Ryzen Owners Zen Garden

[AusWolf]

Didn't you say JEDEC was failing TM5? maybe add a bit of VDIMM at 4800 and see what happens. Bit more VSOC probably helps too since you have faster UCLK at JEDEC than expo lol

The Gene is looking so sad, what an absolute unit of a OC board just sitting around collecting dust while the Raphael UMC trips over itself as soon as it touches DDR5. Meanwhile 13th gen can do it, but doesn't have a decent matx board to do it on

You are right of course, it's just that I paid so much on shipping to RMA the B550I Aorus AX (twice!) that I just don't want to do it again

Yes. JEDEC is failing TM5 and reboots randomly, while EXPO fails to boot on first try when cold, but is stable otherwise.

I don't think the RAM has any issues running at JEDEC - why would it? It's a 6000 kit, after all - but a bit more SoC voltage could help. Or maybe switching back to 1:2 mode and 1200 MHz UCLK.

I don't see the Gene at UK stores. If I decide to ditch this crappy thing, my options will be the ASRock Riptide, the Asus TUF, or the Gigabyte Aorus Elite. The Riptide seems suspiciously cheap, and I'm not a big fan of Gigabyte, so I'm left with the TUF as the only choice. There's also the Asus Prime and Gigabyte Gaming X, but they both look even more poorly equipped than my current MSi Pro.

I know that feeling with the RMA. I RMAd two MSi boards last year (they were the same model, though)... that's why I can't comprehend why I trusted this joke of a brand again.

 

[tabascosauz]

Yes. JEDEC is failing TM5 and reboots randomly, while EXPO fails to boot on first try when cold, but is stable otherwise.

I don't think the RAM has any issues running at JEDEC - why would it? It's a 6000 kit, after all - but a bit more SoC voltage could help. Or maybe switching back to 1:2 mode and 1200 MHz UCLK.

I don't see the Gene at UK stores. If I decide to ditch this crappy thing, my options will be the ASRock Riptide, the Asus TUF, or the Gigabyte Aorus Elite. The Riptide seems suspiciously cheap, and I'm not a big fan of Gigabyte, so I'm left with the TUF as the only choice. There's also the Asus Prime and Gigabyte Gaming X, but they both look even more poorly equipped than my current MSi Pro.

I know that feeling with the RMA. I RMAd two MSi boards last year (they were the same model, though)... that's why I can't comprehend why I trusted this joke of a brand again.

The Gene is $600 lol, do you actually wanna spend that much I only brought it up as it's got the layer count advantage (+ 1DPC advantage!) along with the Gigabytes. Gaming X and DS3H excluded of course, those two are not only 6 layer but sus as hell VRM

I'm sure the 8-layers and Gene will start pulling ahead when next gen CPUs have actual mem OC capabilities, but Raphael is currently holding all the boards back

I won't say that TM5 *can't* catch UMC/Fabric related instability, but if it's printing errors off like mad and rebooting then it's most likely the RAM profile that isn't stable. VDIMM and VSOC are both low by default at JEDEC right?

 

[AusWolf]

The Gene is $600 lol, do you actually wanna spend that much

Oh Jesus, NO! I'd much rather boot twice at every cold start, thanks.

I won't say that TM5 *can't* catch UMC/Fabric related instability, but if it's printing errors off like mad and rebooting then it's most likely the RAM profile that isn't stable. VDIMM and VSOC are both low by default at JEDEC right?

I think they're 1 V or 1.05 V or something like that along with the SoC.

Edit: Oh wait, no. This is my JEDEC profile (I forgot that I took a picture earlier):

 

[tabascosauz]

Oh Jesus, NO! I'd much rather boot twice at every cold start, thanks.

I think they're 1 V or 1.05 V or something like that along with the SoC.

Edit: Oh wait, no. This is my JEDEC profile (I forgot that I took a picture earlier):

Interesting how your UCLK defaults to 2400 1:1 and FCLK to 2000 at JEDEC though. I'm not sure that's supposed to happen? They said default Fabric is 1733 for Raphael. I would think that 1733 FCLK would provide a safe and reliable fallback, but that's just me.

Considering the alleged extra VSOC headroom, seems like you could easily add some VSOC and see what happens.

Might want to play with your VSOC LLC too. Not sure what the board is setting at JEDEC but I doubt it's defaulting to a weird number like 1.03V.

Could also mess with PLL (CPU 1P8) maybe up to 1.9 or 1.95. Doesn't look like that feature has changed much.

 

[Mussels]

I was actually thinking of the Asus TUF. I know it's a 6-layer one, but it has more than decent VRM, has my exact RAM on its compatibility list, and I've never had any problem with any TUF board before.

Asus x570 Tuf boards were their absolute worst, missing BIOS features, overheating chipsets, chipset fans with no control locked at 3K RPM at all times (leading to the myth all ryzen boards did that) and terrible ram clocking

 

[freeagent]

I bought a TUF board once from Newegg. It was DOA. But! I could not believe what I was holding. It felt like such a cheap board. It was dead though, so it really was a piece of junk

I like Asus boards, but the quality of that board was subpar.

 

[AusWolf]

Interesting how your UCLK defaults to 2400 1:1 and FCLK to 2000 at JEDEC though. I'm not sure that's supposed to happen? They said default Fabric is 1733 for Raphael. I would think that 1733 FCLK would provide a safe and reliable fallback, but that's just me.

Considering the alleged extra VSOC headroom, seems like you could easily add some VSOC and see what happens.

Might want to play with your VSOC LLC too. Not sure what the board is setting at JEDEC but I doubt it's defaulting to a weird number like 1.03V.

Could also mess with PLL (CPU 1P8) maybe up to 1.9 or 1.95. Doesn't look like that feature has changed much.

I had a feeling that 2400 UCLK is a bit high, so I dropped it back to 1:2 at 1200, and I couldn't even boot into Windows without a blue screen error. Now I understand how this works even less.

Next, I'll try adding a bit more voltage to the SoC and RAM at JEDEC. 1.2 V should do. If it errors out again, I'll be even more puzzled.

Maybe the best (or least bad) option is to leave it at EXPO and double boot every time.

Edit: TM5 successfully finished at JEDEC 4800 MHz, 1.2 V SoC, 1.2 V RAM, 2000/2400 FCLK/UCLK. It looks like my board's sweet spot is around this... until maybe a BIOS update that actually runs... or until I decide to give up and buy something better.

Edit 2: Again, no errors in TM5 at 5000 MHz, 1.2 V RAM, Auto timings (which are quite crap with CL 42, but whatever), 2000/2500 FCLK/UCLK, but with Auto = 1.065 V SoC. I think it's safe to conclude that 6000 MHz EXPO is a bit too much for this motherboard, even though it officially supports it with single sided DIMMs (aka. it's a piece of MSi crap), but the standard 1.1 V isn't enough for my RAM at JEDEC.

Edit 3 (I'll really stop now): Another stable run at 4800 MHz Jedec, VSoc Auto = 1.02 V, RAM at 1.15 V. I've noticed that HWinfo shows 1.14 V as RAM voltage, so my next suspicion is that it probably isn't stable at JEDEC with stock voltages because something's off with the motherboard's LLC and the RAM voltage drops too far below 1.1 V which I have to manually compensate. I'll just leave it at 1.15 V @ JEDEC for now.

I've got my plan: I'll wait a month or two for MSi to update their BIOS for better EXPO support. If they won't bother, or if I can't install the update like I couldn't install v14, then I'll buy the Asus TUF later.

Asus x570 Tuf boards were their absolute worst, missing BIOS features, overheating chipsets, chipset fans with no control locked at 3K RPM at all times (leading to the myth all ryzen boards did that) and terrible ram clocking

X570 was kind of a hit and miss altogether, imo. I had the B550 TUF, and the Intel B560 TUF in micro-ATX flavour - both are excellent boards.

I bought a TUF board once from Newegg. It was DOA. But! I could not believe what I was holding. It felt like such a cheap board. It was dead though, so it really was a piece of junk

I like Asus boards, but the quality of that board was subpar.

Yeah, old TUF products were quite cheaply made, but the brand has come a long way since then.

 

[AsRock]

Finally got my mobo, Amazon f ing around again haha. Next be some voltage tweaks.

 

[AusWolf]

Finally got my mobo, Amazon f ing around again haha. Next be some voltage tweaks.


View attachment 272939

3000 UCLK at 1.24 VSoC!

Either you have a golden sample, or my motherboard is really crap.

 

[MagnyCours]

Mine also does without problems! I think EXPO automatically sets 1.25v~1.35v VSoC for 6000MHz, well, my motherboard does at least. Though AMD did claim that 6000 is the "sweet spot" for Ryzen 7000, so it is kind of odd that you're having lots of trouble running it.

 

[AusWolf]

Mine also does without problems! I think EXPO automatically sets 1.25v~1.35v VSoC for 6000MHz, well, my motherboard does at least. Though AMD did claim that 6000 is the "sweet spot" for Ryzen 7000, so it is kind of odd that you're having lots of trouble running it.

View attachment 272971

Mine sets 1.36 VSoC, but drops UCLK back to 1:2 at 1500. Then comes the hang-up during cold boot, needing me to turn the PC off and on again. If I set UCLK to 1:1 at 3000, it won't even boot at all.

I think my motherboard is a piece of crap without a doubt.

 

[Super Firm Tofu]

Mine sets 1.36 VSoC, but drops UCLK back to 1:2 at 1500. Then comes the hang-up during cold boot, needing me to turn the PC off and on again. If I set UCLK to 1:1 at 3000, it won't even boot at all.

I think my motherboard is a piece of crap without a doubt.

My gut tells me it's your memory. I'm Running 2000:3000:3000 at 1.02v SoC. Any chance you can get another set of RAM to test with?

 

[AusWolf]

My gut tells me it's your memory. I'm Running 2000:3000:3000 at 1.02v SoC. Any chance you can get another set of RAM to test with?

View attachment 273008

It might be the memory, considering it has trouble running at JEDEC with stock voltage. Although it is fine at 1.15 V.

I still think it is the motherboard, though.

When I set the RAM to run at 1.15 V, HWinfo reports 1.14 V. If it can only run at no less than 1.1 V, and the motherboard drops below that, it can cause problems. That is, something's up with the stock LLC options.

The other thing is the SoC voltage. If you can run 3000 UCLK with 1.02 VSoC, then there's no way the motherboard should apply 1.36 V just to run the RAM at 6000 with 1500 UCLK, but mine does for some reason.

Unfortunately, I'm the first and only one with DDR5 in my circle, so I can't test it unless I buy another kit.

 

[tabascosauz]

It might be the memory, considering it has trouble running at JEDEC with stock voltage. Although it is fine at 1.15 V.

I still think it is the motherboard, though.

When I set the RAM to run at 1.15 V, HWinfo reports 1.14 V. If it can only run at no less than 1.1 V, and the motherboard drops below that, it can cause problems. That is, something's up with the stock LLC options.

The other thing is the SoC voltage. If you can run 3000 UCLK with 1.02 VSoC, then there's no way the motherboard should apply 1.36 V just to run the RAM at 6000 with 1500 UCLK, but mine does for some reason.

Unfortunately, I'm the first and only one with DDR5 in my circle, so I can't test it unless I buy another kit.

Some variance in set and reported VDIMM is to be expected on any motherboard. Most vendors except Gigabyte are usually ±0.01V. Don't take HWInfo at its word either, only a multimeter will tell you for sure.

I don't think @Super Firm Tofu 's 1.02V VSOC is EXPO default, probably an optimal number reached after testing. The others are also setting 1.25V automatically. Silicon quality is at play there and just cause others can do 1.02V @ 3000 doesn't mean yours can (sadly).

If you can pass linpack and memory-biased ycruncher tests in addition to the TM5 testing you've already done, it's a pretty clear indicator that your RAM is just fine, and your board is probably not the issue either. Generally boards running into absolute memory limits will make it pretty obvious they're not happy, and you will not be able to have a good time at 6000.

Which is why it'll be very telling if you do some 1:1 testing at 5200/5400/5600/5800 - if UCLK runs into a brick wall at any one of those speeds then you're just running into the memory controller's limits. At this point AMD saying "sweet spot" is code for "good luck lol"

 

[Mussels]

I'm not getting any WHEA or bug check errors in event viewer, the crashes also don't seem to be logged as they're like hard resets, the only errors that seem to be happening are:



Doesn't seem like they were happening with the 7600X. I did have to reset my PIN at first 2 logins after fitting the 7950X, so not sure if there's some kind of fTPM or security issue or just over-optimistic PBO/CO. Or shit ASrock BIOS. Memtest does 3500% with the current settings so it's probably not the memory as i ran it like that for weeks before.

Saw this error on my 5800x system, googled it, got your post. damned useless errors.

Eventually the google and reddit rabbit hole found out it's a leftover error from thaiphoon burners service
In an admin CMD prompt you need to run
sc delete ssdgio

 

[igralec84]

Saw this error on my 5800x system, googled it, got your post. damned useless errors.

Eventually the google and reddit rabbit hole found out it's a leftover error from thaiphoon burners service
In an admin CMD prompt you need to run
sc delete ssdgio

Thank you for this, it worked and the error is now gone. Never checked how long it's been happening but probably since last summer when i installed it to use the DRAM calculator and then forgot about it.

Since then i'm 95% certain my crashes / hard reboots that don't even show up on the event logs are over-optimistic CO and boost.
When i'm bored, i run core cycler, 3dmark cpu profiler and cinebench, but it seems all core -10 and +125mhz or -13 and + 100mhz are stable or at least didn't crash in 2 days

 

[Super Firm Tofu]

It might be the memory, considering it has trouble running at JEDEC with stock voltage. Although it is fine at 1.15 V.

I still think it is the motherboard, though.

When I set the RAM to run at 1.15 V, HWinfo reports 1.14 V. If it can only run at no less than 1.1 V, and the motherboard drops below that, it can cause problems. That is, something's up with the stock LLC options.

The other thing is the SoC voltage. If you can run 3000 UCLK with 1.02 VSoC, then there's no way the motherboard should apply 1.36 V just to run the RAM at 6000 with 1500 UCLK, but mine does for some reason.

Unfortunately, I'm the first and only one with DDR5 in my circle, so I can't test it unless I buy another kit.

Sorry - I didn't include the whole story. Out of the box with EXPO settings, I also had 1.36 SoC, and 1500. I had to manually change to force 3000. I dropped SoC back to standard and ran all of the stress tests again (core cycler, TM5, y-cruncher) to make sure it was still ok. Everything else with RAM is still EXPO standards as far as voltage/timings. I'm also running CO at -25 all core with reduced PPT, EDC, and TDC (95w TDP mode in the BIOS).

@tabascosauz is probably right that I just lucked out with a decent sample of the 7700x. I wish we were close so we could swap RAM to rule that out - I'd really like to see you able to get it all sorted.

Which is why it'll be very telling if you do some 1:1 testing at 5200/5400/5600/5800 - if UCLK runs into a brick wall at any one of those speeds then you're just running into the memory controller's limits. At this point AMD saying "sweet spot" is code for "good luck lol"

I'd really recommend trying what @tabascosauz is recommending to see if you can get past the need for a double boot @ 6000.

Also, as far as EXPO goes, it appears to me that it's just a WalMart version of XMP where they just crank up voltages and hope for the best. I may have gotten lucky with my specific CPU, but it wasn't plug and play by any means.

As a side note I also had a BIOS update that was supposed to 'Reduce Boot Time' and 'Increase Stability'. Post time dropped by 3 whole seconds (down to 27 seconds whoo-hoo), and while I had no stability problems before, the new update as least hasn't introduced any new instabilities.

 

[AusWolf]

Some variance in set and reported VDIMM is to be expected on any motherboard. Most vendors except Gigabyte are usually ±0.01V. Don't take HWInfo at its word either, only a multimeter will tell you for sure.

I don't think @Super Firm Tofu 's 1.02V VSOC is EXPO default, probably an optimal number reached after testing. The others are also setting 1.25V automatically. Silicon quality is at play there and just cause others can do 1.02V @ 3000 doesn't mean yours can (sadly).

I know. The weird thing is that it can't even do the standard 1500 @ 1.36 VSoC without a double boot, but it can do 2400 @ 1.02 V at JEDEC without a problem. Common sense dictates that increased voltage with decreased frequency should result in much improved stability, but common sense goes down the toilet with my setup.

Which is why it'll be very telling if you do some 1:1 testing at 5200/5400/5600/5800 - if UCLK runs into a brick wall at any one of those speeds then you're just running into the memory controller's limits. At this point AMD saying "sweet spot" is code for "good luck lol"

I'll do that once I have a little time to experiment. The problem is, without a second RAM kit and/or motherboard, there's no way to know for sure if it's the RAM hitting a wall, or the motherboard quality. The only thing I can rule out this way is the memory controller (if it's not some motherboard oddity preventing it from reaching 3000, that is).

Sorry - I didn't include the whole story. Out of the box with EXPO settings, I also had 1.36 SoC, and 1500. I had to manually change to force 3000. I dropped SoC back to standard and ran all of the stress tests again (core cycler, TM5, y-cruncher) to make sure it was still ok. Everything else with RAM is still EXPO standards as far as voltage/timings. I'm also running CO at -25 all core with reduced PPT, EDC, and TDC (95w TDP mode in the BIOS).

@tabascosauz is probably right that I just lucked out with a decent sample of the 7700x. I wish we were close so we could swap RAM to rule that out - I'd really like to see you able to get it all sorted.

That's nice of you. I know myself, I'll probably end up buying an extra RAM kit, or motherboard, or something eventually just to experiment. It's only that Christmas is coming, presents need to be bought, and I'm having the days as unpaid leave (forget about a bonus at my workplace), so it'll have to wait until spring time.

I'd really recommend trying what @tabascosauz is recommending to see if you can get past the need for a double boot @ 6000.

Also, as far as EXPO goes, it appears to me that it's just a WalMart version of XMP where they just crank up voltages and hope for the best. I may have gotten lucky with my specific CPU, but it wasn't plug and play by any means.

As a side note I also had a BIOS update that was supposed to 'Reduce Boot Time' and 'Increase Stability'. Post time dropped by 3 whole seconds (down to 27 seconds whoo-hoo), and while I had no stability problems before, the new update as least hasn't introduced any new instabilities.

I'll do that. I'm also waiting on BIOS updates, but MSi seems to be quite slow with my board. I guess the Pro series isn't too high up on their list of things to fix.

It's a shame that EXPO is such a cranky thing at this point. I didn't have too high expectations, but I was kind of hoping it wouldn't be a lot more difficult to use than XMP. I guess it's true with certain systems if you don't mind a high VSoC. I also took extreme care to choose the right kit for the board. It only supports single-sided kits of two at 6000 MHz, so I made sure to pick exactly that. Although, this specific kit isn't on the official support list - as a matter of fact, there's only 3 Samsung die 6000 kits on the list, the rest are all Hynix. Maybe it isn't the board, it isn't the RAM, but the combination of the two.

Another oddity with the board is that it sometimes forgets to enable RGB, and I can't even access RGB settings from the MSi Centre. A restart doesn't help, but if I turn the PC off and on again, it works.

 

[tabascosauz]

I'll do that once I have a little time to experiment. The problem is, without a second RAM kit and/or motherboard, there's no way to know for sure if it's the RAM hitting a wall, or the motherboard quality. The only thing I can rule out this way is the memory controller (if it's not some motherboard oddity preventing it from reaching 3000, that is).

That's nice of you. I know myself, I'll probably end up buying an extra RAM kit, or motherboard, or something eventually just to experiment. It's only that Christmas is coming, presents need to be bought, and I'm having the days as unpaid leave (forget about a bonus at my workplace), so it'll have to wait until spring time.

It's a shame that EXPO is such a cranky thing at this point. I didn't have too high expectations, but I was kind of hoping it wouldn't be a lot more difficult to use than XMP. I guess it's true with certain systems if you don't mind a high VSoC. I also took extreme care to choose the right kit for the board. It only supports single-sided kits of two at 6000 MHz, so I made sure to pick exactly that. Although, this specific kit isn't on the official support list - as a matter of fact, there's only 3 Samsung die 6000 kits on the list, the rest are all Hynix. Maybe it isn't the board, it isn't the RAM, but the combination of the two.

Hynix A should still be single rank I think?

The double RAM training might go away with a different kit. Different ICs can have different POST behaviour, my 4Gb E-die kit is always a lot slower to POST than the newer DDR4 ICs

Corsair Vengeance RGB DDR5-6000 CL30 (AMD Expo) 2x 16 GB Review

Corsair's Vengeance RGB takes the AMD DDR5 platform to a sweet spot. Using the new EXPO 6000 MT/s profile that's optimized for Zen 4 and tight timings of 30-36-36-76, this kit is designed to fit right into the Corsair ecosystem to provide a great end-to-end user experience.

www.techpowerup.com

BZ has a buying guide for DDR5 - like other OCers he seems to have strong opinions against not-Hynix. Makes sense, Samsung and Micron are really useless on the stronger Intel IMC

 

[AusWolf]

Hynix A should still be single rank I think?

The double RAM training might go away with a different kit. Different ICs can have different POST behaviour, my 4Gb E-die kit is always a lot slower to POST than the newer DDR4 ICs

Corsair Vengeance RGB DDR5-6000 CL30 (AMD Expo) 2x 16 GB Review

Corsair's Vengeance RGB takes the AMD DDR5 platform to a sweet spot. Using the new EXPO 6000 MT/s profile that's optimized for Zen 4 and tight timings of 30-36-36-76, this kit is designed to fit right into the Corsair ecosystem to provide a great end-to-end user experience.

www.techpowerup.com

BZ has a buying guide for DDR5 - like other OCers he seems to have strong opinions against not-Hynix. Makes sense, Samsung and Micron are really useless on the stronger Intel IMC

My system isn't just slow to boot when EXPO is enabled. It doesn't boot at all unless I turn it off and on again.

I guess I'm stuck with it for a while now. Christmas is coming, so I'll hold myself back from buying a new motherboard or RAM kit at least until spring. I'll have a play with what you suggested, and hopefully, MSi issues some BIOS updates as well, one of which will potentially help my case... fingers crossed.

 

[AsRock]

Hynix A should still be single rank I think?

The double RAM training might go away with a different kit. Different ICs can have different POST behaviour, my 4Gb E-die kit is always a lot slower to POST than the newer DDR4 ICs

Corsair Vengeance RGB DDR5-6000 CL30 (AMD Expo) 2x 16 GB Review

Corsair's Vengeance RGB takes the AMD DDR5 platform to a sweet spot. Using the new EXPO 6000 MT/s profile that's optimized for Zen 4 and tight timings of 30-36-36-76, this kit is designed to fit right into the Corsair ecosystem to provide a great end-to-end user experience.

www.techpowerup.com

BZ has a buying guide for DDR5 - like other OCers he seems to have strong opinions against not-Hynix. Makes sense, Samsung and Micron are really useless on the stronger Intel IMC

Yeah that's the video i used as a guide, i never had any luck with low end Corsair ram going all the way back to the Athlon days and glad i did not get it when i build my AM4 build too as it had issue's there as well.

 

[Mussels]

Since then i'm 95% certain my crashes / hard reboots that don't even show up on the event logs are over-optimistic CO and boost

black screen crashes and sudden reboots are usually related to the SoC or undervolting
It all ties into the SoC stability in the end, as it's the RAM, PCI-E and USB ports all in one - when it errors it'd reset (the black screen, with or without it recovering) or have USB devices bing-bong

When you look at it together it comes back as "Oh yeah devices reset like they'd been unplugged and replugged, and the PC didnt like it when critical ones did that"


my original 2700x system had this all the time, and it came down to a faulty corsair RAM stick that went unstable around 35-40C at a mere 2667Mhz

 

[Mussels]

Looks like Asus owners get a free ride


5800x3D ownwers have PBO unlocked in the new 4408 BIOS

Changelog of course, tells us nothing

Found on HardwareLuxx (where they've already modded the BIOS to unlock even more settings)

and reddit

 

[tabascosauz]

Looks like Asus owners get a free ride


5800x3D ownwers have PBO unlocked in the new 4408 BIOS

Changelog of course, tells us nothing
View attachment 273682


Found on HardwareLuxx (where they've already modded the BIOS to unlock even more settings)
View attachment 273683

I've been running the 4303 beta for a couple days. CO and PBO.

Not sure if anything else has changed. New SMU version and AGESA 1208 now, but publicly it was just to address vulnerabilities. Obviously I will try again to break the FCLK 1900 wall but I expect nothing

 

[Mussels]

A new BIOS setting got me all moist in my panties, i'll let you guys know if it works, or just a broken unlockable option

attempt 1: (it has a few choices)
nope

I've been running the 4303 beta for a couple days. CO and PBO.

Not sure if anything else has changed. New SMU version and AGESA 1208 now, but publicly it was just to address vulnerabilities. Obviously I will try again to break the FCLK 1900 wall but I expect nothing

4408 has the new stuff, not 4403

Okay so that features doesnt do anything

set what i previously had and CO doesnt seem to be working oddly enough (and it WAS working)

1.275V and 75C says stock