AMD - Zen chitchat

[CAT-THE-FIFTH]

Hmm,not too bad,its getting close to Intel SKL SKUs at similar clockspeeds.

[DanceswithUnix]

Interesting to see Threadripper doing so badly.

Also interesting to see the Phenom X6 so badly at the bottom of the table. There were a lot of people saying that Piledriver was a backwards step from the older Stars cores. I didn't think the numbers backed that up at the time, and the years haven't been so kind to the Phenom. Still, a 2600X is looking mighty tempting...

[CAT-THE-FIFTH]

Interesting to see Threadripper doing so badly.

Also interesting to see the Phenom X6 so badly at the bottom of the table. There were a lot of people saying that Piledriver was a backwards step from the older Stars cores. I didn't think the numbers backed that up at the time, and the years haven't been so kind to the Phenom. Still, a 2600X is looking mighty tempting...

I am kind of stuck now - its not clear if most review sites have applied the full Intel patches,since this will reduce performance in FO4 as it has done with my IB Xeon E3 1230 V2,but at the same time Intel has the edge due to the way the engine works. Having said that Ryzen 2000 will be actually be a slight upgrade now.

[watercooled]

Anandtech have some interesting results that show AMD mostly ahead of even Coffee Lake. They do have all patches applied it seems - I wonder if that's what's causing the difference? Surely there aren't a bunch of tech websites not bothering to update their Intel results?

https://www.anandtech.com/show/12625...-2600x-2600/19

[CAT-THE-FIFTH]

Anandtech have some interesting results that show AMD mostly ahead of even Coffee Lake. They do have all patches applied it seems - I wonder if that's what's causing the difference? Surely there aren't a bunch of tech websites not bothering to update their Intel results?

https://www.anandtech.com/show/12625...-2600x-2600/19

TR has applied all the patches too,but not finished their games testing.

[watercooled]

Tom's also shows it to be extremely close to the 8700k for gaming: https://www.tomshardware.com/reviews...ew,5571-7.html

Edit!

A very interesting section from Tom's (emphasis mine):

Spectre And Meltdown

Our test rigs now include Meltdown And Spectre Variant 1 mitigations. Spectre Variant 2 requires both motherboard firmware/microcode and operating system patches. We have installed the operating system patches for Variant 2.

Today's performance measurements do not include Intel's motherboard firmware mitigations for Spectre Variant 2 though, as we've been waiting for AMD patches to level the playing field. Last week, AMD announced that it’s making the mitigations available to motherboard vendors and OEMs, which the company says should take time to appear in the wild. We checked MSI's website for firmware updates applicable to our X370 platforms when AMD made its announcement, but no new BIOSes were available (and still aren't).

Unfortunately, we were only made aware that Variant 2 mitigations are present in our X470 board's firmware just before launch, precluding us from re-testing the Intel platforms with patches applied. We're working on this now, and plan to post updated results in future reviews.

The lack of Spectre Variant 2 patches in our Intel results likely give the Core CPUs a slight advantage over AMD's patched platforms. But the performance difference should be minimal with modern processors.

Good on them for disclosing it, but as we know, day 1 numbers are quite important so they could really be doing AMD a disservice if they turn out to make any real difference. The gaming results are very close as-is though, so even a small performance uplift could take Ryzen ahead.

Edit2: Interesting Tweet from Anandtech - perhaps some are testing with Precision Boost disabled because of how it's presented in BIOS? https://twitter.com/IanCutress/statu...26323544649729

[Xlucine]

Surely bios level patches are going to be superior to OS level patches? If they both protect and the bios level comes with a performance penalty, then there's no need to update the bios

[watercooled]

They're not either/or, both types of patches must be applied to fully mitigate the vulnerabilities.

[DanceswithUnix]

Surely bios level patches are going to be superior to OS level patches? If they both protect and the bios level comes with a performance penalty, then there's no need to update the bios

I believe they work in parallel. Bios gives new barrier instructions for the OS to work with.

The spectre workarounds I have seen so far have been minimal impact as well, not like meltdown.

[scaryjim]

... Edit2: Interesting Tweet from Anandtech - perhaps some are testing with Precision Boost disabled because of how it's presented in BIOS? https://twitter.com/IanCutress/statu...26323544649729

Perhaps, but the Hexus review clearly tests with it on and they get Intel clearly ahead in the gaming tests...

Something's up somewhere, and given that AT is the only place that gets Ryzen 2 consistently top in all gaming benchmarks, I'm inclined to question their benchmark scores before I question everyone else's...!

[CAT-THE-FIFTH]

Tom's also shows it to be extremely close to the 8700k for gaming: https://www.tomshardware.com/reviews...ew,5571-7.html

Edit!

A very interesting section from Tom's (emphasis mine):



Good on them for disclosing it, but as we know, day 1 numbers are quite important so they could really be doing AMD a disservice if they turn out to make any real difference. The gaming results are very close as-is though, so even a small performance uplift could take Ryzen ahead.

Edit2: Interesting Tweet from Anandtech - perhaps some are testing with Precision Boost disabled because of how it's presented in BIOS? https://twitter.com/IanCutress/statu...26323544649729

Interesting that - TR also has apparently patched the Intel boards and so did AT.

[watercooled]

Perhaps, but the Hexus review clearly tests with it on and they get Intel clearly ahead in the gaming tests...

Something's up somewhere, and given that AT is the only place that gets Ryzen 2 consistently top in all gaming benchmarks, I'm inclined to question their benchmark scores before I question everyone else's...!

F1 and Far Cry 5 are the ones substantially ahead at Hexus and Anandtech hasn't tested those. TPU have, and while F1 is closer, FC5 still shows a gap between AMD and Intel: https://www.techpowerup.com/reviews/..._2700X/13.html

[CAT-THE-FIFTH]

Perhaps, but the Hexus review clearly tests with it on and they get Intel clearly ahead in the gaming tests...

Something's up somewhere, and given that AT is the only place that gets Ryzen 2 consistently top in all gaming benchmarks, I'm inclined to question their benchmark scores before I question everyone else's...!

AT have not only patched up the systems,but more important have only tested with stock RAM settings - Intel at 2666MHZ and AMD at 2933MHZ. So if AT ran faster RAM and overclocked both sets of CPU,Intel will probably win.

But if you look at the TH review,even with 3466MHZ DDR4 on both platforms,AMD is much closer than expected in a number of titles which tended to perform better on Intel,due to their lightly threaded nature.












After all,AMD Ryzen 2 gained nearly 16% in one of its worse titles,ie,FO4 which has zero optimisation on an engine which tends to love Intel CPUs.

The DAW improvements are enormous and its no wonder when TH measured better cache latency improvements than what AMD said would happen:

https://www.tomshardware.com/reviews...ew,5571-3.html

% Improvement Over 1800X
L1
L2
L3
In-Page
11.11%
51.72%
26.38%
Full-Random
11.11%
53.5%
25.64%
Sequential
11.11%
13.3%
13.3%

[Xlucine]

Some of these are eerily close:

https://www.anandtech.com/show/12625...-2600x-2600/15

https://www.tomshardware.co.uk/amd-r...w-34307-6.html

Assuming the AT results are minimums (because they fit so well!):
-8700k is identical
-7820x lost 12 fps
-2700x gained 9 fps
-1800x is identical
-1700x is identical
-1600x gained 2 fps
-1700 lost 1 fps

AT have not only patched up the systems,but more important have only tested with stock RAM settings - Intel at 2666MHZ and AMD at 2933MHZ. So if AT ran faster RAM and overclocked both sets of CPU,Intel will probably win.

Hexus used the same RAM speeds, and so did tomshardware (i.e. max rated, except TH used 2.7 GHz RAM in all intel 8th gen CPUs).

[scaryjim]

Some of these are eerily close:

Assuming the AT results are minimums (because they fit so well!):
...

AT used Ultra settings, while tom's used High settings, which is why AT's averages look vaguely similar to tom's minimums. Sorry to rain on your parade...

[CAT-THE-FIFTH]

Also here are some apples to apples timed runs:

https://www.youtube.com/watch?v=Mr2B0RJd7Nc&



Another FO4 test,where Ryzen 2 is doing far better than expected. Its a timed run around Swan's Pond in one of the denser parts of the map with more NPCs. Sweclockers was testing around Diamond City which is more a small hub area you don't visit often.

The Core i7 8700K is running at 4.4GHZ and the Ryzen 7 2700X at 4.2GHZ,and this game just loves Intel CPUs. You can tell its a CPU limited area as GPU utilisation drops around the city.

The channel tested the Ryzen 7 1800X and it was definitely behind the Core i7 8700K.

Even GTA V is doing better than expected.

OK,this is getting interesting.

Hexus used the same RAM speeds, and so did tomshardware (i.e. max rated, except TH used 2.7 GHz RAM in all intel 8th gen CPUs).

TH used 3466MHZ DDR4 for all the overclocked results:

First-gen Ryzen processors don't have much memory overclocking headroom, so we're still testing tuned X370 platforms at DDR4-3200. However, the X470 platform was remarkably stable at higher data rates with Ryzen 7 2700X. So, we settled on DDR4-3466 with 14-14-14-34 timings (though we're confident that more time to tune would yield even higher overclocks). We also ran our overclocked Intel processors at DDR4-3466.