Share your AIDA 64 cache and memory benchmark here

[freeagent]

I recently sold my OCZ Reaper HPC’s. They were 800 6-8-6 and could only manage 1000 7-9-7.. I think.. Those are waay better lol..

Edit:

I may be been too gentle with vdimm

 

[freeagent]

Some crazy 2T action

Edit:

But its stable with 1.45v.. not bad..

 

[tabascosauz]

@freeagent that 5.15GHz single core really stepping in to save the day lol, I'm very jealous of your CPU

Viable at 1.5V again, thanks solely to 2 x A14 2000rpm push-pull on the C14S providing testbench-like temps for the DIMMs the underhanging A14 literally rests its rubber corners on the tops of the Trident Z heatspreaders

 

[DR4G00N]

I recently sold my OCZ Reaper HPC’s. They were 800 6-8-6 and could only manage 1000 7-9-7.. I think.. Those are waay better lol..

Edit:

I may be been too gentle with vdimm

Always just comes down to luck, even with the right IC's and PCB they could still be trash. I have some Mushkin 1600 7-9-7 sticks with Elpida BDBG on KO-8117 PCB's, on paper they look promising but they're actually meh.

Or you could have great IC's on crap PCB's and they end up trash.

 

[freeagent]

@freeagent that 5.15GHz single core really stepping in to save the day lol, I'm very jealous of your CPU

View attachment 210267


Viable at 1.5V again, thanks solely to 2 x A14 2000rpm push-pull on the C14S providing testbench-like temps for the DIMMs the underhanging A14 literally rests its rubber corners on the tops of the Trident Z heatspreaders

I would like to say its my tuning but that would be a lie

 

[9_cat_black_6]

This is my poor R5 2600 with dual rank bad B-die. Stock with XMP vs OC as tight as I can get it.

 

[HD64G]

Stock 2600X limited@95W with my poor Hynix CJR (sadly not stable enough to keep it for daily use)

 

[dgianstefani]

GSkill 3800/14 B die SR modules, tuned.
OCCT stable. 1.6v

Got the 1050ti in there as the 3070 was sold, and waiting until my waterblock for the 3080 arrives

 

[freeagent]

So it’s ok to run vddgs and vddp that high? Shit.. I want to try but I am 14 hours from home.. should be home Tuesday..

 

[DR4G00N]

Got a couple i5 750's to try, IMC is way faster than the Clarkdale chips. Board also worked much easier with Lynnfield.
Couldn't get it to do 2150 stable though no matter what, probably something sub-timing related or the cpu can't do 215 bclk stable, tried lower mem speed but same result.

Edit: Here's one on 1366 for comparison, the 920 ran out of uncore (this is at 1.55 or 1.6V VTT). Could probably do better with a Xeon instead.

 

[freeagent]

Stock "ish"

 

[tabascosauz]

So much for it being a "shittier 5600X", eh?

That's a 4-ish nanosecond gain from Renoir apples-to-apples. Did lose a bit of copy compared to Renoir though. Seems to be a consistent thing, might just be a Zen 3 quirk or part of their power optimizations for IF on Cezanne.

 

[freeagent]

What a beaut!

Here’s my shitty 5600X for comparison

Don’t forget to turn spread spectrum off

 

[glnn_23]

Ryzen 5700g

 

[tabascosauz]

Pushing the daily envelope a little further before I move to the MSI B550 ITX

edit: did 5600G make my B-die automagically better?? Less VDIMM needed now

4266 and 4333 16-16-16 verified iGPU stable, but setting 4400 instantly nuked entire BIOS and needed to reflash

 

[glnn_23]

Might be at the limit here

 

[tabascosauz]

Probably going to be the final daily settings achievable on this board.

@glnn_23 what GPU are you using for benching? thinking of putting in either my 5450 or GT610 but both being EOL seem like will be a driver nightmare

 

[glnn_23]

Mate I just use an old Asus GTX560.

 

[Coronajabona]

Hynix CJR

 

[Woomack]

5000 1:1 on Samsung B / quick test and most timings at auto so could be better
Micron is much easier to stabilize, just results are worse.

For comparison Micron / Ballistix 2x16GB DDR4-3600 CL16 @5000 CL18 1:1, also no tweaking, just basic settings

 

[Simkin]

Intel i7-5960X with Corsair Dominator Torque Special Edition 4x8GB b-die @ 3200Mhz 14-14-14-34 2T 1.35V (3200 CL14 Kit)

AMD Ryzen 7 5700G with G.Skill Ripjaws V 2x8GB b-die @ 3866Mhz 18-18-18-38-56 1T (1:1) 1.36V (3600 CL16 kit)

 

[Mbellantoni]

5600x g skill neos 3600cl16 8gb x2 @ 4000cl 14-15-14-21 "ram stable" at 1.58.5v but i get whea 19's randomly. Went back down to 3800/1900. To stop whea 19's ..same timings but 1.56.5v stable. I dont have a pic but with the 3800 profile im at around 57.5k mb/s read and my latency is 54.8 on average. About a 5% loss of performance in high fps 1080p gaming compared to the 4000 profile.

 

[Simkin]

Up from 3866Mhz 18-18-18-38-56 to 4000Mhz 18-19-19-39-56, FCLK raised to 2000Mhz, so running 1:1. Still running 1.36VDIM, not touched any other voltages, or sub-timings. Not gonna chase any big numbers here, its my HTPC only.

But, my write speed is always alot lower then read, have tested 3600 CL16, 3733 CL16 and now 4000 CL18. From other benchmarks in here of users with 4000/5000 APU they are pretty much equal. B550M (mATX motherboard) Also my 3600Mhz 16-16-16-36 L1 Cache numbers was 1980-1068-2107, so higher than it is now, but i guess its normal as the timings where tighter?

 

[tabascosauz]

@Simkin post up a Zentimings screenshot. The latency looks about right for 18-19-19, but your bandwidth is way down at least on copy where I'd expect 3600CL16 to be. I can understand a 1-2GB/s loss due to looser timings, but not 5GB/s. At 4000 16-16-16 on the 4650G:

I did consistently lose a tiny bit of read and copy on evey profile going from 4650G to 5600G and 5700G. But overall Cezanne enjoys ~3ns gain, and a gain in membench as well. AIDA is so fickle I generally only run it once and move onto membench

 

[bissag]

What be the reason for not improving the latency between c14 and c16