• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

New 3080 and suddenly problems

"I do have PBO turned on though using motherboard limits, scalar 2x and boost 500Mhz"

that boost sounds very high, my advice is put PBO=Enabled but dont mess with Advanced settings like the curve, scalar and boost if you are having stability issues

Yes it's high. But the cpu will never boost that high before it hits one of the pbo limits.

It's not the CPU that's the problem it was my ram overclock. I've toned it down to cl16 and so far it's all good.

But it makes me wonder why it worked before and now after installing 3080 it doesn't.

Heat?

But if that's the case what of ram sticks like the 8 Pack cl14 3600 ram that is 1.45v out of the box?

They gonna overheat too? :confused:
 
tbh if its a P400a case with the fans cranked up it will blast the warm air out pretty damn quick

I think you need to turn PBO off and return your ram speeds back to their XMP profile, and keep the IF speeds sensible to see if that helps with stability first. If it does start to raise them back up one by one.

Switching to CL16 ram timings seems (finger crossed) to have solved it.
 
It may be something as simple as the 3080 running in PCIE 4.0 through lanes that are connected to the CPU and putting more stress on the Infinity Fabric/Data Fabric. If I ever get a 3080 I know that my 3600MHz RAM might become unstable with my 3900X as it's on the limits of stability right now (I had to adjust my VDDG and VDDP to get it stable but it's as stable as a rock now with my 1080).

You know what it's like though when you tinker. I keep wanting to put it back to CL14 and adjust more voltages. :D

Previously I left my VDDG and SOC on auto (it was fine)

But it's interesting you mention that because adjusting VDDG was something I was thinking of trying.

Ryzen Master documentation even says "when overclocking ram VDDG and SOC should typically start at 1.1v)

Have a look at page 39.

---

Bah... you now I have to try it. :p
 
I’m running my 3080 FE on an MSI X570 with 5800X and 32gb 3600 cl16 RAM powered by a Corsair RM750 psu. Rock solid with no issues hours into MSFS2020. This is with a Noctua air cooler on the cpu which helps pull warm air away from the RAM.

I was running my ram CL14 at 1.47v though.

At CL16 mine are now 1.35v.

But I keep asking this question... If I am getting what we imagine is temperature related instability on my ram, then where does that leave sticks like the CL14 8 Pack stuff that is 1.45v out of the box?

I've seen only one kit out there on the internet which is similar timings that is at 1.40v every other kit is 1.45v plus.
 
My Teamgroup 8pack CL14 3600 ram does what it says at 1.45v with its XMP profile in my system (x570 MEG Unify and a 5800X)

If you have 3200 CL16 ram set at 3600 CL14 and experiencing instability then.... sorry I don't mean to be mean but wtf?

Yup, on my x370 I had to use 1.1v VDDG to get it to work.

When I bought my x570 I thought I had to use the same settings on my ram so set it to 1.1v on the VDDG. What was stable on x370 was no longer working on x570. Leaving VDDG on stock did the trick.

So that's the same experience as me.
 
This is the RAM model https://www.corsair.com/ca/en/Categ.../Vengeance-PRO-RGB-Black/p/CMW32GX4M4C3200C14

Its rated for XMP 3200 CL14, but it's running stable at 3600 14-15-15-30

As mentioned it does get a little bit toasty while gaming (50c), it's due to the GPU dumping 410w of hot air into the chassis, on my previous GPU it would sit at 40c. But the extra 10c hasn't caused any instability for me.

never the less I am moving the GPU to water cooling soon and expect the RAM temp will come back down again

In my case I have been leaving the SOC on auto and just setting 1.47v dram.

Dunno what the default SOC is but I am tempted to try setting it to 1.1v and see if it behaves.
 
Right as an update to this.

Still experiencing issues.

My Watch Dogs Legion issues I 'fixed' by upping the dram voltage to 1.5v.

However getting freezes and crashes in Valhalla. Ask in the Valhalla thread and others say they get it too.

However also get freezes and crashes in Doom Eternal DLC 2 so now I am suspecting my overclock.

As previously stated, I was stable before the 3080 was installed using 1.47v dram for at least a year.

So now I am experimenting with SOC voltage in case it's the case that installing the 3080 has put more pressure on the memory controller perhaps? :confused:

It's not heat as my case is well ventilated and we are still experiencing winter conditions.

As for SOC I am gonna try 1.125v.
 
If it is heat then this should cause the frequency of people with ram such as the 8 Pack 3600 CL14 stuff to start complaining of instability as well and at XMP too as it's 1.45v out the box..

You see the issue I have with blaming heat. Where does it leave the ram out there that comes out the box at 1.45v +?

I wonder if this is an issue with the FE rather than the AIB cards.
 
OK, so yea dropping to CL16 timings and 1.35v seems to solve the issue.

So it's heat.

But this is strange all the same. During the initial reviews of the FE cards some of the big reviewers like Jayztwocents and GamersNexus et al tested the theory that the FE's blower design would heat up the ram and they concluded that this was not the case and that it was thermally efficient.

However as end users we are experiencing the opposite.

These cards heat up so much that they push ram in to thermally unstable territory!

And what about the people that buy ram such as the 8PackRipped stuff. It will be unstable at XMP with these cards unless you have active ram cooling.

Does it really matter? Not so much.. it's just annoying to have to downscale my ram due to GPU heat.
 
They were probably just testing ram set to XMP though rather than manually tuned to the limit, while the 8 pack ripped runs at 1.45v it's also not configured as tight as manually tuned especially in the sub timings which can increase the heat of the memory chips.

I punched in the same primary timings on my ram at 1.45v and it still froze.

14 15 15 15 35 @ 1.45v

Remember this same ram prior to the 3080 was running the above and it could even run 14 15 14 14 28 @ 1.47v (with tight secondaries) which is what I ran it at for quite a while.

---

Well actually, probably not a good test, because I know that my ram needs at least 1.47v for stability. So perhaps I should try again.
 
Back
Top Bottom