• Guest, The rules for the P & N subforum have been updated to prohibit "ad hominem" or personal attacks against other posters. See the full details in the post "Politics and News Rules & Guidelines."

Official AMD Ryzen Benchmarks, Reviews, Prices, and Discussion

Page 59 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.

inf64

Diamond Member
Mar 11, 2011
3,062
1,783
136
https://twitter.com/CPCHardware/status/836346777267761155

Our primary source for Ryzen information for many months is claiming the IMC is slow.


I had considered that if the IMC really is approaching near 100% of the theoretical bandwidth, this may have been a concious decision by AMD to have higher latency for better throughput.
This would be perfect for Raven Ridge.
Except there is one problem in what Doc from CPC HW is saying: his own gaming benchmarks he performed on early 8C/16T ES @ 3.3Ghz MAX boost. This chip scored ~9% lower than 6900K which boosts to 3.5-3.7Ghz in gaming benchmarks. Also his game test selection is more inclined towards higher IPC/lower thread CPUs than the MTed monsters like Ryzen 7 and BDW-E. Still, that ES chip in development board scored great for a low clock and non-final BIOS/Agesa it had.

He seems overall salty for some reason and even though his own numbers are telling us Ryzen performs great both in apps and games he still is giving weird information via twitter("slow" IMC- how can IMC be slow? ; high TDP, low OCs, low clocks on 4C models etc).
 

lolfail9001

Golden Member
Sep 9, 2016
1,056
353
96
The sisoft sandra benchmark showed totally different clock speeds so who ever did that benchmark probably did not use the xmp profile.
Sandra does not detect clock speed of memory correctly, apparently.
Regarding rated TDP of the cooler: i have a 130W sandy bridge 2011, with a cooler rated 220W, so the rating of the cooler does not mean anything...
Your tower cooler was bundled with CPU?
And they sent 1800X s to reviewers, which is compared to the 6900X, by them ;)
Yep, but i expect the good reviews to compare to all 3 i7s AMD positions Ryzens against.
Except there is one problem in what Doc from CPC HW is saying: his own gaming benchmarks he performed on early 8C/16T ES @ 3.3Ghz MAX boost. This chip scored ~9% lower than 6900K which boosts to 3.5-3.7Ghz in gaming benchmarks. Also his game test selection is more inclined towards higher IPC/lower thread CPUs than the MTed monsters like Ryzen 7 and BDW-E. Still, that ES chip in development board scored great for a low clock and non-final BIOS/Agesa it had.
Yes, so what?
he still is giving weird information via twitter("slow" IMC- how can IMC be slow? ; high TDP, low OCs, low clocks on 4C models etc).
What is weird about it except painting Ryzen as chip that is not perfect?
 

coercitiv

Diamond Member
Jan 24, 2014
3,953
4,272
136
Except there is one problem in what Doc from CPC HW is saying: his own gaming benchmarks he performed on early 8C/16T ES @ 3.3Ghz MAX boost. This chip scored ~9% lower than 6900K which boosts to 3.5-3.7Ghz in gaming benchmarks. Also his game test selection is more inclined towards higher IPC/lower thread CPUs than the MTed monsters like Ryzen 7 and BDW-E. Still, that ES chip in development board scored great for a low clock and non-final BIOS/Agesa it had.
That tweet comments performance contribution of the IMC to benches/games, not overall performance of Zen in games.

On the other side, if I may be allowed to semi-troll a bit, since many consider modern games performance scale solely with memory bandwidth and not latency, one could consider the "slow" IMC of Zen to be an actual advantage :smilingimp:
 

inf64

Diamond Member
Mar 11, 2011
3,062
1,783
136
Sandra does not detect clock speed of memory correctly, apparently.

Your tower cooler was bundled with CPU?

Yep, but i expect the good reviews to compare to all 3 i7s AMD positions Ryzens against.

Yes, so what?

What is weird about it except painting Ryzen as chip that is not perfect?
Nobody is saying Ryzen is a perfect chip. There is no perfect chip out there.
 

looncraz

Senior member
Sep 12, 2011
716
1,638
136
TDP stands for Thermal Design Power.
When an electrical apparatus is powered with X watts, the sum of kinetic energy, electromagnetic energy and thermal energy MUST be X, because of law of conservation of energy.
Kinetic energy is zero, because AFAIK a powered PC does not move.
Electromagnetic energy is negligible because of FCC rules and because otherwise all of us would be sterile due to testicle overheat.
All remaining energy (almost all) does turn in heat.
Regarding rated TDP of the cooler: i have a 130W sandy bridge 2011, with a cooler rated 220W, so the rating of the cooler does not mean anything...
Thanks, I edited that comment - no idea why I called it "Typical Dissipation Power." Probably something to do with the migraine I've had all day (not joking, my migraines often last for days and really play with my language processing... I've been known to write entire sentences backwards without realizing it - just glad they're not really painful any more)... though it may have something to do with my long-ago studies regarding circuit power dissipation (as heat)... no idea.

I was trying to speak in more generic terms regarding kinetic energy - even kinetic energy eventually dissipates as heat through friction... though some might argue that heat is a form of kinetic energy at a certain level... I won't make that argument :p

So... to that TDP rating on a cooler. Yes, it means something :p Or it's supposed to, anyway. The basic formula is the same, but everyone uses different inputs... different ambient, t-case temp, air flow, interface properties, etc...
 
  • Like
Reactions: Drazick

JDG1980

Golden Member
Jul 18, 2013
1,662
565
136
Optimizing the memory controller for bandwidth, at the expense of latency, may have been a deliberate design decision. Zen's large, fast cache should mitigate most of the disadvantages, while higher bandwidth will prove very useful with upcoming Raven Ridge APUs. (Theoretically AMD could have used different memory controllers on Summit Ridge and Raven Ridge, but they may not have had the resources to do so.)
 

french toast

Senior member
Feb 22, 2017
978
803
106
And so the anti amd fud brigade starts sharpening their fangs, you can tell intel has been whispering in the ear of some impressionable reviewers, "focus on sub tests and isolate/magnify the bad parts", you watch AVX256 its magically going to gain importance, even though its next to useless for the foreseeable future.

Point out that ryzen exceeds its tdp rating in some crazy load scenario whilst drawing attention away from the fact ryzen is considerably more efficient overall.
Focus on silly sub tests like latency and max rated memory speed to blackball amds IMC, when in reality AMDs IMC is more efficient- (achieving more bandwidth out of lower specced memory saving yet MORE money for the consumer)- any latency teething issuesare likely to make NO difference in real world applications due in part to ryzens excellent cache system.

Hell why they are at it spread some fud about overclocking, compare ryzen against skylake quads to point out how bad overclocking they are, whilst ignoring the fact intels twice as expensive HEDT only OCs 100mhz more.

Oh and those of you who have clamped onto this latency subtest like a leech, suggesting it will affect games consider this; in almost every game we have seen so far shows ryzen beating out intels HEDT, even the R7 1700 was shown to equal the Messiah 7700k in average fps GTA5- hardly the most modern game out there.

Lets wait for some reviews shall we to see how this terrible IMC holds up in the real world, from leaks we have seen it does very well indeed.
 

inf64

Diamond Member
Mar 11, 2011
3,062
1,783
136
Yep, no Ryzen visible in those :)

edit: Hmm, there is Ryzen in some charts but it is missing in others (Excel for example). Are they benching this chip or taking results from online leaks?
 

Whitestar127

Senior member
Dec 2, 2011
382
8
81
Ryzen is in there, just not in all the benchmarks apparently. Check under 3DMark and Cinebench, for example.

EDIT: Apparently they forgot to mention Ryzen on the Introduction page.
EDIT2: And it was published on January 3.
 
Last edited:
  • Like
Reactions: looncraz

R0H1T

Platinum Member
Jan 12, 2013
2,566
142
106

ASK THE COMMUNITY

TRENDING THREADS