Discussion Ryzen 3000 series benchmark thread ** Open **

[DrMrLordX]

KitGuru:

 

[DrMrLordX]

TPU is up with their reviews.

Again with the low power numbers for 3900X? Sadly they didn't say which cooler they used.

 

[Abwx]

, and much less reason to buy any Matisse other than the 3600.

lol.....

 

[PotatoWithEarsOnSide]

All of these reviews showing PBO having no effect may be due to non-final BIOS.
Anandtech review suggests inability to change PPT etc, which is one of the functions of PBO.

 

[misuspita]

Anand tech review is up

 

[Head1985]

I was expecting more.Still slower than intel in games and crap OC atleast its cheap.

 

[jpiniero]

All of these reviews showing PBO having no effect may be due to non-final BIOS.
Anandtech review suggests inability to change PPT etc, which is one of the functions of PBO.

Toms made it sound like they were getting better results with PBO than even manually overclocking it.

 

[Asterox]

lol.....

Yes jpiniero comment is legit, but only if he is a scenarist for Monty Python series.

But AMD also bribed Adobe so.......................

 

[Head1985]

b450 vs x570

 

[Kenmitch]

Anand tech review is up

I'm not sure what he used during the whole review....You'd think he'd have some b-die laying around.

The figures published on this page are run on DDR4-3200CL16 on the Ryzen 3900X and 2700X at timings of 16-16-16-36, and the i9-9900K was run with similar DDR4-3200CL16 at timings of 16-18-18-36.

 

[Glaring_Mistake]

Sweclockers: https://www.sweclockers.com/test/27760-amd-ryzen-9-3900x-och-7-3700x-matisse

English translation: https://translate.google.com/transl...t/27760-amd-ryzen-9-3900x-och-7-3700x-matisse

 

[birdie]

The Ryzen 3700X CPU looks the sweetest deal for most users out there. Relatively future-proof, relatively inexpensive (still I'd be glad if it cost, say $270). A CPU to run for the next 7 years, lol. Time to upgrade my aging Intel Core i5 2500.

 

[Head1985]

Anyone have 3800x review?

 

[IEC]

Anyone have 3800x review?

Per reviewers, AMD provided R9 3900X and R7 3700X parts for launch. We'll have to wait for reviews of the other parts...

 

[birdie]

Sweet Jesus:

From https://techreport.com/review/34672/amd-ryzen-7-3700x-and-ryzen-9-3900x-cpus-reviewed/16

Sky Lake/Coffee Lake is completely defeated.

 

[jpiniero]

lol.....

Looking at the Sweclockers review, there was one game where there was an actual difference between the 3700X and 3900X, Ass Creed, and that might be more because of clocks than the core count. Since it sounds like there won't be much of a difference in all core clocks once OC'd it does more or less confirm that gaming perf will be basically identical between the different Matisse models.

Does really bode well that Threadripper won't have any downside for gaming purposes.

 

[coercitiv]

inter-core latency comparison from Hardware Unboxed

 

[StinkyPinky]

https://www.newegg.com/promotions/CPU/19-1476/index.html?cm_sp=Cat_CPUs-Processors-_-CPU/19-1476-_-//promotions.newegg.com/CPU/19-1476/1920x360.jpg&icid=524355

You snooze you lose. Happy to wait though.

 

[IEC]

That inter-core latency is comparable to a ring-bus design. Not quite as good as the refined Intel design, but nonethless very impressive for modular chiplet(s) + IO.

The $1000 question for me is: With Zen 2's chiplet architecture apparently paving the way for Threadrippers with no gaming deficits compared to the mainstream Ryzen parts, will Threadripper beat even the 3900X in gaming?

I see some potential benefits:
1) Bigger L3 cache
2) Quad-channel memory
3) Platform/AGESA/BIOS fixes and tweaks by the time TR3 launches in the (late?) fall

If they match or beat the 3900X in gaming then I have no reason not to run TR as my primary rig.

 

[Kedas]

Some results are kind of strange for higher resolution in games.
Ryzen 3000 drops more than previous Ryzen when the resolution goes up.
(even dropping below 2700x)

 

[exquisitechar]

The Stilt's posts on Matisse.

IPC is on par with Coffee Lake S parts in his suite. Many interesting observations, same as last time.

 

[Abwx]

Looking at the Sweclockers review, there was one game where there was an actual difference between the 3700X and 3900X, Ass Creed, and that might be more because of clocks than the core count. Since it sounds like there won't be much of a difference in all core clocks once OC'd it does more or less confirm that gaming perf will be basically identical between the different Matisse models.

Does really bode well that Threadripper won't have any downside for gaming purposes.

One game is not a generality, I would still expect a 3700X to be better than a 3600 in games by the virtue of more cores, not counting that PS/XBX games will benefit from 8C Zen 2 and will be iterated for the PC world, we saw the pace at wich quadcores obsoleted once there were more cores at disposal...

 

[badb0y]

Solid CPUs from AMD but I’m a bit disappointed with the OC potential on these. Either way anyone on the 2600kish era seems to have a nice upgrade path here.

 

[coercitiv]

der8auer on Ryzen 3000 overclocking and delidding: it's over 5000!!! Just kidding, welcome back to reality and what sensible forum posters tried to explain to hype train experimental passengers.

 

[exquisitechar]

der8auer on Ryzen 3000 overclocking and delidding: it's over 5000!!! Just kidding, welcome back to reality and what sensible forum posters tried to explain to hype train experimental passengers.

Yes, the OC potential/voltages required are a reality check for the more optimistic folks. Getting big clock speed increases on a new node is harder than ever.

Seems that the advertised boost clocks on the 3900x are not even reached in reality. Even with very high voltages, 4.6GHz is barely reachable on a single core for brief moments.