Hi peeps. Well I decided to re tune my RAM today since I had not do it for a while. The last time was last year when I got my "old" 3800X, which is now a 5800X. I was getting random crashes in game, and on the different RAM stability programs, which before were 100% stable. I did increase the frequency from 3733Mhz, to 3800Mhz and pretty much left everything as it was. I had the kit running at 1.46v running for over 1 year and never had an issue. After some testing and stuff, I had to dial the voltage to 1.48v. Would you guys say its okay for daily use? My PC info on my Signature. Thanks.
B-Die 1.5v for daily use
- Thread starter Nghtmare
- Start date
Yeah I have an O11 Dynamic with 6 intake fans and a 360 AIO for exhaust on the top. I just wasnt sure how reliable that is but a lot of companies now are bringing out kits running 1.5v out of the box so it should be okay.
The random crashes in games are probably cause by the RTX 3080 heating up the case, I used to run my ram happily at 3800/14 @1.49v but after installing the 3080 I noticed some instability from the ram while gaming so knocked it down to CL16 and dropped the voltage to 1.45 and it's been fine since.
This instability was never picked up in memtest as the GPU wasn't loaded it was only during games when the GPU was pumping out heat.
This instability was never picked up in memtest as the GPU wasn't loaded it was only during games when the GPU was pumping out heat.
I had those random crashes when I was running the 5700XT Red Devil I had in between selling the 2080ti and getting the 3080. My FTW runs at 60C under full load since I have it undervolted. The 5700XT was running at like 73C on Cyberpunk, and I was capping the FPS at 60. I can let the 3080 run wild and it doesnt heat up at all. Btw the Errors show up in memory benchmarks as well, after increasing the voltage to 1.48v they dissapeared.
I was getting the errors with the ram when running Karhu, or TestMem5 - Usmus V3. This was not the case in the past when running it at 3733Mhz, at 3800Mhz is when it started. I guess it needed some voltage since it was running very tight.
Last edited:
You mean the SOC voltage?
I think I am at 1.45v (need to check to confirm, might be bit lower).
I originally ran with no special cooling and was stable, but I feel better since I added the side case fan, the fan was added to cool my m.2 ssd, but is also blowing air over the ram as a bonus.
I originally ran with no special cooling and was stable, but I feel better since I added the side case fan, the fan was added to cool my m.2 ssd, but is also blowing air over the ram as a bonus.
Supposely its okay for BDie as long as you have decent cooling. Like you said, a lot of kits come 1.5v out of the box now days. The SOC is running at 1.15v, the board had it at 1.18v on Auto but I brought it down a bit.
I always had it at 1.1v, now I bumped it up a bit because I am trying to find the new sweet spot. The thing is that on the latest Asrock BIOS it made settings that were previously fine, unstable. The problem is that I HAVE to use the new versions if I want to use Ryzen 5000.
The IMC/CPU doesn't care at all what voltage is pumped into the RAM. As long as it gets the data from RAM.
The thing with B-Die and 1.5v XMP is it can get quite misleading. You are not fine tuning voltages besides primary. When you start to tune, the chips have to work harder so you are creating more heat and this is when issues can creep in.
I'm actually surprised no popular YouTube channel has dug deep into B-Die temperatures.
Good info. Does vdimm have a knock on effect to oc ability for ryzen? I get the impression ryzen is relatively limited in the sense that you've only got a finite amount of resource to play with for OC. IE, if you're pulling v for dimm, soc etc, is that limiting the resource available for cpu?
I wouldn't think so. Remember supplying voltages for dimm etc is the job of the motherboard. The CPU will take whatever it needs at such ridiculously fast intervals this should not be a problem. The only knock on effect I can think of is chip degradation from manually suppplying too much voltage to these areas.
Yes the latest BIOS for the Taichi was pants. I went back to 3.61 and found stable setting there.
I am using the latest bios with a 5950 and experiencing no issues with stability.
But then I always flash the bios by inserting usb stick into the flash slot on back of motherboard.
That entirely wipes any previous settings from memory so you have to enter everything from scratch.
Took photos of bios settings and copied them in and running same as before stability wise.
But then I always flash the bios by inserting usb stick into the flash slot on back of motherboard.
That entirely wipes any previous settings from memory so you have to enter everything from scratch.
Took photos of bios settings and copied them in and running same as before stability wise.
I've been monitoring this. Im at stock 1.5V on my G.Skill 3800 CL14 and the highest I've seen the Ram temp reach after 3 hours of gaming is 45C. I also have the O11D XL with both Rads as intakes so there is some active cooling over the Ram modules. In OCCT without active cooling they hit 50C in a matter of 3 minutes.
The Days of water-cooling Ram for aesthetics only might be over at this voltage though. Im seriously thinking it may be worth it.
I have pictures of my settingss as well, learned the hard way. It happened on my 3800X where I had a setting that was 100% stable, could pass over 8hrs of Karhu for example, and when I tried some new BIOS version (cant remember the number), the PC just wouldnt boot. Some other users on the Taichi thread were reporting problems with stability too.
I never tried using the Flashback thing to update, however doesn't the normal way wipe everything anyways?
Last edited: