AMD Ryzen (Summit Ridge) Benchmarks Thread (use new thread)

[LTC8K6]

My guess is that the SR3 - or the lowest SKU, whatever that is - will be a 4C 4T part. Possibly no L3 cache.

Well, if you and that price diagram are correct, that lowest SKU would cost around $220.

 

[raghu78]

FWIW

http://wccftech.com/amd-zen-summit-ridge-17th-january-launch-rumor/

Zen SR7 at USD 250-300 . overclocks to 4.2 Ghz.

 

[Phynaz]

Ban for linking to wccf

 

[cytg111]

performance on par with 6850k? 6/12 vs 8/16? hmmm. can still mean anything.

 

[swilli89]

FWIW

http://wccftech.com/amd-zen-summit-ridge-17th-january-launch-rumor/

Zen SR7 at USD 250-300 . overclocks to 4.2 Ghz.

I'd take the performance estimates with a grain of salt, but the pricing looks to be accurate. $300+ for locked multiplier 8C and $500+ for 8C/16T fully unlocked, an amazing deal if its near Haswell IPC.

 

[frozentundra123456]

Actually the performance is about what I expected-- 8 core zen competing with 6 core Intel. And don't forget the 6800k is only 400.00 and offers essentially the same performance as the 6850k. So sub 300 dollar pricing seems too low, but it 350ish seems realistic if they want to undercut Intel and gain back market share. I also don't think a January launch is realistic, unless they are just talking paper launch, with real availability in march

 

[deasd]

FWIW

http://wccftech.com/amd-zen-summit-ridge-17th-january-launch-rumor/

Zen SR7 at USD 250-300 . overclocks to 4.2 Ghz.

Again WCCF don't know what they were talking about.....there's no exact price range and performance mentioned in that chiphell chinese leak. Also 4.2Ghz turbo is somewhat a guesstimation either.

 

[Fanatical Meat]

Actually the performance is about what I expected-- 8 core zen competing with 6 core Intel. And don't forget the 6800k is only 400.00 and offers essentially the same performance as the 6850k. So sub 300 dollar pricing seems too low, but it 350ish seems realistic if they want to undercut Intel and gain back market share. I also don't think a January launch is realistic, unless they are just talking paper launch, with real availability in march

This is more or less what I expected in my amateur opinion. I'll likely sit this round out just because I still feel burned by AMD however that could change.

 

[The Stilt]

Also 4.2Ghz turbo is somewhat a guesstimation either.

They base those statements on the fanfiction "leak" from South Africa, where 14nm LPP handles 1.5V voltages and houses are made of gingerbreads.

Moving onward, AMD has optimized the clock speeds for Summit Ridge chips around 3.15-3.30 GHz base and 3.5 GHz boost on FinFET process. The rumors currently mention that Summit Ridge can overclock to 4.2 GHz with ease and voltage can go as high as 1.5V with stability. It’s also mentioned that the chip can only break the 5 GHz mark with LN2 cooling. The clock speeds are based off a rumor from a different source but Maxsun did quote it in their official statement so it might hold some legitimacy.

 

[Arachnotronic]

They base those statements on the fanfiction "leak" from South Africa, where 14nm LPP handles 1.5V voltages and houses are made of gingerbreads.

LOL!

 

[cytg111]

.. an amazing deal if its near Haswell IPC.

If it needs 8c/16t to keep up(@ rumored clocks) with a 6850k then IPC will be well below haswell.

 

[KTE]

If it needs 8c/16t to keep up(@ rumored clocks) with a 6850k then IPC will be well below haswell.

ST 20-25% below Broadwell-E is what I'm expecting.

Sent from HTC 10
(Opinions are own)

 

[cytg111]

ST 20-25% below Broadwell-E is what I'm expecting.

Sent from HTC 10
(Opinions are own)

Count me in at 20%. Hoping for more but circumstances by large seems to indicate otherwise (leaks, #cores, targeted segment and price .. but it would still fit in their +40% IPC promise)

 

[AtenRa]

ST 20-25% below Broadwell-E is what I'm expecting.

Sent from HTC 10
(Opinions are own)

Count me in at 20%. Hoping for more but circumstances by large seems to indicate otherwise (leaks, #cores, targeted segment and price .. but it would still fit in their +40% IPC promise)

If ST is 20-25% bellow Broadwell-E then we are talking about 2011 SandyBridge ST performance while having a 50% wider Core. It will be a complete failure unless, perf/watt is way way higher and closer to Skylake.

Since ZEN development started after IvyBridge was released, i strongly hope the IPC and ST performance target was at least Ivy-Haswell.

 

[TheELF]

Since ZEN development started after IvyBridge was released, i strongly hope the IPC and ST performance target was at least Ivy-Haswell.

Why?You think AMD got hold of intel's plans/blueprints/whatever their called for ivy?

 

[deasd]

performance on par with 6850k? 6/12 vs 8/16? hmmm. can still mean anything.

If it needs 8c/16t to keep up(@ rumored clocks) with a 6850k

Please tell me where did you read this. WCCF??

 

[cytg111]

Please tell me where did you read this. WCCF??

Yes. Hence the *if* .

We are all just rolling with unsubstantiated rumors here, extrapolating what it may mean if they come true or not.

 

[Glo.]

The same fu**** site spread false information, prerelease, about Polaris clocking at 1650 MHz, and being faster than GTX 1070.

Leave the rumors from this site where they should be - garbage.

 

[KTE]

Why?You think AMD got hold of intel's plans/blueprints/whatever their called for ivy?

They usually have a good idea, yup

Sent from HTC 10
(Opinions are own)

 

[AtenRa]

[rumor] January 17th release date for ZEN ??

http://www.guru3d.com/news-story/amd-zen-8-core-summit-ridge-to-launch-january-17th-2017.html

 

[majord]

for Consistency I think we should really stick to IPC "above" not below. "i.e Broadwell E will be xx% higher IPC" and yes it makes a difference. 20% lower, means 25% higher

I'd put my bets on a 10-15% advantage for Broadwell E.. on average - mixed workloads. but it's going vary enough that anyone could be 'right'. The arguments that will ensue as people try to defend their predictions will be pretty painful to watch. there'll be cherry picked results from all corners. . Without a 'standard' set of benchmarks to draw an average from, it's not going to be easy.

but alas, if we're talking an avg over a large, reasonably varied benchmark suite, anything less than that falls outside AMD's claims for one, and would also be a bit disapointing given what we know about the core.

 

[deasd]

Yes. Hence the *if* .

We are all just rolling with unsubstantiated rumors here, extrapolating what it may mean if they come true or not.

OK, let's assume it's correct, on par with i7-6850K in Cinebench R15.

As we see, I'm afraid the 'on par with 6850k' statement is not enough to justify anything, because there's very limited difference between 5960X(8C) and 6850k(6C), not to mention we don't know the exact frequency of Zen.

BTW, it's about twice as fast as FX8350 which clocked at 4Ghz if true.

edit: I dislike Cinebench, anyway.

 

[nismotigerwvu]

[rumor] January 17th release date for ZEN ??

http://www.guru3d.com/news-story/amd-zen-8-core-summit-ridge-to-launch-january-17th-2017.html

Maybe it's a translation quirk, I could see a January unveiling/paper launch but a full launch that soon sounds a bit far fetched.

 

[moonbogg]

That rumor sais it can OC to 4.2ghz fairly easily with conventional cooling. That doesn't explain much, but at least it looks like it won't get bogged down below 4ghz like many thought it would. Even if the per core performance at 4.2 matched my 3930k@4.6, to be honest that's still pretty fast actually, especially with 8 cores. Well, its fast for a 5 year old CPU anyway, but its still a good deal at $300. That price means everything.

 

[nathanddrews]

Exactly, price is going to be the biggest factor. If I have a chance of getting 8 speedy cores for around $300 along with all the modern I/O bells and whistles, that's a good deal.