Share your AIDA 64 cache and memory benchmark here

[Wormisko]

Dont know if its good, but iam fine

 

[Marcoz]

Dont know if its good, but iam fine

try this if you want.

 

[Wormisko]

try this if you want.

will check it out, thanks mate

 

[The King]

will check it out, thanks mate

He is running Samsung B-die RAM and you are not. Those settings will not work for you so I don't recommend you try them.

 

[Marcoz]

will check it out, thanks mate

In fact if you don't have samsung b-die don't try my times I didn't know what kind of memory chip you had.

 

[KokoSloan]

Dual vs. single rank seems to affect only copy speed, boosting it by over 2000 MB/s. But read and write didn't change when using the same settings:
View attachment 288128

Compare your score , having almost best kit money can buy SR and my old 2400 cl15 cheap ancient kit DR ..(Micron A-die on 1.35V )

 

[freeagent]

Compare your score , having almost best kit money can buy SR and my old 2400 cl15 cheap ancient kit DR ..(Micron A-die on 1.35V )

View attachment 289016View attachment 289017

That’s a weird flex.

 

[QuietBob]

Compare your score , having almost best kit money can buy SR and my old 2400 cl15 cheap ancient kit DR ..(Micron A-die on 1.35V )

View attachment 289016View attachment 289017

Nice overclock on that budget kit!
However, results from different CPUs/architectures are not directly comparable. And there are people with much better results than mine on DDR4 in this very thread.

 

[freeagent]

You guys are all doing awesome..

 

[Hardi]

You guys are all doing awesome..

no problem

 

[freeagent]

no problem

 

[Det0x]

Aida doing aida things again..

If you score above 102000 MB/s in the aida64 cache and memory benchmark then the unit of measurement changes from MB/s to GB/s. (1000kb per mb -> 1024kb per mb)
Then your "~102700MB/s" suddenly becomes "~100.7GB/s"

 

[dont whant to set it"']

102739MB/s÷1024=100.3310546875GB/s
I've noticed that as well years ago , that single tests scores above or below the benchmark suite, all within run to run variance I'd say.

 

[freeagent]

Fresh windows install

Edit:

Found a pic of my 5600X on my other board

 

[Athlonite]

Mine at the moment

 

[Marcoz]

Finally my current profile is stable and 100% working at 1.50v

 

[bersy]

Finally my current profile is stable and 100% working at 1.50v

And that's quite a similar one

 

[Braegnok]

New G.SKILL F5-6000J3040G32GX2-TZ5N.

My 32GBx2 2023 Mar kit is Hynix A-die with Richtek OC PMIC.

Seems not all the 32GBx2 6000CL30 kits are A-die, check the SKU number above the barcode for last letter "A",.. some kits are M-die.

 

[bersy]

DDR4 M8H OEM 4x8 4000CL17 CR1 1.55V – AMD 5600G APU, MSI B450M Mortar (daisy-chain)

tPHYRDL 26 on both channels.

It was quite challenging to stabilize 4 modules in such a working mode. It boots stable at 4133CL18 CR1 too, but 4000 is a ceiling as for CL17.

M8H have relatively low RFC ~190 ns, but RCD is higher than M8E.

 

[The King]

DDR4 M8H OEM 4x8 4000CL17 CR1 1.55V – AMD 5600G APU, MSI B450M Mortar (daisy-chain)

tPHYRDL 26 on both channels.

Any WHEA 19 errors at 2000 FCLK and above?

I was looking at a 5700G maybe 5600G to do some RAM OCing above 2000 FCLK but was worried my board (B450M Mortar Max) may not handle the higher MCLK/FCLK.
Seems a APU makes up for this has my Vermeer CPU struggles with WHEA at 2000 FCLK and higher.

I can pass TM5 but if you run something like Y-rcunher 2.5B the WHEA errors begin to pile up. So not practical for daily use.

 

[bersy]

Any WHEA 19 errors at 2000 FCLK and above?

WHEA at APU? It never happens. The same goes to APU based CPUs (4100, 4500 and 5500) afaik.

my board (B450M Mortar Max) may not handle the higher MCLK/FCLK.

Mine one is capable to boot at 4733 MHz with high bin ICs.

 

[The King]

There is no WHEA at APU at any MHz at all.


Mine one is capable to boot at 4733 MHz with high bin ICs.

I can boot has high has 4800MT but unsynced.

Some of the guys on OCN forums can boot over 5000 FCLK using a 5700G and B550 Unify-X which is insane for DDR4 and 1:1:1 MCLK/FCLK/UCLK.

 

[bersy]

I can boot has high has 4800MT but unsynced.

Well I actually meant 1:1 mode.

Some of the guys on OCN forums can boot over 5000 FCLK using a 5700G and B550 Unify-X which is insane.

That's a specific, memory overclock oriented mobo. So it doesn't look so insane to me.

 

[The King]

That's a specific, memory overclock oriented mobo. So it doesn't look so insane to me.

It more to do with the RAM IC than the board itself he got similar results using a entry level Gigabyte B550.

The Samsung B-die IC are not the best for doing very high MLCK, Micron Rev.E/B and some Hynix IC seems to be key in getting these higher MCLK stable.

Anyway cool that you can run that on a B450 was always impressed just how well this cheap board can OC RAM even able to run GDM disable real 1T on both my B450s.
Many with high end board struggle to even get that to run stable.

 

[bersy]

Anyway cool that you can run that on a B450 was always impressed just how well this cheap board can OC RAM even able to run GDM disable real 1T on both my B450s.

Same here

DDR4 M8J OEM 2x16 4200CL17 CR1 1.48V w/ custom ReCold heatsink