• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

[Motley Fool] Intel should consider going fabless

Page 3 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
IGP likely only a big issue vs NV for discrete card gaming laptops. They should have surely have Intel covered OK?

Probably. Intel has made up a fair bit of ground over the last few years, but that was while AMD's parts were hamstrung by their weak CPU cores. Zen APUs should put a fair bit of distance ahead of their Intel rivals, albeit with the latter now being just about viable (though certainly not optimal) for casual gaming.
 
Probably. Intel has made up a fair bit of ground over the last few years, but that was while AMD's parts were hamstrung by their weak CPU cores. Zen APUs should put a fair bit of distance ahead of their Intel rivals, albeit with the latter now being just about viable (though certainly not optimal) for casual gaming.

I agree, AMD should be easily ahead for casual gaming vs Intel iGPU.

As was show is that sweet spot for Ryzen performance/watt is at lower clocks, around 2-3 ghz. There it's actually better than intel! (well after all GF 14nm is a low power process). I think the bigger issue here could actually be Vega iGPU taking Vega FE as indication.

Is there anything definitive on this. So much of the testing is different core/thread counts, different clock speeds that it really blurs the picture, so it is hard to get a handle on real perf/watt.

But if you compare actual work, of identical core/thread counts, Intel does Better on perf/watt (with the possible of exception of the their HEDT quad memory channel chips, which seem to have issues).

Here is two benchmarks that include power usage (including one of AMDs favorite benches Cinebench). Compare the Ryzen 1500X to the 7700K. Both 4C/8T, with Intel running higher clocks, so no surprise it is faster:
https://www.techspot.com/review/1379-and-ryzen-5-1600x-1500x/page3.html
CineBench:
7700K 941 points 137 watts
1500X 797 points 154 watts

Excel:
7700K 3.26 seconds 133 watts
1500X 4.06 seconds 148 watts

But Intel also using less power while running at higher clocks and delivering more performance,
in both Cinebench and Excel benchmarks.

Intel performs better and uses less power. Will lower clock speed reverse that? I'd like to see a demonstration.
 
Last edited:
Here is one of AMDs favorite benches Cinebench: Compare the Ryzen 1500X to the 7700K. Both 4C/8T, with Intel running higher clocks, so no surprise it is faster:
https://techspot-static-xjzaqowzxaoif5.stackpathdns.com/articles-info/1379/bench/Cinebench.png
But it also using less power while running at higher clocks and delivering more performance:
https://techspot-static-xjzaqowzxaoif5.stackpathdns.com/articles-info/1379/bench/PowerCinebench.png

Same thing happens in their Excel benchmarks. Intel performs better and uses less power.

Yeah put as I wrote Ryzen is clocked way beyond the sweet spot of the process which is below 3 ghz. TheStilt in Ryzen technical thread tested this. Performance/watt is a lot higher at lower clocks, say 2.5ghz because power use (voltage) is much lower. Thats why epyc 32 core can be clocked at 2.7 ghz all core.
 
Yeah put as I wrote Ryzen is clocked way beyond the sweet spot of the process which is below 3 ghz. TheStilt in Ryzen technical thread tested this. Performance/watt is a lot higher at lower clocks, say 2.5ghz because power use (voltage) is much lower. Thats why epyc 32 core can be clocked at 2.7 ghz all core.

And you don't think Intels CPUs will also be much more power efficient running significantly under-clocked and under-volted? People have been doing this for ages to run passive setups with Intel chips. It works for them too.

I have a hard time believing the above example would reverse itself, until I actually see a demonstration.

Edit: Anyway we are getting way off topic. If we want to talk Raven Ridge, there is a thread for that.

I still haven't seen a compelling argument that Intel should go Fabless. The original story is more clickbait than anything serious IMO.
 
Last edited:
Back
Top