Question Post your MaxxMem² score

[igor_kavinski]

Thanks to Det0x for using this utility in his benchmarking and thanks to AdamK47 for posting his score, thus making me aware of this utility.

Det0x's score: https://forums.anandtech.com/threads/zen-4-builders-thread.2605910/post-40890534

AdamK47's score: https://forums.anandtech.com/threads/zen-4-builders-thread.2605910/post-40890893

This is what I got on my Dell Optiplex work PC:



This is the best of three runs.

No other software should be running as it will result in lower scores.



Compare feature for DDR3 shows bandwidth is acceptable but latency could be better.

Download link: http://maxxpi2.de/pages/downloads/maxxmemsup2---download.php

Click the name next to the tiny Winzip icon:

 

[igor_kavinski]

Office PC stuck in single channel offers better latency than my i7-4770 work PC.

 

[igor_kavinski]

Wow. That latency is baddddddd. Never felt the system had poor responsiveness. Maybe the L4 cache is helping. However, this tool behaved weird on this system. It basically got stuck a lot longer than normal when doing the latency calculation. So possible that the tool is getting confused somehow.

 

[igor_kavinski]

Previous PC's Copy score was relatively high whereas this laptop's Write score is high. I wonder why.

 

[AdamK47]

Small tweaks while maintaining stability.

 

[lopri]

6400 Hynix


6400 Samsung


6200 Together (With 4 sticks at 6200 the board changes command rate to 2T and I cannot find where to change it in the ASUS BIOS)

 

[igor_kavinski]

6400 Hynix
View attachment 73075

This seems to be your best kit. But for some weird reason, the extra bandwidth isn't available. Your bandwidth should theoretically be higher than AdamK47's DDR5-6200. Maybe the extra cache on the 7950X is playing some role.

 

[lopri]

Adam knows how to get more out of his hardware, that is why.

 

[igor_kavinski]

Adam knows how to get more out of his hardware, that is why.

He tuned for better timings. Timings shouldn't have any bearing on bandwidth availability. I think there is some limitation in Zen 4's memory controller.

 

[AdamK47]

I was testing out DDR5 6000 over the weekend. Still stable with lower timings compared to 6200. Score went up.

 

[lopri]

tRFC 160?! Holy kangaroo

 

[igor_kavinski]

Zen 4 IMC is really counter-intuitive!

 

[AdamK47]

With 6000 I am running 1T. 6200+ automatically uses 2T. That on top of the other lower timings allows me to pull ahead slightly.

 

[BoomerD]

Thanks to Det0x for using this utility in his benchmarking and thanks to AdamK47 for posting his score, thus making me aware of this utility.

Det0x's score: https://forums.anandtech.com/threads/zen-4-builders-thread.2605910/post-40890534

AdamK47's score: https://forums.anandtech.com/threads/zen-4-builders-thread.2605910/post-40890893

This is what I got on my Dell Optiplex work PC:

View attachment 71307

This is the best of three runs.

No other software should be running as it will result in lower scores.

View attachment 71309

Compare feature for DDR3 shows bandwidth is acceptable but latency could be better.

Download link: http://maxxpi2.de/pages/downloads/maxxmemsup2---download.php

Click the name next to the tiny Winzip icon:

View attachment 71310

Chrome blocks the MaxxMem2 link as a trojan.

Edit: got it from MajorGeeks.

 

[Zor Prime]

It detects nothing on my PC for some reason even if I run it as administrator. Win11 ASRock X570 Taichi 5800X3D box.

 

[igor_kavinski]

It detects nothing on my PC for some reason even if I run it as administrator. Win11 ASRock X570 Taichi 5800X3D box.

Maybe it's getting confused by the V-cache? Drop a message to the software's author.

 

[Zor Prime]

Maybe it's getting confused by the V-cache? Drop a message to the software's author.

That's what I was thinking.

 

[HutchinsonJC]

It didn't populate OS, CPU, Memory, Mainboard, etc but it gave a result for read, write, copy, bandwidth.

Read: 42982
Write: 28877
Copy: 57640
Bandwidth: 43.17

Latency populated at the bottom as 0.0 ns, which is on the way to infinity bandwidth, I guess (jokes)

Windows 10
AMD 5950x
Asus Crosshair VIII Hero
G.Skill Trident Z Neo Series 32GB (2 x 16GB) 288-Pin SDRAM PC4-28800 DDR4 3600 CL16-19-19-39 1.35V Dual Channel Desktop Memory Model F4-3600C16D-32GTZNC

 

[igor_kavinski]

Read: 42982
Write: 28877
Copy: 57640
Bandwidth: 43.17

Interesting. 5950X IMC behavior is similar to 7950X. Write is slowest and copy is fastest. It's like they just took the 5950X IMC and tacked on DDR5 support to it.

 

[lopri]

Write speed scales with Infinity Fabric regardless of memory speed.

 

[igor_kavinski]

Write speed scales with Infinity Fabric regardless of memory speed.

I keep reading about the IF speed limitation almost everywhere. Is this a TSMC process limitation that prevents AMD from increasing it? Do the Epyc CPUs have higher IF speed?

 

[In2Photos]

My daughter's new PC. No tuning, just enabled the EXPO setting. She has the same kit as lopri so it looks like I could tinker and get better results, but I've never messed with RAM timings and it is my daughter's PC after all, so I'll probably leave it as is.

 

[igor_kavinski]

HP ProBook 450 15.6 inch G9 Notebook

Core i5-1235U with dual channel 2x8GB DDR4-3200

HT enabled - HT disabled



Disabling HT increases the bandwidth a bit. Latency not showing probably due to some bug in software.

 

[AdamK47]

I was testing out DDR5 6000 over the weekend. Still stable with lower timings compared to 6200. Score went up.

More tweaking. Same speed of 6000.

 

[AdamK47]

Up up up I go. Not sure I can get any better than this at 6000. Ran stress testing for over three hours for stability. Will run more to confirm.