ZEN ES Benchmark from french hardware Magazine

Page 26 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

raghu78

Diamond Member
Aug 23, 2012
4,093
1,475
136
Perhaps I've missed something, but there doesn't seem to be any indication that Ryzen actually matches BW-E IPC on avg. Rather is slightly behind.. Enough to be considered more like Haswell in throughput, or slightly lower.. This is based on the Canard tests which show a 6900K at a 14.5% advantage, (including blender) and which, by all accounts has a <10% clock advantage. So perhaps people shouldn't be getting there knickers in a twist 're who's predictions were right or wrong quite yet? Moreover, the 40% is an ST claim, and none of the benchmarks so far have been run 1T per core. So until someone does this, no one knows.

By my own testing of Excavator in 1T/module vs Skylake i3 1T/core , I stiill avg around 60% to SKL no matter what I add or remove from the bench set to try and get a realistic avg (which is difficult due to the wildly different architectures), that puts Skylake approx 15% > A theortical Zen. Which in turn plants it pretty firmly in between ivy and Haswell.

A bit more than The Stilt's original prediction, and still sounds low to some, but personally I still predict Ryzen will have greater SMT Yeild (%) than Broadwell, which means of course that throughput can be closer to Haswell/BW than it's ST performance is.

As for why I think this will be the case:- Distributed Int Schedulers, and a higher number of statically partitioned structures are the main reasons. Intel's more dynamic approach here should result in higher ST performance, but IMO is a mistake in regards to throughput perf/watt. if not perf/mm , It makes sense to me that AMD chose to avoid such a large unified scheduler.. This is after all an architecture designed to be as balanced as possible, But given it's 2017 now (Is here in Australia!) likely biased towards throughput wherever it didn't impact ST significantly,


rZ734LS.png

yeah you sure have missed the fact that Canard PC had a buggy A0 revision ES. They acknowledge their sample had SMT and uop cache issues. If anybody is deriving any conclusion from canard PC benchmarks its bound to be wrong. I would say the final reviews of the production chip with all bugs fixed, proper turbo and firmware optimizations is what we should wait for. AMD has already demonstrated Blender and Handbrake with similar and better performance respectively wrt 6900k. If you do not want to believe AMD controlled benchmarks thats perfectly fine. But then jumping to conclusions based on a buggy ES sample is not the right approach. You would be right to disregard all benchmarks till launch but drawing your conclusion based on a buggy ES benchmark from canardpc while ignoring more recent benchmarks done by AMD based on recent steppings points to an agenda of your own.
 
  • Like
Reactions: Doom2pro

DrMrLordX

Lifer
Apr 27, 2000
21,582
10,785
136
CPU cores are fed from one plane and all of the other high power parts from the SoC plane. Carrizo / Bristol Ridge requires three planes by the design, however on AM4 BR parts both of the SoC & GFX planes are fed from the same plane as AM4 has only two major power planes. This required some software configuration changes to BR and hurts the overall efficiency slightly. Things are not as simple with Zen based parts ;)

The phase configuration obviously cannot be changed after the design has been done. The specific controller just allows more flexible configuration (i.e same controller can be used with e.g 6+1 and 1+6 phase boards). The Gigabyte board uses 4+3 configuration.

Oh okay, I misinterpreted what you said. So it doesn't switch dynamically.

I'd still like to know what's the purpose of that extra choke though.
 

cytg111

Lifer
Mar 17, 2008
23,047
12,715
136
Perhaps I've missed something, but there doesn't seem to be any indication that Ryzen actually
matches BW-E IPC on avg. Rather is slightly behind..

- Yup, just to reiterate my earlier prediction (now that we are all throwing our money to the center of the table), I am going with ~5% behind BW IPC. We should keep a record of sorts so we can pick a winner - after the fact. :)
 

lolfail9001

Golden Member
Sep 9, 2016
1,056
353
96
They acknowledge their sample had SMT and uop cache issues.
They also hint repeatedly that they sacced stability for having both of these enabled, while noting that changes to get rid of the bug lead to Sandy Bridge level of IPC, not what was shown.
I would say the final reviews of the production chip with all bugs fixed, proper turbo and firmware optimizations is what we should wait for.
I just hope they do come before end of February.
 

The Stilt

Golden Member
Dec 5, 2015
1,709
3,057
106
I'd still like to know what's the purpose of that extra choke though.

It's for some of the minor voltage rails in the CPU with TDC of < few amps. Something which draws more current than is efficient to generate using a LDO.
 

raghu78

Diamond Member
Aug 23, 2012
4,093
1,475
136
They also hint repeatedly that they sacced stability for having both of these enabled, while noting that changes to get rid of the bug lead to Sandy Bridge level of IPC, not what was shown.

I just hope they do come before end of February.

The nature of a bug is it generally leads to erroneous results and reduced performance. So why do you want to arrive at any conclusions based on canard PC's buggy sample. Digitimes is saying late Feb launch and March mass shipments. So I think your wait is going to be roughly 2+ months.
 

TheELF

Diamond Member
Dec 22, 2012
3,967
720
126
The nature of a bug is it generally leads to erroneous results and reduced performance. So why do you want to arrive at any conclusions based on canard PC's buggy sample. Digitimes is saying late Feb launch and March mass shipments. So I think your wait is going to be roughly 2+ months.
Because they commented on the bug,ran some MT benches,including blender and handbrake,so if they would have found way less performance on those two benches than what AMD showed they would have commented on that as well,since they didn't comment on it we have to assume that blender and handbrake just run very well on the zen arch (which would be why AMD choose them) and that the rest of the MT benches ran much worse.
 

raghu78

Diamond Member
Aug 23, 2012
4,093
1,475
136
Because they commented on the bug,ran some MT benches,including blender and handbrake,so if they would have found way less performance on those two benches than what AMD showed they would have commented on that as well,since they didn't comment on it we have to assume that blender and handbrake just run very well on the zen arch (which would be why AMD choose them) and that the rest of the MT benches ran much worse.

You are again making assumptions on what happened. Canard PC did not confirm that their sample performed as well as AMD's chip demoed at New Horizon even after accounting for lower clocks. Canard PC give no assurances that this is the performance we can expect from a production chip.
 
  • Like
Reactions: Doom2pro

Doom2pro

Senior member
Apr 2, 2016
587
619
106
It's for some of the minor voltage rails in the CPU with TDC of < few amps. Something which draws more current than is efficient to generate using a LDO.

It could also be for some various other silicon on the motherboard as well... I know AM4 is light on chipsets, but there could be other supportive hardware like ASMedia USB3 buffers and such.
 

TheELF

Diamond Member
Dec 22, 2012
3,967
720
126
You are again making assumptions on what happened. Canard PC did not confirm that their sample performed as well as AMD's chip demoed at New Horizon even after accounting for lower clocks. Canard PC give no assurances that this is the performance we can expect from a production chip.
You are making assumptions as well when you are saying that the bug lowered performance,you just don't know,I don't know,nobody knows.Canard PC did not confirm that their sample performed as well as AMD's chip demoed at New Horizon,but neither did they refute it.
In general reporters will comment on the weird,the out of usual,like the bug and not on the normal stuff unless there is some controversy.
 

majord

Senior member
Jul 26, 2015
433
523
136
yeah you sure have missed the fact that Canard PC had a buggy A0 revision ES. They acknowledge their sample had SMT and uop cache issues. If anybody is deriving any conclusion from canard PC benchmarks its bound to be wrong. I would say the final reviews of the production chip with all bugs fixed, proper turbo and firmware optimizations is what we should wait for. AMD has already demonstrated Blender and Handbrake with similar and better performance respectively wrt 6900k. If you do not want to believe AMD controlled benchmarks thats perfectly fine. But then jumping to conclusions based on a buggy ES sample is not the right approach. You would be right to disregard all benchmarks till launch but drawing your conclusion based on a buggy ES benchmark from canardpc while ignoring more recent benchmarks done by AMD based on recent steppings points to an agenda of your own.


Well no I didn't miss the business about the steppings, I also didn't reach any conclusions, and lastly, my only agenda is the truth.

Chasing extra performance in later steppings usually ends up in disappointment, and disappointment leads to irrational backlash against the company people have been drawn into expecting miracles from. So bear in mind, if I'm right, don't come in here all disappointed.

I'm not discounting the possibility of Zen matching/beating Broadwell in throughput/clk (on avg across the board), and your'e sadly mistaken if you think I would be disappointed if it exceeded expectations in this way, but there's just not enough evidence of this. It seems to me no matter how good the leaks are, people still want to push the envelope further.
 
  • Like
Reactions: coercitiv
Aug 11, 2008
10,451
642
126
yeah you sure have missed the fact that Canard PC had a buggy A0 revision ES. They acknowledge their sample had SMT and uop cache issues. If anybody is deriving any conclusion from canard PC benchmarks its bound to be wrong. I would say the final reviews of the production chip with all bugs fixed, proper turbo and firmware optimizations is what we should wait for. AMD has already demonstrated Blender and Handbrake with similar and better performance respectively wrt 6900k. If you do not want to believe AMD controlled benchmarks thats perfectly fine. But then jumping to conclusions based on a buggy ES sample is not the right approach. You would be right to disregard all benchmarks till launch but drawing your conclusion based on a buggy ES benchmark from canardpc while ignoring more recent benchmarks done by AMD based on recent steppings points to an agenda of your own.
So basically you are saying we should accept 2 very likely cherry picked benchmarks from the manufacturer while ignoring a wider suite of benchmarks from an independent source. OK, whatever you say.
 

zinfamous

No Lifer
Jul 12, 2006
110,511
29,096
146
Funny how a website that only peddles clearly faked rumors also is the only one that shows favorable performance for a company known to deceptively market their products using hype.

lol. who hired you, my man? You are now 2-for-2 on posting pure conspiracy trollop and retort-via-source discrediting. I'm not sure this forum can withstand so many fallacies.

But again: kudos on the irony-laden avatar. Sweepr will be along at any moment to make you 2-for-2 on likes, though. Don't you worry!

But Guys, try to keep this on Zen.
 

Tup3x

Senior member
Dec 31, 2016
944
925
136
Chasing extra performance in later steppings usually ends up in disappointment, and disappointment leads to irrational backlash against the company people have been drawn into expecting miracles from. So bear in mind, if I'm right, don't come in here all disappointed.
It's not about chashing extra performance but fixing bugs that are hindering the performance and causing stability issues. The final chip is not going to perform worse.
 

KTE

Senior member
May 26, 2016
478
130
76
It's not about chashing extra performance but fixing bugs that are hindering the performance and causing stability issues. The final chip is not going to perform worse.
It's unknown to perform better, too.

Sent from HTC 10
(Opinions are own)
 

SlowSpyder

Lifer
Jan 12, 2005
17,305
1,001
126
Funny how a website that only peddles clearly faked rumors also is the only one that shows favorable performance for a company known to deceptively market their products using hype.

How does a company deceptively market their products using hype? If anything I think if AMD over-hypes Zen it'll lead to disappointment at launch if the real product doesn't live up to the hype.
 

KTE

Senior member
May 26, 2016
478
130
76
Correct me if I'm wrong but there is precedence of bug fixing resulting in lower performance, isn't it? So it could go any way in the end.
Yep, if it's patched as a quick fix.

Sent from HTC 10
(Opinions are own)
 
  • Like
Reactions: prtskg

Abwx

Lifer
Apr 2, 2011
10,847
3,297
136
Correct me if I'm wrong but there is precedence of bug fixing resulting in lower performance, isn't it? So it could go any way in the end.

CPC stated that their measurement should be taken as a "a minima" case, wich mean minimal values to expect from the chip, so it cant go in any other way than an eventual improvement...