8 PACK MEMORY RANGE GROWING: SAY HELLO TO 8 PACK RIPPED EDITION & 32GB KITS!!!

michty_me said:
I have a 5800X with a 3080FE.
Audio setup is a G6 (Used as a DAC feeding a Schiit Asgard 3 amplifier)
I'm not sure, I'll need to carry out a benchmark test when I get home. I'm due home this week so I'll download and run that as a comparison.
It must be a similar setup to you then just different RAM.
Click to expand...
Yeh, you have my exact setup by the sounds of it apart from diferent RAM!

I generally dont care for super high FPS as have an Acer X34 100Hz screen and run 3440 x 1440. I lock vsync & gsync on and cap FPS in NCP to 97. It's the minimums the certainly count when gaming
 
Guest2 said:
Yeh, you have my exact setup by the sounds of it apart from diferent RAM!

I generally dont care for super high FPS as have an Acer X34 100Hz screen and run 3440 x 1440. I lock vsync & gsync on and cap FPS in NCP to 97. It's the minimums the certainly count when gaming
Click to expand...

I'll try the Unigine benchtest and report back. Both having very similar setups will give a decent indicator on the RAM differences between 3733 CL16 and 3800 CL14.
I'll hopefully report back by the end of the week.
 
michty_me said:
I'll try the Unigine benchtest and report back. Both having very similar setups will give a decent indicator on the RAM differences between 3733 CL16 and 3800 CL14.
I'll hopefully report back by the end of the week.
Click to expand...
Thanks, please do. Might think up some other good benchmarks to compare minimums
Are you running 1866.50 FCLK currently?
 
Guest2 said:
Thanks, please do. Might think up some other good benchmarks to compare minimums
Are you running 1866.50 FCLK currently?
Click to expand...

No problem. It would be good to compare actually. See what the differences between the two are.
I am currently running at 1866 yes. I tried putting it to 3800/1900 and booted fine, Ran for a good while. About 20 minutes into a game the system restarted so obviously not stable. Likely needs a voltage increase but haven't looked into it further and left it running 3733/1866.
 
michty_me said:
No problem. It would be good to compare actually. See what the differences between the two are.
I am currently running at 1866 yes. I tried putting it to 3800/1900 and booted fine, Ran for a good while. About 20 minutes into a game the system restarted so obviously not stable. Likely needs a voltage increase but haven't looked into it further and left it running 3733/1866.
Click to expand...
Best keeping it 1:1 with memory anyway.

DRAM voltage is a fine line ive found.

There are other setting too i’ve got set like vsoc, vddp,vddg iod, vddp ccd, northbridge llc and nb switching frequency.
 
Guest2 said:
Best keeping it 1:1 with memory anyway.

DRAM voltage is a fine line ive found.

There are other setting too i’ve got set like vsoc, vddp,vddg iod, vddp ccd, northbridge llc and nb switching frequency.
Click to expand...
Yup, Was there a bug with certain AGESA where it wouldn't like posting above certain frequencies?
I think Micron Rev E can be a bit iffy with voltages but I'll be reading up on it before changing anything further.
 
michty_me said:
Yup, Was there a bug with certain AGESA where it wouldn't like posting above certain frequencies?
I think Micron Rev E can be a bit iffy with voltages but I'll be reading up on it before changing anything further.
Click to expand...
For the most part i've had no issues with BIOS updates. Even the first BIOS and beta were fine. The most annoying thing is re-doing all BIOS settings like timings again.

The new Tomb Raider game (Shadow of the tomb raider) has a built in benchmark and is included in the demo/trial, that might be a good one if it shows minimum fps instead of just average. I'll give it a try and see
 
Guest2 said:
For the most part i've had no issues with BIOS updates. Even the first BIOS and beta were fine. The most annoying thing is re-doing all BIOS settings like timings again.

The new Tomb Raider game (Shadow of the tomb raider) has a built in benchmark and is included in the demo/trial, that might be a good one if it shows minimum fps instead of just average. I'll give it a try and see
Click to expand...

Yeah, I agree with that. It does get a bit time consuming which is why I rarely update the BIOS much if everything I need is working fine.
I have seen the inbuilt benchmarks appearing in some games now. I think I used CapFrameX last time when I was checking minimum, 1% lows etc. It seems pretty good.
 
Updated timings and voltages for anyone looking at going 3600C16 > 3800C14

I had to increase VSOC, VDDP, CCD & IOD voltages as Karhu was failing with an error. I expect it was stressing the IF or CPU too much.
I also disabled GDM and set T1 with AddrCmdSetup set to 56

Also on latest AGESA (1.2.0.5)

QSfnrRw.png

@MrPils I did try tweaking some timings further, but hit a black screen and errors when testing with much lower. I suppose I could add 0.50V but 1.55V on the RAM is pushing it a bit probably and would probably cause errors anyway.

So far so good with these timings. I'll run Karhu overnight tonight.
Edit.
Karhu stopped with an error after even shorter time with the increased VSOC, VDDP, CCD and IOD voltages.

I wonder what it is that's making it error in Karhu.
RAM seems fine with other tests like TestMem 5 with extreme anta config, realbench is fine, memtest85 fine. No crashes from desktop or gaming for over a year. Maybe Karhu doesnt like CPU NB LLC on mode 1 or the CPU NB switching frequency that's on 1000KHz (or requires more/less) I do have it set to test CPU cache too (CPU cache enable in advanced settings)

I now have a stable Karhu (well, 8.5 hours with CPU cache enabled)
I set NB LLC from mode 1 to mode 3
This results in VSOC 1.1750V droop to 1.1688V

Oddly I did try increasing VSOC to 1.2000, IOD to 1.1500, VDDP and CCD to 1.000. It seemed stable, no black screens or errors but Karhu failed after 8 mins.

These settings with mode 3 NB LLC seem to work
huFeiYd.png
 
Last edited:
michty_me said:
@Guest2
As promised, Here are a few tests carried out this morning

QZYhhwc.jpg
Click to expand...


Did you manually set your own timings and voltages?

tRFC 580 is quite high. For 3800Mhz the tRFC range is usually between 200 and 250

It's the multiple of tRC (58 x 10)

Have a read through MrPils advice on this page and pages around 47 (im just quoting him!)
https://forums.overclockers.co.uk/t...ack-ripped-edition-32gb-kits.18885405/page-47

As to the benefits of getting newer RAM, the latency would come down by maybe 10ns.
Whether this shows in your daily / general use though I suppose it depends on what you use your machine for.
Heaven 4 looks about right, minimums might come up 10-15fps but this wouldnt show in newer games as much.

Having said that, a friend of mine also has a 3080 FE, 32GB RAM and a 3900X CPU.
Playing Warzone and now Warzone Caldera in 2560 x 1440, he says his FPS fluctuates usually from 70 - 150 uncapped. I lock mine at 97 for my Gsync 100Hz monitor and it only ever drops below 97fps on the plane scene at the start at the start of a game.
I also play on a 3440 x 1440 monitor so either his system is knackered or the RAM and CPU are seriously helping me. I know I have a 5800X and he has a 3900X but the 3900X is still a great CPU with 4 more cores than the 5800X. Turbo on the 5800X is a bit higher at 4850MHz where the 3900X is 4600MHz
 
Guest2 said:
Did you manually set your own timings and voltages?

tRFC 580 is quite high. For 3800Mhz the tRFC range is usually between 200 and 250

It's the multiple of tRC (58 x 10)

Have a read through MrPils advice on this page and pages around 47 (im just quoting him!)
https://forums.overclockers.co.uk/t...ack-ripped-edition-32gb-kits.18885405/page-47

As to the benefits of getting newer RAM, the latency would come down by maybe 10ns.
Whether this shows in your daily / general use though I suppose it depends on what you use your machine for.
Heaven 4 looks about right, minimums might come up 10-15fps but this wouldnt show in newer games as much.

Having said that, a friend of mine also has a 3080 FE, 32GB RAM and a 3900X CPU.
Playing Warzone and now Warzone Caldera in 2560 x 1440, he says his FPS fluctuates usually from 70 - 150 uncapped. I lock mine at 97 for my Gsync 100Hz monitor and it only ever drops below 97fps on the plane scene at the start at the start of a game.
I also play on a 3440 x 1440 monitor so either his system is knackered or the RAM and CPU are seriously helping me. I know I have a 5800X and he has a 3900X but the 3900X is still a great CPU with 4 more cores than the 5800X. Turbo on the 5800X is a bit higher at 4850MHz where the 3900X is 4600MHz
Click to expand...

Yes, All set manually. From what I remember, Micron Rev E can't go very low on tRFC using dual rank dimms which is why its that. I'll need to double check that of course.
I haven't played on this machine for a number of weeks now but will give it a try this afternoon and see what sort of fps I am geting. I don't cap mine though.
I've also set my settings similar to a Youtube video so not running on max settings.
 
Got a set of the 8Pack 32GB when it was cheap but decided to push my existing B-die a bit harder.

I've got Corsair 4x16G modules 3466mhz CL16 - normally run only 2 sticks as 4 is such a pain to get working.

Just tried pushing IF to 1900mhz, RAM at 3800mhz CL16 and it seems stable with Memtest so far.
 
Minstadave said:
Got a set of the 8Pack 32GB when it was cheap but decided to push my existing B-die a bit harder.

I've got Corsair 4x16G modules 3466mhz CL16 - normally run only 2 sticks as 4 is such a pain to get working.

Just tried pushing IF to 1900mhz, RAM at 3800mhz CL16 and it seems stable with Memtest so far.
Click to expand...
IF increasss stresses the CPU and FPU too.
Memtest86 is also easy enough to pass.
Try TestMem 5 with the extreme anta config to test RAM only.
Karhu with cache enabled will test IF increases. AIDA stress testing will too
 
I think I may buy that Dominator kit though for something to tinker with in the winter days.
Just found someone on Reddit running it at 4400 15-15-15-15-258 @ 1.6V (Not that I would ever need that or get it on Ryzen) but I would imagine it would have to be a good bin.

Edit: The person said they had it running fairly easily 4000Mhz CL14-13-13-28-238-2T (Gear 1)
 
Last edited:
michty_me said:
I think I may buy that Dominator kit though for something to tinker with in the winter days.
Just found someone on Reddit running it at 4400 15-15-15-15-258 @ 1.6V (Not that I would ever need that or get it on Ryzen) but I would imagine it would have to be a good bin.

Edit: The person said they had it running fairly easily 4000Mhz CL14-13-13-28-238-2T (Gear 1)
Click to expand...
Interesting, do you have the reddit post link? Be good to see what the results are including the full timings & voltages
 
Minstadave said:
Ok I gave in and installed the 8 pack kit.

CL14 at 1900mhz IF/3800mhz RAM - very nice.

Can't boot without enabling GDM though, is there much of a penalty from that?
Click to expand...
Nice work :)
Have you ran Ryzen DRAM bench, AIDA 64 cache & memory bench other stability tests?
What voltages & other BIOS settings are you running?
AddrCMDSetup can be changed to allow GDM to be turned off. There is a small penalty but not as much as it dropping to 2T under GDM
 
Last edited:
You must log in or register to reply here.
Back
Top Bottom