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

[MajinCry]

Bah, we're relying on a bloody strategy game fer draw calls, and the results are all o'er the place. I say we get a bunch o' computers, chuck Fallout 4 on 'em @ ultra details + ENB for draw call measurements, then high tail it tae the top of the Corvega factory.

Give us a proper gameplay scenario 'n' performance comparison, dangnammit. I want ta know if AMD have finally gotten over there draw call deficit, where intel's equivalent CPUs in performance (e.g, i7 920 == 965 BE) will get 3-4x higher framerates, when the draw calls are the decidin' factor.

 

[TheELF]

Does anyone has a pic of the ZEN AoTS benchmark run used in the OP ??

Just run the benchmark with my Core i7 3770K @ 4.44GHz + 16GB ram at 2154MHz + R9 390 default with Crimson 16.7.3 Win 10 64bit

Wcc has all of them
http://wccftech.com/amd-zen-es-benchmarks/

 

[cytg111]

I searched but found nothing about java benchmark. Reason why java run slow is not just because of poorly threaded, java is already obsoleted and there's very few optimization on it. I think there's no need to consider a new rig only for java. IF you need, i3 or A10 which close to 4Ghz clock might help. As for Zen, it might not widely available before January.
Don't spend more money on inappropriate hardware.

Java is not obsolete, infact it is dominating every segment that is not to-the-metal, lookup the scores. There is a reason that these exits ; JRuby, Jython, Scala etc. there is a life for most languages in the JVM.. And with java8 and lambdas it wont be long before you see JMono i bet.
If you are not coding c/c++ you are coding java.. or less

Are these posts a joke?
Why would any of you ever think that AMD was going to be the competitive gaming chip? Its been what? 10? 11? years since that was the case. You didn't honestly believe Jim Keller alone could come in and make a gaming chip that would lay low the mighty Intel with its R&D budget that's what, 900% more than AMD's?
I guess hope really does spring eternal for AMD fans.
The only saving grace this chip can claim is it might be a contender for mid-range workstations. A $300 price point with 75-80% of the i7-6900K performance would be a huge win for AMD.

I can make a few guesstimations that is enough to make me less certain about the absolute truth as you describe.
Intel has been going heavy for mobile. Intel has been going strong on atom for mobile. How much of the R&D behind recent process nodes have been geared towards those two goals?(mobile atom is now dropped)
Goals that have been set AFTER the first incarnation of Core!
Intel may have a big R&D budget but they may also have a process tech that is pointing one way(mobile) and a selling segment that points another.. I dont know, I wouldnt count the competition out just yet.. seems premature.

 

[aigomorla]

Speaking of it. How to get my forum navigation sitebar back?

You can ask any forum related question here:
https://forums.anandtech.com/forums/technical-forum-issues.17/


This is the LAST warning i am giving anyone after that post.
You drift off TOPIC, you will get a infraction. That means anything posted now should have something to do with Zen, or it will be considered OFF TOPIC.

Moderator Aigo

 

[dark zero]

I saw that Zen can scale to 5 watt... There won't be a 5- watt variant to mobile?

 

[AtenRa]

Wcc has all of them
http://wccftech.com/amd-zen-es-benchmarks/

thanks

 

[AtenRa]

OK this is with Standard preset

 

[blublub]

I have found some information, regarding older AMD ES, that the Turbo boost is often disabled.
Taking this into account wouldn't performance of the 2 CPUs just leaked be very good? - many people seemed to estimate IPC via Turbo clocks

 

[The Stilt]

I have found some information, regarding older AMD ES, that the Turbo boost is often disabled.

Few examples (SKUs) on products which had turbo available on the final product but not on the ES?

 

[blublub]

Few examples (SKUs) on products which had turbo available on the final product but not on the ES?

Nope, no SKUs sorry. I came across this site where it was mentioned:

http://www.pcgameshardware.de/AMD-Z...ring-Samples-Revision-Basistaktraten-1191370/

and searched a little and a few others sites also mentioned something similar - I choice of words wasn't the best I should have better written: "have found some information, regarding older AMD ES, that the Turbo boost could be disabled."

 

[Abwx]

OK this is with Standard preset

Could you do the test at same CPU frequency but with the RAM speed at say 1600MHz..?.

 

[AtenRa]

Could you do the test at same CPU frequency but with the RAM speed at say 1600MHz..?.

Everything the same except RAM at 1614MHz

 

[Abwx]

Everything the same except RAM at 1614MHz

Thank you.

So 33% higher RAM speed increased the average framerate by 8% but the CPU framerate did increase by 15%, so this latter estimation by Aots is highly RAM speed dependant, this correlate with the point made by a member a few pages earlier...

 

[blublub]

So if the ZEN was using slow RAM or settings the result is crippled

 

[Abwx]

So if the ZEN was using slow RAM or settings the result is crippled

It is crippled mainly for the CPU framerate numbers but for the actual bench it lose somewhat less, given that it s working likely at 2.8GHz it should largely match the 5960X despite this latter s 4 channels.

Also the main concern for ES delivered to OEMs is stability, so they are often slightly overvolted in respect of the defnitive specs, hence the 2.8 base frequency.

 

[Sweepr]

It has to match a (stock) 4C/8T Haswell in this benchmark first before you can dream about 5960X performance.

- Core i7-4770 4C/8T 3.4-3.9 GHz (2013 Haswell)
Average: 66.0 FPS
Normal batch: 74.5 FPS
Medium batch: 69.5 FPS
Heavy batch: 56.6 FPS

- Summit Ridge 8C/16T ES - 2.8-3.2 GHz (2016/2017 Zen)
Average: 58.9 FPS
Normal batch: 65.8 FPS
Medium batch: 62.8 FPS
Heavy batch: 50.5 FPS

 

[Sweepr]

Edit.

 

[Abwx]

It has to match a (stock) 4C/8T Haswell in this benchmark first before you can dream about 5960X performance.

As usual no source for the scores in your post, where are the settings..?.

Neverless here Guru 3D submission of their 3.2GHz 5960X, FPS are no better than the ones of the 4770, besides Aots doesnt record the actual frequency so we can assume that anything better than Guru 3D is an ocked chip.

http://www.ashesofthesingularity.co...-details/5d5ab818-bc9e-4340-8a33-9e243a097a1f

And Zen :

http://cdn.wccftech.com/wp-content/uploads/2016/08/standard-1080p-zen-1d.png

So with 14% higher clock and undoubtly vastly bigger RAM bandwith the 140W 5960X does 15-20% better than the 95W RAM bandwith gimped Zen, if anything that look quite promising, at least for some people, for others it may well be a big concern instead....

 

[blublub]

As usual no source for the scores in your post, where are the settings..?.

Neverless here Guru 3D submission of their 3.2GHz 5960X, FPS are no better than the ones of the 4770, besides Aots doesnt record the actual frequency so we can assume that anything better than Guru 3D is an ocked chip.

http://www.ashesofthesingularity.co...-details/5d5ab818-bc9e-4340-8a33-9e243a097a1f

And Zen :

http://cdn.wccftech.com/wp-content/uploads/2016/08/standard-1080p-zen-1d.png

So with 14% higher clock and undoubtly vastly bigger RAM bandwith the 140W 5960X does 15-20% better than the 95W RAM bandwith gimped Zen, if anything that look quite promising, at least for some people, for others it may well be a big concern instead....

If ZEN does gain 10% clock until production it looks to be pretty good performance/watt at least for the 8c model even though it will not be the performance king in terms of raw performance.
Since everything is probably better than current CPUs it will sell if the price is ok..an maybe it will kick Intel's butt just a little so that prices can drop

 

[Sweepr]

As usual no source for the scores in your post, where are the settings..?.

Sources in the OP. Alternatively look at the benchmarks above, where a slightly OCed 3770K with gimped DDR3 speeds is faster than Summit Ridge with DDR4 and twice the cores (4C/8T vs 8C/16T).

So with 14% higher clock and undoubtly vastly bigger RAM bandwith the 140W 5960X does 15-20% better than the 95W RAM bandwith gimped Zen, if anything that look quite promising, at least for some people, for others it may well be a big concern instead....

Stock i7 4770 from 2013 is faster in this benchmark, and yes, it scales past 4C/8T (1, 2 - 30% gains with 8C/16T vs 4C/8T). Don't try to spin it, it might or not be representative of final performance (overall) but the results in AotS are simply awful.

for others it may well be a big concern instead....

Maybe if I was claiming 5960X performance at 95W it would.

 

[Abwx]

Sources in the OP. Alternatively look at the benchmarks above, where a slightly OCed 3770K with gimped DDR3 speeds is faster than Summit Ridge with DDR4 and twice the cores (4C/8T vs 8C/16T).

Slightly overclocked at 4.44GHz, lol...

Stock i7 4770 from 2013 is faster in this benchmark, and yes, it scales past 4C/8T (1, 2 - 30% gains with 8C/16T vs 4C/8T). Don't try to spin it, it might or not be representative of final performance (overall) but the results in AotS are simply awful.

Maybe if I was claiming 5960X performance at 95W it would.

What is awfull is the way you are trying to blur the numbers, i noticed that you keep using the CPU framerate in your deliberatly misleading OP, what the card render is the Framerate and this is the only real bench made by Aots, the very one you are trying to obfuscate..

Beside Guru3D 5960X is at a fixed 3.2GHz, so above its TDP rating, so much for your wishfull estimation in this respect...

 

[Abwx]

edit

 

[Glo.]

I think all of discussing people here are drawing too many conclusions too early.

You do not know if this was A0 version of Zen, which this analysis points to: http://www.bitsandchips.it/english/...speculations-about-zen-after-our-april-s-fool

3050 MHz points to exact mentioned 3GHz clock frequency that Fottemberg has written. Now its all about the quality of silicon.

Realistically, I was expecting 3.5 GHz base clock, 4 GHz turbo. If it will be higher it might be even better.

 

[Arachnotronic]

I think all of discussing people here are drawing too many conclusions too early.

You do not know if this was A0 version of Zen, which this analysis points to: http://www.bitsandchips.it/english/...speculations-about-zen-after-our-april-s-fool

3050 MHz points to exact mentioned 3GHz clock frequency that Fottemberg has written. Now its all about the quality of silicon.

Realistically, I was expecting 3.5 GHz base clock, 4 GHz turbo. If it will be higher it might be even better.

6900K from Intel runs at 3.2ghz base and 3.7ghz in a 140w tdp and has far less integrated on the die manufactured on a superior process. Your expectations seem very unreasonable IMO.

 

[jpiniero]

Realistically, I was expecting 3.5 GHz base clock, 4 GHz turbo. If it will be higher it might be even better.

In case it wasn't obvious, AMD clearly missed their power consumption target for Polaris; but unlike the GPUs they can't just raise the TDP since they presumably committed that they would release the 8C Zen at 95W. So down the clocks go.

If the leak did indeed come from someone from a mobo vendor or OEM I imagine we will see more leaks soon.