BIOS updates for AM4 motherboards

was back to stock to test 3200.

Well that sucks, seems they have it very close to working for 3200 but I think alot of it is just up to what the memory controller can handle. Good news is it looks like they have Zen+ working good on the memory front. So if your planning like me to wait out for Zen 2 then I think 16 gig memory sticks wont be a issue by then.
 
I removed my ProcODT setting at 96 that had been required to get 2933 previously and noticed that tRC was set to 75 instead of the 52 that the kit supports. After fixing this, it completed 500% in HCI Memtest Pro overnight at 3200 16-16-16-36-52 1T

Looking good so far.

The Q-Fan settings are much more aggressive now, it goes to 100% at 50C now.
 
I added X470 board links to the first page. I must say, in my opinion, that this launch appears to be much more stable and without the hiccups of the first generation. Platform is maturing. The motherboards are really shining here. For you B350 chipset guys, whatever your processor, the X470 is your upgrade. Huh, can't shake the feeling there is a 2800X waiting in the wings.;)
 
Welp, count me in for a test-run for the 2700x on the X370 Professional Gaming board.


I'll be updating from 4.40 to 4.60, I'll let you guys know how it goes.
 
Looks like another C6H BIOS available: http://www.overclock.net/forum/27198569-post35517.html

Unverified change log:

- DOCP Profile 1-5 removed (also in 6004)
- BCLK Divider is back (also in 6004)
- S3 resume stuck at E3 when overclocking fixed
- Q-Fan Tuning fix for DC mode

C6H/C6E Beta BIOS 6101

http://www.mediafire.com/file/mcaddy...-ASUS-6101.zip sha256 1cdd6c8913b63a32226824df8c8c328ff75248ebe3ecf24d0c 2d394ce0198b72
http://www.mediafire.com/file/4o199n...-ASUS-6101.zip sha256 8990f33b333afa69520883cb6793628e972378b13ba3307ecc 58b55c28d40646
http://www.mediafire.com/file/z9dss6...-ASUS-6101.zip sha256 b213d1c20d14f92046f51a2b61499832459a6ac0a9670866c2 230511c846834f
 
I added X470 board links to the first page. I must say, in my opinion, that this launch appears to be much more stable and without the hiccups of the first generation. Platform is maturing. The motherboards are really shining here. For you B350 chipset guys, whatever your processor, the X470 is your upgrade. Huh, can't shake the feeling there is a 2800X waiting in the wings.;)

Is it? I think B350 works fine except for overdrive. The precision boost 2 and XFR2 are built into the chip, but you can't get the extra precision boost 2 overdrive (which removes amperage limits) without x370/x470.

I'll find out in an hour or two.
 
Just an update, it works and it feels to be working very well.

The only setting I've changed from the 4.6 Bios is the XMP profile for 3200 memory.
One weird thing is that HW Monitor shows it running steady @ 3950, up to 4200 at any given moment during gaming.
But, will have a max clocks # of:

5391
5391
5358
5358
5391
5358
5358
5391

Did I win the silicone lottery?
 
Make sure you're using the latest version. Someone else noticed much higher clocks than almost everyone and it turned out to be an older version of the monitor software.
 
I'm using the latest HWMonitor, 1.35.0 Driver version 146

It looks like it's not afraid to gobble up voltage with the Max Voltage #'s of 1.494V across 5 cores and 1.5000 on the other 3.
 
I'm using the latest HWMonitor, 1.35.0 Driver version 146

It looks like it's not afraid to gobble up voltage with the Max Voltage #'s of 1.494V across 5 cores and 1.5000 on the other 3.

HWMonitor completely stinks compared to HWinfo64. More accurate readings and less hiccups.
 
gigaxtreme1
You should change the link for the MSI Titanium to actually go to the BIOS tab instead of the Drivers one :p
https://www.msi.com/Motherboard/support/X370-XPOWER-GAMING-TITANIUM.html#down-bios

For everyone using the Titanium, there's a new BIOS out (v1.F) as of May 4th, running the new AGESA 1.0.0.2a! :D
I've not yet grabbed it, but till be soon. I don't anticipate any new goodies to unlock via modding it, but fingers crossed heh

Full 1.F changelog:
- Improved USB mouse compatibility.
- Improve M.2 compatibility.
- Update AGESA code 1.0.0.2a
 
So there I was, pouring through the latest Titanium BIOS's list of Strings and.... what in the fffff.......?? o_0

(NOTE: these fall right in the midst of the new PB Overdrive strings and a bunch of Core Performance Boost [CPB] strings that are unfamiliar to me)
0x139E Performance Regulator
0x139F Enable specific option to gain extra score for benchmark.
0x13A0 Cinebench R15
0x13A1 Cinebench R11.5
0x13A2 GeekBench
0x13A3 AIDA64 Memory
0x13A4 LN2 OC [I'm not specifically sure if this is related or not to the Perf Regulator option]

I'm not sure how exactly to take this... It feels suspiciously like cheating. Given that I have not come across where these are even located in the Setup portion, I have no way to access these to even try them out so I really can't pass a fair opinion on them considering.

Also the mysterious "Mode0" that others have noticed, seems to be related to PB Overdrive, but I'm just as much in the dark in terms of what it does as anyone else... :p
 
So there I was, pouring through the latest Titanium BIOS's list of Strings and.... what in the fffff.......?? o_0

(NOTE: these fall right in the midst of the new PB Overdrive strings and a bunch of Core Performance Boost [CPB] strings that are unfamiliar to me)
0x139E Performance Regulator
0x139F Enable specific option to gain extra score for benchmark.
0x13A0 Cinebench R15
0x13A1 Cinebench R11.5
0x13A2 GeekBench
0x13A3 AIDA64 Memory
0x13A4 LN2 OC [I'm not specifically sure if this is related or not to the Perf Regulator option]

I'm not sure how exactly to take this... It feels suspiciously like cheating. Given that I have not come across where these are even located in the Setup portion, I have no way to access these to even try them out so I really can't pass a fair opinion on them considering.

Also the mysterious "Mode0" that others have noticed, seems to be related to PB Overdrive, but I'm just as much in the dark in terms of what it does as anyone else... :p
Maybe a hidden option for when pb3/4 are selected? Unused, or automatic for now?
 
it could just be bios settings/internal versions from testing to keep track of changes that never got cleared out..
 
I'm certain that it's not a setting that is for public usage, but what my initial thought was that it was a special BIOS they would provide to those participating in overclocking competitions. Only issue is that with a lack of BClk on the TItanium, that seems weird that anyone would bother to use it (and the fact people weren't that impressed by the Nikos MOSFETs used).

But yes, overall, there are tons and tons of strings in there that simply don't belong. While I'm certain that these ones ARE for AM4 (due to its location amongst the strings), my proof of there being tons that don't belong come from the massive amounts of Intel related stuff. Mainly things for the Game Boost feature, to "dial an OC". I guess they just keep it in there to function as "padding" to make up the proper file size, but I dunno.

From a curiosity standpoint it'd be fun to have access too :D I mean I know I could futz with it, there are EFI "addons" that people have made that you can execute from the UEFI Console which let you manipulate the BIOS on a hex level, but Hex is way above me unfortunately :p
 
MSI finally has a bunch of new updates for their boards. I'm talking about the AGESA 1.0.0.2a bios updates. The B350 Tomahawk, X370 Gaming Pro Carbon are examples with updates, but there are more as well.
 
Well damn... Not sure in which version MSI removed these, but the Titanium no longer has the "CPU Power Duty Control" and "CPU NB Power Duty Control" options in the DigitALL Power menu. They let you set a Thermal Balance or Current Balance, to keep VRMs within a temperature, or to ignore temp and maintain a current draw. Whether or not that was of any benefit, I dunno, but it sucks to lose any settings IMO. The Strings for the options are still present in the BIOS, just not the menu entry anymore.

Other than that I haven't tried out whether the RAM can clock better yet. However, they also removed "BankGroupSwapAlt" from the Gen1 Ryzens, for who knows what reason. It's in the BIOS, it's just not accessible to us 1st-geners as it's in a different Timings menu. Which yes, I too, ask why they need to do that, which, I haven't a clue.

That being said, as usual, the OC Profile from the older version does not work and thus I had to manually re-enter things. Sadly, I completely forgot to look at my fan profiles so now I'll have to flash BACK to 1.74 just so I can write that all down :meh:
Though from the looks of it, that's not going to matter... I don't think. The onboard debug LED doesn't change anymore and is locked on "24". I'll flash the unmodded BIOS first after I flash-back to get my fan settings, then we'll see what's what. (AIDA is able to read the temps though, so it may just be the onboard readout that is not functioning, but still unfortunate due to how much I monitor the debug LED)
 
Gigabyte aorus gaming K5
2018/04/17
F23d: Update AGESA 1.0.0.2a + SMU FW 43.18
 
Ended up having to use MSI's BIOS FLASHBACK+ to restore v1.74 since it would appear that now MSI is also locking out flashing back to older BIOS versions :( At least it worked well and as advertised. Name the file MSI.ROM and put it on a blank FAT32 USB drive that's plugged into the vertical port on the back (next to the Clr_CMOS button), press the button for a couple seconds next to the USB3.1 header (bottom right corner) and it was either when I released it restarted or it just restarted on its own, but yea... It worked slick. Flashy lights indicated it was working and after it finished, it restarted again.

I initiated it while IN the BIOS already, as the manual wasn't really clear on how to do it. I initially tried pressing it with system off, no dice. Then I held it down and pressed Power, and it lept into the BIOS. But at that time I had followed earlier directions of not naming it "MSI.ROM" so it wasn't detecting the file. Either way, after changing filename and putting the drive back in, it worked fine.

Also work mentioning, is on v1.F0, the CPU voltage is uncomfortably high. It's not "kill your CPU" high, but it was 1.48V on mine. Pair that with not really having a properly reading CPU temp (was showing 24/25C like the onboard readout) and it could be an issue.

I'll be submitting my bugs to MSI tomorrow though, on the voltage, temp readings, and the missing BankGroupSwapAlt and the other two DigitALL options. Power company is about to do some sort of upgrade at midnight though, so I'd rather not have the computer on for that sudden shutdown of the grid. :p

EDIT: Alright, given their forum is completely user-ran and there was no mention of how to submit bug reports... I just filled out a Support Ticket on it. So I'll just keep my fingers crossed! heh
 
Last edited:
Asus has now officially released the 0601 BIOS for the Crosshair VII Hero Wifi.
2018-05-11
1. Update SMU Firmware to version 43.18.0
2. Update RAID driver to version 9.2.0.41.
3. Update Asus Grid notification behavior

Note: It has the same MD5 checksum as 0601 that was already available from [H] in the AMD Precision Boost 2 and Wraith Prism Deep Dive article on this page here.
 
Well damn... Not sure in which version MSI removed these, but the Titanium no longer has the "CPU Power Duty Control" and "CPU NB Power Duty Control" options in the DigitALL Power menu. They let you set a Thermal Balance or Current Balance, to keep VRMs within a temperature, or to ignore temp and maintain a current draw. Whether or not that was of any benefit, I dunno, but it sucks to lose any settings IMO. The Strings for the options are still present in the BIOS, just not the menu entry anymore.

Other than that I haven't tried out whether the RAM can clock better yet. However, they also removed "BankGroupSwapAlt" from the Gen1 Ryzens, for who knows what reason. It's in the BIOS, it's just not accessible to us 1st-geners as it's in a different Timings menu. Which yes, I too, ask why they need to do that, which, I haven't a clue.

That being said, as usual, the OC Profile from the older version does not work and thus I had to manually re-enter things. Sadly, I completely forgot to look at my fan profiles so now I'll have to flash BACK to 1.74 just so I can write that all down :meh:
Though from the looks of it, that's not going to matter... I don't think. The onboard debug LED doesn't change anymore and is locked on "24". I'll flash the unmodded BIOS first after I flash-back to get my fan settings, then we'll see what's what. (AIDA is able to read the temps though, so it may just be the onboard readout that is not functioning, but still unfortunate due to how much I monitor the debug LED)

The new 1.0.02A agesa code improves OpenGL performance on cinebench 15 up to the levels we saw prior to 1.0.0.7a . Now I am getting 130fps again for multicore cpu. Of course I am on the X470 Gaming M7 AC now. Wanted the improved vrm on the X470 M7 and the better heat sink/heat pipes for the m.2 drive slots.
 
The new 1.0.02A agesa code improves OpenGL performance on cinebench 15 up to the levels we saw prior to 1.0.0.7a . Now I am getting 130fps again for multicore cpu. Of course I am on the X470 Gaming M7 AC now. Wanted the improved vrm on the X470 M7 and the better heat sink/heat pipes for the m.2 drive slots.
Since I had a sure-fire way to get 1.74 on if I so desired, I tossed an un-modded 1.F0 on mine just now... Temps are still frozen as far as I can tell (on the Diag LED), though the BIOS did show it moving on that end, since it was running at 31C while flashing. Voltage still showed 1.48V after starting with 1.F0 which is disconcerting, but I dropped it back to the 1.25V I run it at for 3.7GHz (I'm not overclocked currently).

NOW... the important part of why I'm bothering to post... lol
I changed 3 things a few minutes ago: CPU Voltage as mentioned, CPU-NB/SoC Voltage set to 1.05V (I usually run 0.975V at 3200), and "Memory Try It !" set to DDR4-3466 16-18-18-38 (which auto set DRAM to 1.4V).

And...
it...
Booted :D First try, into Windows, no BSOD yet, just been idling on the desktop for the past 15 minutes while I typed another post! So THAT makes me excited and super happy! Time to see exactly how well it's working though and try to run some benches.

If it does much much tweaking to be done, as all the sub-timings were left on Auto.

EDIT: Gaaaa, drat, so close! I'll sort out my CPU Voltage first before I call it, as the Boost was still functioning and I KNOW that 3.8GHz isn't stable at 1.25V, so for sure 3.9 wouldn't be.
Got through the Read and Write tests on AIDA's Cache & Memory bench. 51900MB/s and 49700MB/s, but BSODed during Copy. Not too bad IMO considering stock CPU speed, high timings AND fully auto sub-timings. I'll run the Latency test first this time.

EDIT2: Alright so I set VCore to 1.35V. Manually running the last 2: Latency is impressive and the same as my old settings on BIOS 1.74: 70.4ns. Copy is awful and shows to not be tuned: 431700MB/s. Time to see if she'll hold up to Cache Mem test in its entirety now...
Bazinga! :D I think I misread the Read/Write speeds before it crashed, but it complete it this time!

Whether or not related, my keyboard isn't functioning now. It's plugged into a USB3 port connected to the internal header, but the weird part is it has a USB Hub in which I have my mouse connected to... Keyboard = not registering but Mouse = functional >_> Color me confused.*shrug* Ignore me. It's all fine. heh

Either way... 3466!! lol *cracks knuckles* Alright my pretties, I know I haven't even check stability on you yet but how do you feel about 3600? :shame:
 
Last edited:
So there I was, pouring through the latest Titanium BIOS's list of Strings and.... what in the fffff.......?? o_0

(NOTE: these fall right in the midst of the new PB Overdrive strings and a bunch of Core Performance Boost [CPB] strings that are unfamiliar to me)
0x139E Performance Regulator
0x139F Enable specific option to gain extra score for benchmark.
0x13A0 Cinebench R15
0x13A1 Cinebench R11.5
0x13A2 GeekBench
0x13A3 AIDA64 Memory
0x13A4 LN2 OC [I'm not specifically sure if this is related or not to the Perf Regulator option]

I'm not sure how exactly to take this... It feels suspiciously like cheating. Given that I have not come across where these are even located in the Setup portion, I have no way to access these to even try them out so I really can't pass a fair opinion on them considering.

Also the mysterious "Mode0" that others have noticed, seems to be related to PB Overdrive, but I'm just as much in the dark in terms of what it does as anyone else... :p

I decided I was sick and tired of MSI's lack of "Help" strings for DRAM and grabbed the CH7 Hero's latest BIOS to steal their's lol
That being said, there is the EXACT same option in the ASUS BIOS! In there it IS available to use (whether or not hidden from users, I don't know). It's in the Extreme Tweaker section and labeled as "Performance Bias" with the string reading "Different Values may help different Software's performance", providing the same 4 options (which seems to rule out LN2 OC as one, in my listing of MSI strings).

There are also tons of options, that's for sure lol Again, which ones are available by default, I have nooo idea. I'm also kinda glad I don't have this board as I'd feel like I wouldn't be getting everything I could out of my setup, and as such, be frustrated that I couldn't figure out how some of the options work.
For example, this one labeled "NB Frequency" with the help string of
"Adjust NB Frequency.
The NB Frequency can be lowered to allow more OC margin for the memory, or higher to boost memory performance. Note that the relative ratio between the NB Frequency and the Memory Frequency has to be 1.25 : 1 or bigger for proper operation system."


But... so much for that idea. The CH7 provides ZERO help on the DRAM Settings, which is rather surprising, considering my Titanium actually provides some help on a few of them.

Does anyone know if Gigabyte, ASRock or Biostar boards provide any useful tidbits on their Timing options? (I'm obviously looking for things beyond explaining that tWTR is Write to Read, etc)
 
I heard back from MSI support on the missing BIOS options on the Titanium (and possibly other MSI boards), and it's kinda hard to follow due to language barrier (and their ticket system sucking shit by removing line breaks), but the long and short of it is:
High voltage by default is apparently due to the AGESA and the BIOS engineer claims that MSI applies no voltage on Auto. He tested it on a Titanium and an ASUS Prime X370 Pro, both having the same results.
BankGroupSwapAlt, if I understand correctly, is said to be a Ryzen APU feature only. Seems odd, since so many boards have/had it for quite a long time.
The DigitALL options they removed (CPU Power Duty Control, CPU NB Power Duty Control) are because of: "MSI has optimized the HW design toward the thermal/current balance of MOSFET and we notice this options does not do much in term of performance. Hence, the options are removed by spec".
The Debug LED (POST Codes) no longer displaying the CPU temp is intentional: "We remove this feature due to the native CPU temperature offset behavior of AM4 CPU. Please use command center or 3rd party software to read CPU temperature."

I followed up with saying that I do not feel it right that they remove features like that (referring to the Temp readout and the Power Duty Control options), as it's sort of a bait-and-switch. Personally, the temp readout was a selling point to me and I liked that. I requested there be a BIOS option to enable/disable[default] to allow those of us who understand there is a temperature offset to still be able to use that feature, further supplying the idea of having it auto calculate the offset based on the CPU installed and subtract it from the readout's value. Should the latter be too much work, then at least offer the option to enable it again if we so choose.
As for the Power Duty Control options, my supplied reason was that while it was determined to "not do much", it still did something and even if not, it'd at least provide a placebo effect, as again it sucks having features removed.
Finally, I requested to please provide P-State control like all the other board makers provide. I mean, there's an option in there finally, but it makes zero sense. All it does is provide options of "PState 0/PState 1/PState2", but selecting them doesn't provide any further options so.... *shrug*

They offered to supply me with a Beta v1.G2, but :shy: I had already asked the MSI forum mod if there were any new ones and he linked me to it.

Changes of interest in Beta v1.G2....
-They REMOVED all of the PBO options now, possibly at AMD's request? Either way, these were removed (not sure if you Zen+ had access to it anyways?): "Precision Boost Overdrive", "PPT Limit", "TDC Limit", "EDC Limit", "Precision Boost Overdrive Scalar", and "Customized Precision Boost Overdrive" (the previous one is Auto/Disabled/Manual. This one is the actual scalar with 2x-10x options).
-After version after version of the hidden CPU BClk setting being stuffed right before the "Game Boost" options, they have now curiously given it its own spot... More curiously, they have even renamed them? Instead of being "CPU Base Clock (MHz)" it is "FCH Base Clock (MHz)", which really doesn't make sense to be since AFAIK they haven't called anything on Ryzen a "Fusion Controller Hub" due to being forced to abandon the Fusion branding (due to a lawsuit). I doubt it has any more headroom than previously, but we'll see. (I always make it visible in my modded BIOS)

Other than that, there's some new behind the scenes options, but they're related to APU's iGPU clocks. It appears to be essentially (maybe) letting you define the clocks for each Performance Table, so like GPU PStates (it doesn't really explain it to any degree). It's just called "Low Power PT GPP Clock Force Output" for the menu and each of the 8 options (0-7) are "GPP Clock # Force Output" with Auto/Enabled/Disabled choices. So your guess is as good as mine :p
 
C'mon Asrock, update your X300 series to AGESA 1.0.0.2a. People are starting to complain and I had to buy an MSI X370 Plus,
 
C'mon Asrock, update your X300 series to AGESA 1.0.0.2a. People are starting to complain and I had to buy an MSI X370 Plus,

it's sort of there, it's in the 4.7 bios for the taichi. it just doesn't say it anywhere. but there's a 1.0.0.4a beta bios available for the pro gaming board that fixes the bandwidth bug on zen+ so hopefully one appears for the taichi soon.. problem is asrock just keeps passing blame off on AMD when things don't work even though their boards are the only ones that suffer from the problems.. even though i hate MSI you didn't make a bad choice switching. i regret ever buying the taichi.
 
I was thinking they want to sell their new boards to saturation before they roll out the updates. I got a AsrockB350M-HDV as a stopgap but the deal was there and the R5 1600 will go in that one. If anyone buys the last gen go with X370s, stronger chipset better VRMs mostly.
 
I'm happy with my X470 Tachi - it did recognize my DDR4-3466 right away and runs it at stock timing
/haven't tweaked the CAS yet
 
I like my asrock x370 fatality pro gaming, But the slowness of the firmware updates are a real drag, when I go x499? with my 7nm TR3 next year it will be on an Asus board of some sort!
 
I had a lot of problems getting 4x8GB to work on an Asus Prime Pro X470 w/latest bios and 2700X. Has to be a bios bug. Because I could put 3 in any slots just fine, but 4 nope reboot loop. So I manually entered the same auto timings with 2 sticks on one channel (either channel), and bam boots like a champ with 4 and has been working fine ever since. I should go back and manually change each manual entry back to auto one at a time, and find out which one is the problem. But I'm too lazy.

I stopped updating my Asus Prime B350 w/1500X when the bios added new apus, as I wasn't having any problems w/openSUSE Tumbleweed Linux, so no need to roll the dice with the bigger bios changes. The system is over a year old now anyway.
 
Last edited:
Asus has now officially released the 0702 BIOS for the Crosshair VII Hero Wifi
2018-06-22
Update Agesa Code to 1.0.0.2c.
Update compatibility protocol for 3rd party hardware monitoring software.
Fixed miscellaneous issues with fan calibration/options.
Improve memory compatibility.

I have not tried it yet. Probably will do it this weekend sometime.
 
It only took a fucking year, but FINALLY I can OC my ram to 3333mhz on factory volts, I was stuck at 2933 and 1.45mv....
 
0602 available for Crosshair VI Hero. Using it with R7 2700X and 3200 RAM with no problems. All cores sit at 4.2GHz almost all of the time...
 
Asus has now officially released the 0804 BIOS for the Crosshair VII Hero Wifi
2018-07-20
1. Improve system stability
2. Improve Secure Erase function on NVMe devices
3. Update LPPT firmware

Edit: There is also a new version of the Chipset driver for Windows 10 64bit as well. Released 2018-07-18.

Does anyone know what "LPPT firmware" is?
 
Last edited:
Back
Top