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。

Pushing AMD's Infinity Fabric to Its Limit

229 点作者 klelatti6 个月前

4 条评论

majke6 个月前
This has puzzled me for a while. The cited system has 2x89.6 GB&#x2F;s bandwidth. But a single CCD can do at most 64GB&#x2F;s of sequential reads. Are claims like &quot;Apple Silicon having 400GB&#x2F;s&quot; meaningless? I understand a typical single logical CPU can&#x27;t do more than 50-70GB&#x2F;s, and it seems like a group of CPU&#x27;s typically shares a mem controller which is similarly limited.<p>To rephrase: is it possible to cause 100% mem bandwith utilization with only or 1 or 2 CPU&#x27;s doing the work per CCD?
评论 #42236442 未加载
评论 #42237162 未加载
评论 #42235848 未加载
评论 #42239911 未加载
评论 #42241556 未加载
cebert6 个月前
George’s detailed analysis always impresses me. I’m amazed with his attention to detail.
评论 #42232445 未加载
评论 #42240548 未加载
Agingcoder6 个月前
Proper thread placement and numa handling does have a massive impact on modern amd cpus - significantly more so than on Xeon systems. This might be anecdotal, but I’ve seen performance improve by 50% on some real world workloads.
评论 #42236834 未加载
评论 #42236282 未加载
AbuAssar6 个月前
Great deep dive into AMD&#x27;s Infinity Fabric! The balance between bandwidth, latency, and clock speeds shows both clever engineering and limits under pressure. Makes me wonder how these trade-offs will evolve in future designs. Thoughts?
评论 #42235115 未加载