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 Announces 7nm Rome CPUs and MI60 GPUs

291 pointsby ccwilson10over 6 years ago

14 comments

berbecover 6 years ago
With AMD closing the gap with Intel competition in the CPU market is back after a ten year abscense.<p>What I&#x27;m waiting for is an AMD GPU that can compete with a top-tier NVidia offering. Vega is nice, but not really a contender on the mid to top end. The G series cpus with vega inside are great, but where is the 2080ti, or even 1080ti, killer? Even something a bit slower, but close, would be great.<p>Is it too much to ask of AMD to handle both? I am unsure, but I would love to see NVidia in a price and performance war at the same time Intel is. Competition makes the resulting products better. Do you think the 9th Gen Intel chips would be octocore without Ryzen?
评论 #18393943 未加载
评论 #18393682 未加载
评论 #18394396 未加载
评论 #18397260 未加载
评论 #18395714 未加载
评论 #18393604 未加载
评论 #18395433 未加载
评论 #18394247 未加载
评论 #18394097 未加载
评论 #18396315 未加载
tambourine_manover 6 years ago
What happened to Intel? Have they given any official explanations yet?<p>How does a company, who for decades has been 1-2 node processes ahead of the industry, suddenly gets 1-2 node processes <i>behind</i> everyone else.<p>There must be an amazing story behind this that no one seems to be digging into.
评论 #18393940 未加载
评论 #18393547 未加载
评论 #18393823 未加载
评论 #18395237 未加载
评论 #18394639 未加载
评论 #18393818 未加载
评论 #18394117 未加载
at1asover 6 years ago
One slide noted that their 7nm processor had 13.28 billion transistors per 331 mm^2.<p>Although 7nm&#x2F;10nm have mostly become marketing terms, is this something directly comparable between fabs&#x2F;companies?<p>Does anyone have a comparison for Intel? Best I can find is this list: <a href="https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Transistor_count" rel="nofollow">https:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Transistor_count</a> (and if that&#x27;s accurate and up to date, it looks as though AMD will be well ahead).
评论 #18393923 未加载
评论 #18393926 未加载
评论 #18393859 未加载
评论 #18394171 未加载
51lverover 6 years ago
This is really fun to watch. AMD is giving people EXACTLY what they want (again), and intel is having to fight dirty (again).<p>Possible example (not at all out of character for intel): why are so many people parroting that 10nm technical superiority junk without supplying sources?
评论 #18396025 未加载
评论 #18394984 未加载
stagger87over 6 years ago
Increasing the vector width to 256 bits (assuming no crazy thermal throttling) is a pretty big deal and would get me to move off Intel, unless Intel can figure out 512 bit widths without massive throttling.
评论 #18399737 未加载
arnonover 6 years ago
This AMD card can compete with NVIDIA&#x27;s high end Tesla V100 accelerator.<p>At 7.4 TFlops of double-precision, it is smack in the middle between the PCIe version of the V100 at 7.0 and the NVLink version at 7.8.<p>Memory bandwidth for the MI60 is a bit better at 1000GB&#x2F;s, compared to the Tesla V100&#x27;s 900GB&#x2F;s.<p>However, AMD&#x27;s problems are usually not the actual hardware, but the software around it. NVIDIA has done amazing work with CUDA and the surrounding frameworks, while AMD has not really. They really need to catch up on software that makes writing code for their GPUs more trivial.
utternerdover 6 years ago
64 core EPYC chips based on Zen 2 is what really blows my mind.
评论 #18394401 未加载
kevin_thibedeauover 6 years ago
Fictional process geometries are the new MHz wars.
评论 #18393971 未加载
tmd83over 6 years ago
Is anyone knowledgeable to comment about the memory bandwidth. I thought Zen-1 was eight channel with 32 core, now the Zen-2 is the same eight channel with 64 core. Wouldn&#x27;t that cause issue or can the new memory system be that better?<p>The other interesting thing is that they said the memory access would be more uniform kind of like NUMA independent given that the controller is no longer part of the individual chip but a common element. Which definitely makes good performance easier with such a beast of cheap but does it do so at the cost of the lowest possible latency as in when in Zen-1 the memory access was from a channel in the same CPU. I would hope that a massive single piece of IO chip would allow them to design the thing better but does anyone know or care to guess?
评论 #18398865 未加载
lxcidover 6 years ago
I have a feeling this is not a AMD vs Intel or Apple vs Intel.<p>This is TSMC vs Intel. TSMC basically make the 7nm chip for Apple and AMD.<p>It looks like this company HQ at Taiwan provided the bragging right for Apple and AMD…
评论 #18398908 未加载
Symmetryover 6 years ago
Well, looks like Semiaccurate was right about Rome&#x27;s architecture. Personally I&#x27;m excited about the wider vector units.
sergiotapiaover 6 years ago
Video: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=GwX13bo0RDQ" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=GwX13bo0RDQ</a><p>Dr. Lisa Su is so inspiring when she speaks - very clear, sure of her product and confident. Hopefully I get there one day!
doctorsherover 6 years ago
Non-AMP link: <a href="https:&#x2F;&#x2F;www.tomshardware.com&#x2F;news&#x2F;amd-new-horizon-7nm-cpu,38029.html" rel="nofollow">https:&#x2F;&#x2F;www.tomshardware.com&#x2F;news&#x2F;amd-new-horizon-7nm-cpu,38...</a>
评论 #18393441 未加载
评论 #18393430 未加载
airstrikeover 6 years ago
nm is the new megapixel
评论 #18394587 未加载