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

Page 209 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.
Status
Not open for further replies.

OrangeKhrush

Senior member
Feb 11, 2017
220
343
96
Assuming linear scaling Ryzen per single thread (passmark) is 46% faster

8370 (4.0/4.3) = 1622*3.4/4.3 = 1283
Ryzen (3.4) = 2046


My HTPC w/Out Turbo (3.2) = 1920
My HTPC (3.2/3.4) = 2026

1920*3.4/3.2 = ~2040 (linear scale)


 

PhonakV30

Senior member
Oct 26, 2009
979
351
136
Well acording to passmark


If its running at 3.4Ghz its exactly Haswell IPC, at 3.8Ghz it may be below Haswell IPC, personally i think its running at 3.4Ghz, the table says "Turbo Clock: N/A" like with the I3s, it may also be because the program is unable to detect it,
Sandy above Haswell ? 2500K > 4130 ?
 

OrangeKhrush

Senior member
Feb 11, 2017
220
343
96
Passmark only checks frequency CPU is running at before running tests themselves. MSR reports 3.4Ghz and then it boosts up during tests. Bah-dum, it is entirely invisible.
Can you prove it? funny how the database shows no turbo, like my 4460 when I disabled turbo is the only CPU with Ryzen that has no turbo range listed.


Very strange right, all the other CPU's are running their turbos as the 4460 shows with two different scores that equal up more or less with scaling at clockspeed.
 

lolfail9001

Golden Member
Sep 9, 2016
1,056
353
96
Can you prove it?
That Passmark runs frequency check before tests? It tells you so straight up.
funny how the database shows no turbo, like my 4460 when I disabled turbo is the only CPU with Ryzen that has no turbo range listed.
So what? The entire point of argument is that Ryzen does not list turbo range. It does not mean a thing on Ryzen, hence the turbo state is "invisible". You simply cannot make conclusions on it's actual running clock from lack of reported turbo.
 

PhonakV30

Senior member
Oct 26, 2009
979
351
136
my 2500K is running at 4.2Ghz, that bench suite only shows base clocks on the bars.
oh yes.so Passmark doesn't show current clock during test.so either turbo was OFF or 2046 is with turbo.
If turbo (2046) => 3.8 Ryzen ~ 3.4 Haswell
If Off ( 2046) => 3.4 Ryzen ~ 3.4 Haswell
 

Shivansps

Diamond Member
Sep 11, 2013
3,634
1,314
136
oh yes.so Passmark doesn't show current clock during test.
It shows them on the table, when you add them, but it doesnt shows them on the bars, the I3 have no turbo, so the listed speed is correct. the Ryzen has no information for turbo like the I3, so it may be because the turbo is disable or the program desent read it.
 

Shivansps

Diamond Member
Sep 11, 2013
3,634
1,314
136
2500K at 4Ghz 2172

2172 * 4.2 / 4 = 2226 which is more or less what that score was
yes i was getting 2256 at 4.2Ghz, i dont see where all this disscusin comes from, its very clear to me, the bars only shows base clocks at the comparison, I3 have no turbo, so the listed speed is correct. The Ryzen turbo info is "N/A" so it may have turbo disabled to just be a bad reading, and my 2500K is running at 4.2Ghz turbo, bars shows the base 3.3ghz, and it scores similar to Haswell 3.8ghz, and it seems correct to me.
 

OrangeKhrush

Senior member
Feb 11, 2017
220
343
96
That Passmark runs frequency check before tests? It tells you so straight up.

So what? The entire point of argument is that Ryzen does not list turbo range. It does not mean a thing on Ryzen, hence the turbo state is "invisible". You simply cannot make conclusions on it's actual running clock from lack of reported turbo.
Why does it matter? Simple really we want coherant information not more gimping which is going on, why are people so scared of Ryzen's performance that every bench out has had some kind of impediment to it.

The difference is at 3.4 which I believe it was with turbo disabled either by firmware or in BIOS makes it Haswell or better, at 3.8 it s more like Ivybridge or a 3Ghz 5960X so little omissions can create false sentiments. I don't believe AMD would be so actively wanting to parade around Ivy level performance in 2017 on 14nm, that would be like slitting your wrists in front of everyone.
 

lolfail9001

Golden Member
Sep 9, 2016
1,056
353
96
why are people so scared of Ryzen's performance that every bench out has had some kind of impediment to it.
It performs well enough in circumstances as is, the truth here is that claiming it was running without turbo with literally 0 evidence is only making hype train ride too fast.
I don't believe AMD would be so actively wanting to parade around Ivy level performance in 2017 on 14nm, that would be like slitting your wrists in front of everyone.
Do i need to remind you how far did Intel go in that time?

Heck, i took 4.5Ghz 6900k. Single thread is 20% faster, but average of multithreaded tests (that use workloads from single thread test) is 30% faster. Here, your evidence to turbo being active.
 
  • Like
Reactions: Glo.

OrangeKhrush

Senior member
Feb 11, 2017
220
343
96
It performs well enough in circumstances as is, the truth here is that claiming it was running without turbo with literally 0 evidence is only making hype train ride too fast.

Do i need to remind you how far did Intel go in that time?

Heck, i took 4.5Ghz 6900k. Single thread is 20% faster, but average of multithreaded tests (that use workloads from single thread test) is 30% faster. Here, your evidence to turbo being active.
Intel has it's own fair share of problems but CPU's don't just evolve relative to the other. AMD's performance gradiant has zero relation to intel and there 14nm tick, tock, tock, tock. What are people afraid of? why is Intel viewed on a higher pedistal when Barrons released a very dire report on how overspent Intel's FABs is, how intel have squandered 20bn on sunken investments and how their CEO is the biggest moron around. If you want to see what a moron can do to a company look at Hector Ruiz.

Lisa Su spent all her time getting rid of non engineers in critical positions and replaced them with top engineers, if you think Ryzen was some slam jam throw together m'aam job, then nothing can help your cynicism and you are very well entitled to them. So far despite the amount of negitivity that is coming from the same group Ryzen continues to impress. I use Intel and Nvidia and that doesn't mean a rejuvinated AMD under a vibrant leader is not something I would love to see, because truth be told, my 4460 and 5960X will be just fine for now.

Nobody has ever claimed that Ryzen is going to smack Kaby lake at low clocks, and if they have it is there own perogative.
 

Glo.

Diamond Member
Apr 25, 2015
5,016
3,639
136
Intel has it's own fair share of problems but CPU's don't just evolve relative to the other. AMD's performance gradiant has zero relation to intel and there 14nm tick, tock, tock, tock. What are people afraid of? why is Intel viewed on a higher pedistal when Barrons released a very dire report on how overspent Intel's FABs is, how intel have squandered 20bn on sunken investments and how their CEO is the biggest moron around. If you want to see what a moron can do to a company look at Hector Ruiz.

Lisa Su spent all her time getting rid of non engineers in critical positions and replaced them with top engineers, if you think Ryzen was some slam jam throw together m'aam job, then nothing can help your cynicism and you are very well entitled to them. So far despite the amount of negitivity that is coming from the same group Ryzen continues to impress. I use Intel and Nvidia and that doesn't mean a rejuvinated AMD under a vibrant leader is not something I would love to see, because truth be told, my 4460 and 5960X will be just fine for now.

Nobody has ever claimed that Ryzen is going to smack Kaby lake at low clocks, and if they have it is there own perogative.
Usually products speak for themselves. If they need help from people on forums, they are bad products. If the don't need help, and are downplayed, by fans of other brands - they are great products.

You personally registered few days ago. You have been banned from one of other forums. There is huge problem with your credibility, and nobody in their own mind will believe you in the first place, until we will see hard facts. Ergo: Reviews of Ryzen chips.

So lets wait for reviews all, and do not hype, or downplay the CPUs.
 
  • Like
Reactions: ZipSpeed

bjt2

Senior member
Sep 11, 2016
784
180
86
Yup, the devil is in the details. The concept sounds incredible, the execution however will decide how it will pan out.
Judging the papers on Bristol Ridge, should be good. But on BR the clock was limited to the ones advertised...
On Zen should be without limits and with a 25MHz granularity...

In short: the periodic calibration made on the 1300 mini replicas is used to construct a table of Vcore and frequencies and from temperature a Vmax is calculated, that is the Vmax to apply to remain under safety constraint and to have a low wear, under the wear limit.
With those data (Vmax and frequency/Vcore table), the frequency is upped if the power consumption is under the limit AND the temperature is under the limit AND the Vcore needed for the new frequency is under the limit.
Otherwise the frequency stay the same or is lowered if any of this limit is exceeded.

With this mechanism, if you lower the temperature, with a good cooler, you can go upper because:

- Lower temperature means lower leakage, so less power drawn
- Lower temperature means less vcore for same frequency, that in turns means less power drawn
- Lower temperature means higher Vmax, that means you can go upper in frequency

So the mantra will be: keep'em cool
 
  • Like
Reactions: .vodka

bjt2

Senior member
Sep 11, 2016
784
180
86
I can almost imagine XFR refusing to hit advertised boost clocks at summer.

I'll reveal you a terrible secret: it is not motherboard that reports clocks.
There are not advertised boost clocks. The top turbo (4GHz for the top model) is guaranteed with 32C of ambient and stock cooler. If you don't have air conditioner, then you are right. But with 25/26C (sane summer air conditioning setting) XFR should kick in even with base cooler...
 

OrangeKhrush

Senior member
Feb 11, 2017
220
343
96
Usually products speak for themselves. If they need help from people on forums, they are bad products. If the don't need help, and are downplayed, by fans of other brands - they are great products.

You personally registered few days ago. You have been banned from one of other forums. There is huge problem with your credibility, and nobody in their own mind will believe you in the first place, until we will see hard facts. Ergo: Reviews of Ryzen chips.

So lets wait for reviews all, and do not hype, or downplay the CPUs.
haven't claimed anything, just based off the information leaked in what is now yet another benchmark. I haven't posted any guarantee on numbers or hypothesis that is not available from the numbers provided.

As for Toms, they apparently have mods that don't like you asking to put a person who claims something to proof. Said person claims he knows exactly what AMD's peformance is and it is set in stone. Obviously non of that is falling in place so far. Apparently it is a call out, well yeah because I am interested in knowing what this set in stone is.
 

zinfamous

No Lifer
Jul 12, 2006
108,303
25,130
146
I have q6600 cpu, with old MB from 2004 I guess, and I can boot from ssd. That computer works without problems, but now is too slow:). A friend of mine with phenom x4 965 be it can t install Windows on ssd. That's my concern. Don't want to be off topic, the point is this issues. And he is not a noob in computers. I want to get myself a new build from amd and I will do that. Hope there will be with not to many problems, like I mentioned or performance drop on ssd, USB and other devices connected. And there are bios bug issues, that scared me.

Sent from my HTC One_M8 using Tapatalk
That's strange. My first SSD, Corsair? maybe, 90gb something purchased ~2012--I installed on a then ~3-year old x4 965 with Win7 and never had a problem with installation or booting.

HOWEVER--I did move that SSD over to a new system (Xeon v3 2031, I think), as a backup SSD with a new Evo 950 500gb as the boot drive, and it could not recognize the older Corsair. I don't get it, but I did read somewhere that if you format the SSD before or after, and move it to another SATA port on Mobo--new MoBO or not--you might need to re-reformat for it to be recognized. Heh, I think that SSD is still connected and I haven't bothered to get it working, now 3 years later.
 
Status
Not open for further replies.

ASK THE COMMUNITY