[Dekaohtoura]

Whatever I may try, I can't get my Corsair Vengeance LPX 3000CL15 (CMK16GX4M2B3000C15, Hynix, AFR) to run stable even at 2866CL16 on a Taichi X370 (latest BIOS) with a R1700.

Tried almost 20 different combos, tried 2866 with 3000 safe, tried upping Vsoc, DRAM volt, nothing works.

Really really frustrated...

4.81 is public?

 

[Dekaohtoura]

L4.81 is on the support page at the bottom under Beta BIOS.

Yup, found it...TY.

Tried it, and the dam.ed thing worked...

I don't know if these helped as well (found them in various posts around here):

Global c-states=Enabled
Disable Core Performance Boost
Mem Integer=Channel
Integer size=512
Mesh=Enable

Don't know about stability, don't know if it's going to cold boot, don't know what I'm doing...honestly.

Waiting for the extra mounting kit for my D15 to arrive and I'll start messing with core oc (nothing fancy...anything around 3800 will be great).

Hope to have good news soon.

 

[Dekaohtoura]

@Dekaohtoura

Could you please post your current settings - i guess you are running 2666MT? A screenshot from Stilt's Ryzen Timings Checker would be perfect. It would also help if you'd post a screenshot of Thaiphoon Burner http://www.softnology.biz/files.html

I managed to run dual-rank Hynix LPX at 2933MT (MFR) and had really tough time with anything above 2666MT. I suspect that you might be bit unlucky and actually having a dual-rank kit as well.

Here you are:

Thaiphoons

https://www.overclock.net/forum/attachment.php?attachmentid=224478&thumb=1
https://www.overclock.net/forum/attachment.php?attachmentid=224476&thumb=1
https://www.overclock.net/forum/attachment.php?attachmentid=224480&thumb=1

And my current working settings

https://www.overclock.net/forum/attachment.php?attachmentid=224482&thumb=1

 

[Dekaohtoura]

Got a vengeance as well, 3200 hynix afr but led model. Single rank, and before 4.81 could never run it above 29XX, anything above this would crash. When I flashed 4.81 I could set 3133 and is running with no problems, but still not able to run 3200 in any way, no success using calculator, increasing SOC, vram, etc..

Sent from my HTC U11 using Tapatalk

Ok, got it (prays silently for BIOS v5 with newer AGESA).

Just a clarification, though:

Did you load the XMP profile and chose freq. 3133 on the Asrock oc mode or the AMD one?

 

[Dekaohtoura]

Just XMP and chosen 3133 nothing else voltage 1.35v, soc auto. What do you mean with AMD one? Ryzen Master thru Windows? If yes, looks like there's a way to set 3200mhz using it, I've seen it few pages ago here, gonna try, but I dont like doing it from Windows..

I also tried XMP with higher frequencies, increasing soc, dramvoltage etc, but no luck.

So far from I've searched, these vengeance led sticks I've got are meant to work better with intel..

Sent from my HTC U11 using Tapatalk

I mean on the BIOS OC page, there are two options:

1) Asrock OC, that let's you choose XMP profile etc, and

2) AMD OC that doesn't...you can set everything manually though.

 

[Dekaohtoura]

Asrock one then. the other is amd cbs? well, that one does not allow me to set ram clock for example. maybe thru the advanced meny only, but never tried.

Sent from my HTC U11 using Tapatalk

Yes, AMD CBS.

Well, on Asrock mode, best I can do is 16-17-17-35-52-1T (even when choosing the XMP-3000 profile 15-17-17-32-50...it boots @16).

I'm not sure if I'm stable, but at least it (kinda) works.

On 4.80, I couldn't get even to 2800-16...

 

[Dekaohtoura]

we have a formidable tool like the ryzen dram calculator to setup ram manually, leave the "user friendly oc tool" where they deserve to be: disabled.

Already tried it...to no avail...exactly like almost anyone with a X370 Taichi+4.80 BIOS.

That's what I mean by "safe" and "combos" at the previous page.

 

[Dekaohtoura]

Ok... it seems that I was mistaken and you are running single-ranks after all... Could you please tell me how does the board fail to boot? On which code it stops or loops?

Anyway, I would say that you are limited by CPU's IMC. My dual rank kit behaved that way as well, no matter how loose the timings were.

Did you try disabling GearDownMode and upping CR to 2T? Ultimately, this is what allowed me to reach 2933MT.

Np.

Can't say that I've checked error codes or anything. Next time that I find some time, I'll set it to the default 3000 profile and see what will come up.

Geardown on/off doesn't change much...ctl15 is out of the question either way.

CR=2 seems to help upon selecting restart (cr1T causes the pc to shutdown, cr2T let's it restart normally).

edit:

Scratch that last point about CR2 an Restart...sometimes it will, sometimes it won't

 

[Dekaohtoura]

I would drop CL15 or other odd values, try CL16, drop to CR=2 and disable GDM. You should do that on lower frequency, make sure the values are applied and try 2933MT afterwards.

I'm already using cl16 (15 is a no-go), GDM on or off, CR2 or 1 doesn't make a difference.

The system seems to be stable...can do CB, Blender, IBT-IVX, hours of gaming, different benches...everything except TM5 (error/crash message upon running it).

Sometimes, without any obvious reason, the system won't restart, it will just shut down.

I honestly can't understand.

I'll try dropping to 2933 and see where this will get me.

 

[Dekaohtoura]

I have CMK32GX4M2B3000C15, MFR, dual rank, nearly similar to yours, check out my timings, username cx5 in 1usmus spreadsheet statistics row 126, 127, 128, 132, 176. Each has a link at far right comments with screen capture of timing, voltages, speed, etc.

now running 2933, ram_1.33V, SOC_0.95V, 14,15,15,15,then low low secondary timings.

Actually, I've been stable(ish), using 3000-16-17-17-35-52-1T, at 1.38V and 1.1V VSoC.

I'll see what I can get from your stats (TY!) and maybe experiment a bit.

 

[Dekaohtoura]

Sorry to bother you with another non-oc (I think, at least) problem, but...

For the last 3 days, whenever I cold boot, the SSD/HDD queue is messed up and the board can't find the OS disk.

Upon entering BIOS, everything else is as it should (oc, mem, fan profiles etc), but all the drives are on a random priority.

Set them right, save+exit, all is fine, until next cold boot (but not if I shut the system down and reboot it after 2-3min...some hours must pass).

Any ideas? Could it be the mobo battery failing? But why does everything else remains as it should?

TY.

 

[Dekaohtoura]

try setting your system again, format the hard disk and install windows, but with only the system drive connected to the board and don't connect the hard drives until windows in full installed with all the drivers, the Chipset driver and everything

Thanks for your suggestion, but it's a bit excessive.

I rearranged and reconnected the drives to the mobo, and the OS ssd is now on SATA3_1 (was on 6, before).

Hopefully this will fix the problem.

edit:

Btw, I tried the AMD OC mode (CBS), and was surprised to find out that it seems to work.

When I first set up the board (about a month ago), using the 4.80 BIOS, it was an utter and complete mess, completely unstable and didn't import anything over from the Asrock mode.

After I've flashed the 4.81, I didn't have a chance to test it.

Today, I just saved my Asrock mode profile and changed to the AMD one...everything was already there, neatly set up, and working.

Performance wise, I don't think there is something notable (same latency, same mem b/w, same bench scores), but I'll give it a try and see where this will get me.

Any help and tips, will be much appreciated.

 

[Dekaohtoura]

Thanks for your suggestion, but it's a bit excessive.

I rearranged and reconnected the drives to the mobo, and the OS ssd is now on SATA3_1 (was on 6, before).

Hopefully this will fix the problem.

edit:

Btw, I tried the AMD OC mode (CBS), and was surprised to find out that it seems to work.

When I first set up the board (about a month ago), using the 4.80 BIOS, it was an utter and complete mess, completely unstable and didn't import anything over from the Asrock mode.

After I've flashed the 4.81, I didn't have a chance to test it.

Today, I just saved my Asrock mode profile and changed to the AMD one...everything was already there, neatly set up, and working.

Performance wise, I don't think there is something notable (same latency, same mem b/w, same bench scores), but I'll give it a try and see where this will get me.

Any help and tips, will be much appreciated.

Nope...instability all over everything, actually.

Back to Asrock oc mode.

 

[Dekaohtoura]

SSD/HDD/Optical configuration on the BIOS, has nothing to do with Windows, Linux etc.

The BIOS itself rearranged my configuration (for no apparent reason) and the OS couldn't find the boot disk. BIOS...prior to OS.

After I moved my OS disk to SATA_1, the problem was (re)solved.

Now, the part of me trying to use the AMD oc mode, is a completely different issue. It has to do with a post that I found on this thread, where it was stated than Asrock mode resulted in a 13-14% mem b/w loss (also, I wanted to experiment a bit). I compared both modes using AIDA64 ,and there was no difference at all (less than 1%), so something has changed since this post was written.

I didn't try to import anything, btw. It was automatically done, by the BIOS (as it should, imo). Even then, I checked every single value with my notes from the Asrock mode, and they were identical.

So, using exactly the same settings/values, should (in theory) produce the same results. Unfortunately, that was not the case, and I experienced multiple instabilities when I benched my system (IBT-AVX, 3dMark etc) or when gaming. Reverting back to Asrock mode, completely eradicated any issue.

Have a nice weekend.

 

[Dekaohtoura]

Any ideas when we'll get AGESA 1.0.0.6+?

Was about to ask the same question.

ASUS, MSI, have already started updating their series (admittedly, from top to bottom, latest to older) and here is an Asrock rep reply, just some days ago:

" 1) Is there a new bios with a new agesa code to come for my motherboard?

2) Dear *reducted*,

I'm not aware of any new BIOS with different Agesa code, why are you asking?

A new Agesa code must come from AMD, then it needs to be properly tested, and in most cases we need to write a complete new BIOS.

3) Noted. I just read about a beta bios for an ASUS CH7 with agesa 1.0.0.6 which debuted a month ago on their site, but taken down from it a coupple days later.

This is why i asked about it, since my Taichi is a high end product for the X470 chipset.

4) If Asus has taken it down, it most likely is because there was an error in the Agesa code. We have also taken down BIOS releases because of buggy AMD Agesa codes, or Intel Microcodes.

We test the BIOS thoroughly, but sometimes customers report problems, and the BIOS is taken offline. It can also be that this Agesa code was supposed to fix a specific Asus problem, or we just tested better and didn't release a new BIOS because it failed the test. BIOS versions taken offline usually mean that they cause unexpected side effects, or in some cases the system doesn't even start.

With the first Spectre/Meltdown Microcode updates Intel has sent us, some system crashed, and it was impossible to reflash the BIOS, within a day all manufacturers took it offline.

Keep in mind that BIOS updates are not made because manufacturers have nothing better to do, it either solves an issue, or allows new CPU's to run, and it takes a lot of testing and time, before a BIOS is released.

When there is a new Agesa code, we will also publish a new BIOS, but it will probably be when a new CPU is launched that is unsupported with the current BIOS."

Source: https://www.insomnia.gr/forums/topic/689836-ryzen-7-2700x-thread/?do=findComment&comment=56871046 (it's a Greek tech forum/community)

 

[Dekaohtoura]

Gigabyte released AGESA 1006 BIOSs for their X370 lineup.

MSI released AGESA 1006 BIOSs for their X370 lineup.

ASUS, doesn't specify if their X370 releases include the latest AGESA, but their X470 do.

Asrock..."released" only a beta BIOS for their flagships, through an indirect source...and claims there is no reason at all to update a 1.5 year old 250€ motherboard...

Great!

 

[Dekaohtoura]

Yeah don't expect much until the 3000 series Ryzens drop. IMO 4.40 is the best and that came out prior to the 2700X launch last April. It has the scalar option that has been missing ever since. Hopefully we can drop the 3700X and have full performance and only have to miss out on the pcie 4.0 which isn't a big deal unless you have alot more IO you want to add like more NVMEs @ x4 If not then it looks like I'll have to buy an X570 and hook some family member up with an upgrade, lol

edit: the X470 Taichi bios release/ support hasn't been anything to write home about either.

TY, I may give 4.40 a try, though anything below 4.81 doesn't like my ram very much, and I can't go over 2933, no matter what I try (now 3000 16-17-17).

Let's hope they'll release a proper and properly functioning BIOS for the 3XXX series, then.

 

[Dekaohtoura]

I am one of the lucky ones with Samsung B-dies. Most of the memory improvements from that later AGESA /bios are for Hynix. If 4.81 allows you the best memory performance then you should probably stick with that. 4.81 is one of the few bios revisions that I have read good things about in a long time in this tread.

"Unfortunately", Hynix owner here, so 4.81 seems to be a "must" for the time being.

I've send my "thoughts" to Asrock support and wait for their response.

 

[Dekaohtoura]

So, here is Asrock's "support" reply to my inquiry regarding the latest AGESA release...

"Ηello,

Last Summit PI AGESA version was 1.0.0.6b.

Latest Pinnacle PI AGESA version that I know of is PinnaclePI-AM4_1.0.0.4 Patch C.

Despite the lower number, this is a newer AGESA version than the 1.0.0.6b.

I will ask a colleague in Taiwan if a newer AGESA version has been released by AMD.

Kind regards,

ASRock Support

ASRock Europe B.V."

My response was not very kind (polite, though ironic).

I sent them the specific links of various MSI ang Giga BIOS releases, with the newer code.

I'm curious about their reply.

 

[Dekaohtoura]

So, here is Asrock's "support" reply to my inquiry regarding the latest AGESA release...

"Ηello,

Last Summit PI AGESA version was 1.0.0.6b.

Latest Pinnacle PI AGESA version that I know of is PinnaclePI-AM4_1.0.0.4 Patch C.

Despite the lower number, this is a newer AGESA version than the 1.0.0.6b.

I will ask a colleague in Taiwan if a newer AGESA version has been released by AMD.

Kind regards,

ASRock Support

ASRock Europe B.V."

My response was not very kind (polite, though ironic).

I sent them the specific links of various MSI ang Giga BIOS releases, with the newer code.

I'm curious about their reply.

Update, after my reply...

"Hello,

Thank you for the update. Actually I did have a quick look at a high-end model on Asus website, but that did not show detailed info about the AGESA version. I did not check MSI and Gigabyte website, so thank you for the links.

My colleague in Taiwan has confirmed that our BIOS team is working on new version with the new AGESA code. I will use your mail to put more pressure on it. But it might still take a while for before the BIOS is ready.

Kind regards,

ASRock Support"

I guess we'll wait and see.