3700X Memory Issue

[moproblems99]

So, today is not my day. The 3700X is running and I am trying to get the specs dialed in.

So I have this kit of mem: https://www.newegg.com/g-skill-32gb-288-pin-ddr4-sdram/p/N82E16820232861
3700X
X570 Taichi

I can't for life of me get it to boot on anything other than 3600 16 19 19 19 39 with IF uncoupled at 1600. Anything else gives me a 0x07 on DrDebug which appears to be a lock after microcode (and not in my manual).

Sigh...

 

[TheMadDutchDude]

07 typically occurs when there's a CPU error that can't allow it to go further during POST.

What are you trying to run it at? I would try running your IF coupled at 1800. Should be doable.

 

[moproblems99]

07 typically occurs when there's a CPU error that can't allow it to go further during POST.

What are you trying to run it at? I would try running your IF coupled at 1800. Should be doable.

Negative. It POSTS and then dies and I have to hard power down and clear cmos. If I don't run 3600 16 19 19 19 39 with IF uncoupled at 1600 then I pretty much can't boot.

 

[Xzibit]

Might have to wait for AsRock to improved compatibility. They only list the Royals for 2/4 and OC

 

[moproblems99]

Might have to wait for AsRock to improved compatibility. They only list the Royals for 2/4 and OC

They do also list Trident Z which is basically what these are.

While also superficial and likely not useful, my CPU scored terribly in CPUZ bench lol

 

[Xzibit]

They do also list Trident Z which is basically what these are.

While also superficial and likely not useful, my CPU scored terribly in CPUZ bench lol

Probably due to difference in RGB

 

[moproblems99]

Probably due to difference in RGB

Yeah....RGBs...for $50 less I can deal with them lol

 

[TheMadDutchDude]

What sort of voltages are you running on the CPU?

 

[moproblems99]

What sort of voltages are you running on the CPU?

Everything is on auto at the moment. I literally finished putting it together Monday night and just got some stuff installed on it yesterday. First day of tweaking was today...

Edit: So far the highest boost I have is 4245mhz. I am on the newest BIOS as well.

Welp, I now can no longer boot unless the memory is on 2133.

Well, I made it back to windows again. Enough for tonight

 

[eidairaman1]

@Mussels @TheLostSwede might be able to help

 

[TheLostSwede]

07 typically occurs when there's a CPU error that can't allow it to go further during POST.

What are you trying to run it at? I would try running your IF coupled at 1800. Should be doable.

That's not a CPU error, 07 is memory/memory controller related. It generally means that the memory controller doesn't accept the memory timings, but it might be a range of other issues from what I've been told by a mate at a motherboard manufacturer.

So far, from what I have seen, this has only been an issue with Gigabyte and MSI boards, but it seems now also ASRock. Most people with this issue have RMA:ed their CPUs in the end and that have solved the problem. It doesn't appear to be a board issue. Also, most people with this issue have had to do a fresh UEFI flash to get the boards to boot again after it happens.

What Agesa is the board on? If it's not on 1.0.0.3 ABBA, this could be part of the problem, so make sure the UEFI is updated to the latest version.

Also, get Thaiphoon burner and the DRAM Calculator to see what they spit out about your RAM.

From what I can tell, it's Hynix CJR, which should work fine with Ryzen 3000, at least based on my own experience, although my modules are from Patriot, not G.Skill.
Looks like they can run at the same timings as my modules, based on this review https://www.newegg.com/Product/SingleProductReview?ReviewID=10258805
As such, I can only guess two things, CPU or UEFI related problem.

 

[TheMadDutchDude]

ASUS has the D7 (or is it 07?) error too. I get it when pushing the IF too far. It still cannot recover itself, needs a manual reset.

 

[TheLostSwede]

ASUS has the D7 (or is it 07?) error too. I get it when pushing the IF too far. It still cannot recover itself, needs a manual reset.

Right, but you're getting it for an entirely different reason. A lot of people are seemingly getting the 07 error at random and the only way to recover for a lot of people, has been to use the USB UEFI flash button on the board and restore the UEFI from a USB key. This shouldn't be happening at all for one, but it has happened on at least Gigabyte and MSI boards from what I've seen. That you get a memory error that you can recover from by doing a clear CMOS, isn't a big deal, but when you can'r recover, it's a different matter altogether.

 

[TheMadDutchDude]

True. I saw the video from Jay about his MSI board failing POST...

What an absolute pain.

 

[TheLostSwede]

True. I saw the video from Jay about his MSI board failing POST...

What an absolute pain.

The problem is that there has been zero communication from AMD with regards to this issue, so it's not clear what causes it. Is it a flawed memory controller, something weird with certain RAM combinations, a poor UEFI/AGESA implementation, or a combination of all three? The two things that have solved it for some people have been to RMA the CPU and AGESA 1.0.0.3 ABB/ABBA. Swapping memory has in general not solved the problem.

 

[moproblems99]

What Agesa is the board on?

It is on 2.10 which should be 1.0.0.3 ABBA. At times last night, it would start up, and then shutdown three times before booting with a default bios.

 

[eidairaman1]

True. I saw the video from Jay about his MSI board failing POST...

What an absolute pain.

He had an Engineering Sample model.

 

[moproblems99]

I put in for a return so we'll see.

 

[eidairaman1]

I put in for a return so we'll see.

For what?

 

[TheMadDutchDude]

May have been ES, but it was very close to the retail sample. I’m not buying that it’s so massively different as ES vs retail. They only shipped them weeks before the retail samples... so they were in the final stages as ES boards.

 

[moproblems99]

For what?

3700x

 

[eidairaman1]

May have been ES, but it was very close to the retail sample. I’m not buying that it’s so massively different as ES vs retail. They only shipped them weeks before the retail samples... so they were in the final stages as ES boards.

It is, here you go

3700x

So you are returning the 3700X?

 

[Xzibit]

May have been ES, but it was very close to the retail sample. I’m not buying that it’s so massively different as ES vs retail. They only shipped them weeks before the retail samples... so they were in the final stages as ES boards.

Hardware Unboxed didn't have an issue with theirs

 

[TheMadDutchDude]

That’s my point. It’s extremely unlikely that it was radically different to the point that the BIOS would matter. It sounds like a BS response to cover their butts. That’s my take on it after having dealt with many a pre-release board and NEVER having an issue flashing with a newer BIOS.

 

[eidairaman1]

That’s my point. It’s extremely unlikely that it was radically different to the point that the BIOS would matter. It sounds like a BS response to cover their butts. That’s my take on it after having dealt with many a pre-release board and NEVER having an issue flashing with a newer BIOS.

Either way i dont use msi, they went down the toilet after AM2