpu per node)