Share your AIDA 64 cache and memory benchmark here

[Makaveli]

Not on it per se.. above it yes..

Was able to tweak some voltages a bit, I haven't tried to lower vdimm at all yet.. seems to max out at about 39c after a few hours of TM5 Absolute..

View attachment 393703

My setup currently..

View attachment 393704
View attachment 393705

sub 40c temps running TM5 are incredible temps.

I hit like 55C and i'm at 1.4 on vdimm.

I have two 140MM fans pushing air out in my case so nothing blowing air on the ram.

 

[freeagent]

I am super happy with myself for getting this to pass TM5 over night. Voltage is kinda high.. but that's what it wants so who am I to say no..

 

[Sammoonryong]

My oldie at its peak exactly 1 year ago. I've eventually settled for lower CPU oc and lowered RAM frequency to 3400MT CL15-16-16-32 CR1. Less heat, less noise but still going strong

 

[matja]

My oldie at its peak exactly 1 year ago. I've eventually settled for lower CPU oc and lowered RAM frequency to 3400MT CL15-16-16-32 CR1. Less heat, less noise but still going strong

The power of quad channel!

 

[Mystic1978]

Here’s my 14900KF with dual-rank DDR4 sticks.

Long live DDR4, still delivering outstanding performance. All 13 global top scores in the 3DMark 5070 Ti category I achieved with DDR4 and a 5070 Ti on air cooling, running PCIe 4.0 x16.

DDR4 Is NOT Dead! 💀 13 World Records in 3DMark with 5070Ti on Air & PCIe 4.0

 

[FamousFighter]

Pretty expensive used b die but honestly I am very happy with it. I probably can get a little more out of this kit but this is an awesome improvement over my old kit.

Secondary rig got a B-die upgrade. I'll see if I can get the IF stable at 1900 MHz with the newest AGESA:

View attachment 376217

getting below 60ns on zen 2 is insane. awesome you were able to pull that off

 

[DoncicBR]

8600g with iGPU overclocked to 3150Mhz | 1250mV, and PBO +200Mhz boost | Curve Optimizer -30.

Suggested changes for better results? Tried some 1:1 6000Mhz CL30 configs but couldn't get stability. Hynix A chips 6800Mhz CL34

 

[MagnusTS]

Hynix M-die

 

[freeagent]

Pretty much just my EXPO profile with a couple of minor changes..

 

[freeagent]

Nifty..

 

[leghmoh]

 

[Athlonite]

View attachment 397842View attachment 397843

Why is your North bridge clock only at 1550MHz it should be more like 2067MHz

 

[maur0]

I have a question: does turning off all mitigation protections on the CPU cause any more memory bandwidth, or not?

whoever wants to test

download

https://github.com/ionuttbara/windows-defender-remover/releases/download/release_def_12_8_2/DefenderRemover.exe

run, select S disable all security mitigations

restart computer

on powershell run this command with admins level Get-ProcessMitigation -System

all options must be off or false


After that, test Aida64 normally. Will it give you any more speed in your memories without the CPU mitigations getting in the way a little bit of performance?

 

[ShrimpBrime]

I have a question: does turning off all mitigation protections on the CPU cause any more memory bandwidth, or not?

whoever wants to test

download

https://github.com/ionuttbara/windows-defender-remover/releases/download/release_def_12_8_2/DefenderRemover.exe

run, select S disable all security mitigations

restart computer

on powershell run this command with admins level Get-ProcessMitigation -System

all options must be off or false


After that, test Aida64 normally. Will it give you any more speed in your memories without the CPU mitigations getting in the way a little bit of performance?

No, this does not get you more bandwidth. You can tell what the bandwidth is by the frequency.

you can also calculate Cas latency without running aida64.

All this does is free up cpu cycles.

 

[freeagent]

You would be better off running in safe mode for the biggest numbers..

 

[Rootax]

Hello !

I build my 265k system last week, and I find than my L3 results are pretty low compared to reviews I see online. My ring is a 40x, D2D and NGU at 30x.

Now with 7.65 I see that L0 is taken into account, I don't know if this can affect the displayed results ? I'm in high performance mode during the bench, and the frequencies stay where they should be.

I've disabled virtualisation and all that... Any help would be appreciated

Thx !

 

[Rootax]

Hello !

I build my 265k system last week, and I find than my L3 results are pretty low compared to reviews I see online. My ring is a 40x, D2D and NGU at 30x.

Now with 7.65 I see that L0 is taken into account, I don't know if this can affect the displayed results ? I'm in high performance mode during the bench, and the frequencies stay where they should be.

I've disabled virtualisation and all that... Any help would be appreciated

Thx !

Little update, it was the trefi value. It was at 8k by default, I don't know why. I put it at 32k, and bandwidth and latency are more logical. I put a low rpm 12cm fan over the sticks to keep them under 50c during memtests, working great (it was going at 60 before, even with 8k trefi).

Happy !

 

[Athlonite]

Little update, it was the trefi value. It was at 8k by default, I don't know why. I put it at 32k, and bandwidth and latency are more logical. I put a low rpm 12cm fan over the sticks to keep them under 50c during memtests, working great (it was going at 60 before, even with 8k trefi).

Happy !

While you're playing with settings you should also adjust CR from 2 to 1 aswell

 

[Rootax]

While you're playing with settings you should also adjust CR from 2 to 1 aswell

I need to increase some timings for CR1 to be stable, so in the end, worse performances.

 

[Athlonite]

I need to increase some timings for CR1 to be stable, so in the end, worse performances.

Ah righto then