AMD Ryzen 9 3900XT Memory Scaling Performance Under 100 Different Tests

[phoronix]

Phoronix: AMD Ryzen 9 3900XT Memory Scaling Performance Under 100 Different Tests

For those thinking of picking up one of the new AMD Ryzen 3000XT series processors and weighing whether it's worthwhile on your budget picking up DDR4-3600 memory or other higher frequency DDR4 modules, here are some fresh benchmark results with the Ryzen 9 3900XT looking at 100 different tests on Linux and showing how the performance changes from DDR4-2133 through DDR4-3800...

http://www.phoronix.com/scan.php?page=news_item&px=Ryzen-9-3900XT-DDR4-Memory

[schmidtbag]

It's interesting how (if memory serves me correctly, pun not intended) a high-clocked many-core Zen2 CPU seems to be less affected by memory speed than a Zen1 CPU with half the cores and a lower top speed. And that's a good thing - high-speed RAM is expensive.

[bug77]

It's interesting how (if memory serves me correctly, pun not intended) a high-clocked many-core Zen2 CPU seems to be less affected by memory speed than a Zen1 CPU with half the cores and a lower top speed. And that's a good thing - high-speed RAM is expensive.

It's not less affected. Zen2 runs IF at higher frequencies and Michael "cheated" a little. 3800 is about as high as IF will go. If faster memory was tested, the chips wouldn't have been able to run memory and IF 1:1. And you would have seen the same drop you see for earlier chips.

[ntropy]

Intresting. I had my R9 3900X going at 3200 Mhz mem freq for some time but now went back to 2400 Mhz. I dont really feel a difference there, except the cpu fan is quieter. But regarding normal day C compilation and gaming, I see no difference.

Only thing I reckoned with the 3900X is compression sometimes takes awfully long. Seems like my i7-9750H laptop cpu was faster on that. Yet the Intel does only zip and unzip, dunno how good the zip compression is.

[zxy_thf]

It's interesting how (if memory serves me correctly, pun not intended) a high-clocked many-core Zen2 CPU seems to be less affected by memory speed than a Zen1 CPU with half the cores and a lower top speed. And that's a good thing - high-speed RAM is expensive.

High Speed JEDEC RAMs are expensive, but high speed XMP 2.0 RAMs are not.
They are mostly JEDEC DDR4-2133 overclocked at 1.35V.

[nranger]

Intresting. I had my R9 3900X going at 3200 Mhz mem freq for some time but now went back to 2400 Mhz. I dont really feel a difference there, except the cpu fan is quieter. But regarding normal day C compilation and gaming, I see no difference.

Only thing I reckoned with the 3900X is compression sometimes takes awfully long. Seems like my i7-9750H laptop cpu was faster on that. Yet the Intel does only zip and unzip, dunno how good the zip compression is.

I'm happy for you that you have a good experience with slower RAM. Having a cooler, quieter computer is a valid trade off for higher performance, and you've made the choice that makes sense for you.

That being said, I don't think your anecdotes contradict any of the data we've seen from Phoronix and elsewhere regarding Ryzen memory performance.
1) Your computer probably saves at most a few watts running the memory controller & Infinity Fabric at lower clocks. But your cores are also going to draw less power if they're sitting idle waiting for data.
2) Code compilation is much more cache and storage bound than memory bandwidth or latency dependent, so I'm not surprised you don't see a difference in that workload.
3) Ryzen RAM speed only affects gaming if you're not already CPU bound. And unless you're running a top-top end GPU while gaming at low resolutions and high framerates, you won't see a difference.
4) Ryzen 3000 is a beast in compression/decompression workloads, but the public benchmarks I've seen are usually 7zip or WinRAR. On the same compression software, a 12 core 24 thread desktop Ryzen will crush a 6 core Intel mobile chip. If compression is a workload you use often, you may want to optimize your system and workflow -- are you using a modern compression format, are you using the multi-threaded options in your tools, etc.

I'm a tinkerer - so I'm happy to waste a few hours tuning memory speeds and timings just to see what the results are. For someone just wanting to get work done, a few percent in a select group of applications isn't going to be worth it. In your case, if you're happy with 2400Mhz RAM, good for you (though you might get some benefits leaving the RAM at 2400 and setting Infinity Fabric clock up to 1867 or 1900Mhz). In the future if you get a better GPU, or a high refresh rate monitor, the extra performance is there waiting for you.

[Melcar]

Even with Ryzen 1xxx CPUs RAM speed was not that important. Gaming was/is where you see differences, but only in CPU bound scenarios.

[nranger]

Even with Ryzen 1xxx CPUs RAM speed was not that important. Gaming was/is where you see differences, but only in CPU bound scenarios.

Very true. The difference was "measurable" but it was debatable how important it was. Youtuber & Tech Press reviewers have a mostly unconscious (or at least unstated) goal of showing graphs where there are products with higher numbers and products with lower numbers. The tests that produce those graphs are oftentimes unrealistic.

[frief]

"performance with a pair of 2 x 8GB GSKILL DDR4-4133 modules operating at different frequencies"

So just the frequencies were different and wait states were not reduced for lower frequencies?
I may have overread something but I didn't see wait states mentioned.

[frief]

just for the record: part of the (274 lines of) output of decode-dimms

[FONT=Courier New]# decode-dimms | grep "tCL-"
tCL-tRCD-tRP-tRAS 9-9-9-24
tCL-tRCD-tRP-tRAS as DDR3-1333 9-9-9-24
tCL-tRCD-tRP-tRAS as DDR3-1066 7-7-7-20
tCL-tRCD-tRP-tRAS as DDR3-800  6-6-6-15[/FONT]

probably politely tells me that I should consider an upgrade.