BIOS updates for AM4 motherboards

New v1.73 Beta for the MSI X370 Titanium BIOS up on the MSI Forums Beta BIOS thread.
Here's to save a couple clicks:
https://drive.google.com/open?id=0B-Z98LONNR0hU0FXSXNjekdQRlU

Digging through it with AMIBCP now...,
New setting in CPU Features called "Relaxed EDC throttling".
Help string reads:
Disable - Part-specific EDC throttling protection enabled.
Enable - Reduce the amount of time the processor will throttle.
Auto - AMD's recommendation (Disable)​
(It's actually one long line, so I'm interpreting the line breaks)

New RAM Timing options...
MISC ITEM
BankGroupSwapAlt (this is in addition to the BankGroupSwap option; no help string)​
CAD BUS TIMING CONFIGURATION:
AddCmdSetup - Setup time on CAD bus signals Range 0~63
CsOdtSetup - (same help string)
CkeSetup - (same help string)​
CAD BUS DRIVE STRENGTH CONFIG:
ClkDrvStrength - 120Ohm/60/40/30/24/20 - (no help string)
AddrCmdDrvStren- 120Ohm/60/40/30/24/20 - (no help string)
CsOdtDrvStren- 120Ohm/60/40/30/24/20 - (no help string)
CkeDrvStren- 120Ohm/60/40/30/24/20 - (no help string)
​

EDIT: Eeh... Either my system booting with v1.72 at 3200 15-15-15 was a fluke like my rig posting with 3466, or 1.73 has once again lost the capability of Odd numbered CAS latencies :| I set 15, it runs 16. I know for a fact that 15 worked that one time, as even CPU-z reported it. :(

Other thing of note is now they're auto applying the CPU-NB with 1.2V when using 3333 or above :\ If AMD recommended not going higher than 1.2V, I'm not sure I'm onboard with applying so much voltage. Maybe it's just me though. At least the difference between 15s and 16s on the main timings, doesn't result in any performance impact. So, time to see if I can tighten up the Subtimings.
 
Last edited:
I could have sworn that was the description for 2.3? or am i mistaken? I have been waiting for an update for a while, still on 2.1 because that was the last actual update instead of just software reporting. Would love for my Cool and quiet and memory to work please bring out the 1.06 faster asrock

Edit - Feels like they just adjusted the number to bring it up inline with more of their boards, like my other board the b350 pro4 which is 2.4 as well for the same update.
 
gigaxtreme1 said:
Gigabyte beta BIOS appears to be the only one officially to have AGESA 1.0.0.6 code.
Click to expand...
What's the version that the BIOS reports? For Beta 1.72 and 1.73 (5/22 and 5/26 build dates) the Microcode is 8001126. Last one I remember looking at before that, after 1.0.0.4a came out, was reading 800111C.
 
Formula.350 said:
What's the version that the BIOS reports? For Beta 1.72 and 1.73 (5/22 and 5/26 build dates) the Microcode is 8001126. Last one I remember looking at before that, after 1.0.0.4a came out, was reading 800111C.
Click to expand...


I am sure they are agesa 1.06. The moderator on ther MSI forum verbally attacked me when I said HWinfo64 said 1.71 was 1.04a. Apparently the developers forgot to change the agesa identification string. He said I was deliberately spreading false information and banned me for a week. A bit sensitive I would say. Formula, does anubody understand the new settings in 1.73 and if they offer any overclocking advantage especially for 4 dimms??? With 2 dimms I am at 3333mhz with CL14-14-14-14-35 2T on DDR4 3200 FlareX. No real performance advantage in fps on Cinebench probably because of the 2T command rate.
 
The gigabyte K5, K3 and some of the other boards without (many) updates received a beta bios today, finally.
Agesa 1006, ez-oc, new memory registers, moar memory multipliers, and iommu grouping.

Word from other k5 users is it still lacks LLC, absolute voltages, etc..
 
cageymaru said:
AsRock just updated the bios for the X370 Killer SLI/ac to version.... 1.0.0.4a AGESA! Who else is excited for 1.0.0.4a!
http://www.asrock.com/mb/AMD/X370 Killer SLIac/#BIOS
Click to expand...
Formula.350 said:
Ooh ooh! ME ME!! .... Last month... :\ lol
Click to expand...

I keep thinking someone has had to seen this cluster fuck coming at AMD or at AIB , like the over dramatic slow motion effect you see in movies the accident prolly seemed funny at the time lets say 6 months back.
AMD: hey lets release AM4/Ryzen as a financial surprise for our stock holders to finish of the year, at the same time screw over our AIB and customers at the same time they get to clean up our mess and we will be so happy after we did this (everyone cheering at this point in time).

I mean how can you run a company and not see this as a result of your actions, the only silver lining is that it is a good product and in the end we will get a good bios but can't AMD hire at least some people which do have a clue on what they are doing ?
 
Pieter3dnow said:
I keep thinking someone has had to seen this cluster fuck coming at AMD or at AIB , like the over dramatic slow motion effect you see in movies the accident prolly seemed funny at the time lets say 6 months back.
AMD: hey lets release AM4/Ryzen as a financial surprise for our stock holders to finish of the year, at the same time screw over our AIB and customers at the same time they get to clean up our mess and we will be so happy after we did this (everyone cheering at this point in time).

I mean how can you run a company and not see this as a result of your actions, the only silver lining is that it is a good product and in the end we will get a good bios but can't AMD hire at least some people which do have a clue on what they are doing ?
Click to expand...

I could agree with that especially with how far away the vaga launch was.. I also think they were scared of what intel had up their sleeve if anything detailed about ryzen got out to the public way before launch.. I mean this was probably one of the most tight lip releases I've seen in a long time for a completely new architecture from any of the major companies. So they decided the potential backlash was less of a risk then intel releasing something that could screw up their ryzen release. Either way I think they were damned if they do damned if they don't as far as releasing ryzen when they did.

luke51087 said:
I could have sworn that was the description for 2.3? or am i mistaken? I have been waiting for an update for a while, still on 2.1 because that was the last actual update instead of just software reporting. Would love for my Cool and quiet and memory to work please bring out the 1.06 faster asrock

Edit - Feels like they just adjusted the number to bring it up inline with more of their boards, like my other board the b350 pro4 which is 2.4 as well for the same update.
Click to expand...


There's a 2.34 and 2.36 beta versions floating around for the Taichi and gaming boards.. people that have used them have seen good results with the 2.34 version but not the 2.36 version.. so it'll depend on which one they go with. But I'd suspect the official update is coming soon.
 
Last edited:
sirmonkey1985 said:
I could agree with that especially with how far away the vaga launch was.. I also think they were scared of what intel had up their sleeve if anything detailed about ryzen got out to the public way before launch.. I mean this was probably one of the most tight lip releases I've seen in a long time for a completely new architecture from any of the major companies. So they decided the potential backlash was less of a risk then intel releasing something that could screw up their ryzen release. Either way I think they were damned if they do damned if they don't as far as releasing ryzen when they did.




There's a 2.34 and 2.36 beta versions floating around for the Taichi and gaming boards.. people that have used them have seen good results with the 2.34 version but not the 2.36 version.. so it'll depend on which one they go with. But I'd suspect the official update is coming soon.
Click to expand...
Sounds good i am just going to wait for my killer/ac board to get the official one
 
Well, I just updated my MSI Tomahawk B350 Bios to the beta 1.62 with I can only assume is AGESA 1.0.0.06 and my Corsair Hynix chips 3200 is running at 3200 finally. Broke PS2 keyboard support, but switched to USB keyboard and it's working. Yay!
Edit- PS2 keyoard support is probably still working. Turns out the PS2 connector fell apart and was stuck in the PS2 port on the Bulldozer machine.
 
Last edited:
os2wiz said:
I am sure they are agesa 1.06. The moderator on ther MSI forum verbally attacked me when I said HWinfo64 said 1.71 was 1.04a. Apparently the developers forgot to change the agesa identification string. He said I was deliberately spreading false information and banned me for a week. A bit sensitive I would say. Formula, does anubody understand the new settings in 1.73 and if they offer any overclocking advantage especially for 4 dimms??? With 2 dimms I am at 3333mhz with CL14-14-14-14-35 2T on DDR4 3200 FlareX. No real performance advantage in fps on Cinebench probably because of the 2T command rate.
Click to expand...
hah There are plenty of people who know... Chew is one of them :) However, if you saw my post (on the same page you posted on), you'll see what the problem is... Most of the big time overclockers who really, truly know the ins and outs of these settings, keep that information close to the vest. It sucks, but slowly, I think we'll end up finding out the things we need to know.

He occasionally shares some stuff. It sucks the thread is so long. If you just search of my replies, I'm usually the "idiot" asking about everything, and so he will reply to me and give some nuggets. For some it might be helpful. Often times for me, it isn't QUITE the answer I'm hoping for, as often times I don't know what the "ideal" setting/value is to start with. But I had shared those settings that they added, on the previous page that I replied to you on. In short, he said the lower Drive Strengths should aid compatibility... however, did follow up that for AMD it had generally been worded and not listed in a value, so he isn't sure quite how it functions with them, pointing out that TheStilt would be the guy to ask (I think he hands out on Anandtech now). I really really wish I knew more, not just for my own sake, but to spread that information wealth. heh
 
Formula.350 said:
hah There are plenty of people who know... Chew is one of them :) However, if you saw my post (on the same page you posted on), you'll see what the problem is... Most of the big time overclockers who really, truly know the ins and outs of these settings, keep that information close to the vest. It sucks, but slowly, I think we'll end up finding out the things we need to know.

He occasionally shares some stuff. It sucks the thread is so long. If you just search of my replies, I'm usually the "idiot" asking about everything, and so he will reply to me and give some nuggets. For some it might be helpful. Often times for me, it isn't QUITE the answer I'm hoping for, as often times I don't know what the "ideal" setting/value is to start with. But I had shared those settings that they added, on the previous page that I replied to you on. In short, he said the lower Drive Strengths should aid compatibility... however, did follow up that for AMD it had generally been worded and not listed in a value, so he isn't sure quite how it functions with them, pointing out that TheStilt would be the guy to ask (I think he hands out on Anandtech now). I really really wish I knew more, not just for my own sake, but to spread that information wealth. heh
Click to expand...


I just raised the overclock on my 4 dimm configuration of FlareX to 2800mhz on Titanium bios 1.73. CL14-14-14-14-48 CR T2. I used proc 60 ohms. SOC v. 1.15 v, dram voltage 1.40 v , BGS enabled, VDDR 1.0 volts. It appears stable I have run aida64,cinebench 15, cpuz bench. I will run memory stress test tomorrow. Do you know where I can download GSAT memory test??
 
Formula.350 said:
hah There are plenty of people who know... Chew is one of them :) However, if you saw my post (on the same page you posted on), you'll see what the problem is... Most of the big time overclockers who really, truly know the ins and outs of these settings, keep that information close to the vest. It sucks, but slowly, I think we'll end up finding out the things we need to know.

He occasionally shares some stuff. It sucks the thread is so long. If you just search of my replies, I'm usually the "idiot" asking about everything, and so he will reply to me and give some nuggets. For some it might be helpful. Often times for me, it isn't QUITE the answer I'm hoping for, as often times I don't know what the "ideal" setting/value is to start with. But I had shared those settings that they added, on the previous page that I replied to you on. In short, he said the lower Drive Strengths should aid compatibility... however, did follow up that for AMD it had generally been worded and not listed in a value, so he isn't sure quite how it functions with them, pointing out that TheStilt would be the guy to ask (I think he hands out on Anandtech now). I really really wish I knew more, not just for my own sake, but to spread that information wealth. heh
Click to expand...


My general position is that beneficial knowledge shoiuld NEVER be the property of a few but collective sharing of information is beneficial to society. Less competition and MORE collectivity would make this world a better place.
 
Quick question: If I am on BIOS 1.1 (hypothetical), and BIOS 1.2 comes out with AGESA update, and then 1.3 with some other changes, will going from 1.1 to 1.3 update the microcode as well? Seems like it should be no problem skipping BIOS to the latest, but just wanted to make sure.
 
mnewxcv said:
Quick question: If I am on BIOS 1.1 (hypothetical), and BIOS 1.2 comes out with AGESA update, and then 1.3 with some other changes, will going from 1.1 to 1.3 update the microcode as well? Seems like it should be no problem skipping BIOS to the latest, but just wanted to make sure.
Click to expand...

yes but you can usually check by looking at the file size as well, if it's significantly smaller than the previous patch then it may not have it.
 
sirmonkey1985 said:
yes but you can usually check by looking at the file size as well, if it's significantly smaller than the previous patch then it may not have it.
Click to expand...

Something you might expect from beta bios , but those final ones should have latest AGESA and from the official support page for your motherboard.
 
I'm curious to know if anyone is running a B350 Gaming Pro Carbon with a Ryzen chip. Also about BIOS issues and all that jazz.
 
Not using that one, but the Tomahawk B350. I chose that one because it was cheap and heard good things about the AXMP profile for getting higher memory speeds. My Hynix chips Corsair are now running at rated 3200 speed, with the new beta BIOS. Have to go to the MSI beta BIOS forum to get the betas so far..
 
sirmonkey1985 said:
yes but you can usually check by looking at the file size as well, if it's significantly smaller than the previous patch then it may not have it.
Click to expand...
I don't know if that's a fair assessment of whether a BIOS includes an updated AGESA or not. For instance, they could have added new code for menus, but removed a ton of irrelevant code or Strings that now makes for roughly the same size compressed file. (NOTE: The only way to compare them as he's stating is by the archive, not the uncompressed BIOS, as they all are 16,384KB due to 'padding')

Similarly, a BIOS with a new AGESA and tons of removed Strings might end up being smaller.

Alas, it will likely influence the size. Beta 1.62 with 1.0.0.4a on my Titanium was 8,661KB and the Beta 1.72 with 1.0.0.6 was 8,760KB.
MSI doesn't post the change logs on their Betas though :( So I have really no clue what changes in instances like now when from 1.73 to 1.74, the file size has only changed 300 bytes lol


AND on that note...

MSI Released Beta 1.74 for the X370 Titanium sometime late yesterday :)
https://forum-en.msi.com/index.php?topic=283344.msg1615498#msg1615498
I looked through it in AMIBCP and I can't really say I see anything new/different, so I assume it's stuff like changing the AGESA String to now actually display 1.0.0.6 instead of .0.4a heh
 
I installed the MSI Titanium 1.6 BIOS and boot times are much better, in the 10 - 15 second range which is great.

Unfortunately, I had an issue where the machine wouldn't boot, just a black screen forever. I reset BIOS and rebooted a bunch of times and finally got into Windows. But that makes me a little nervous.
 
I've found that you have to load defaults in the BIOS and it will clear bad settings, then start tweaking the settings. It helped recovering memory settings for XMP on the Tomahawk.
 
KD5ZXG said:
https://www.asus.com/us/Motherboards/PRIME-B350-PLUS/HelpDesk_Download/
PRIME B350-PLUS BIOS 0803
Beta BIOS with AGESA 1.0.0.6
Click to expand...

Thanks for posting that, I have been waiting for an update.

Memory compatibility does seem better. I tried two sets of RAM that previously had issues:

F4-3200C16C-16GTZ - TridentZ, no RGB, mine is Samsung D-die, and double rank, so it didn't work well in the past. It was best just to leave it at 2133, but I could only get it to 2400 before, and because I changed something, then it switched from 2T to 1T which made me nervous.

It will not run at 3200 with the new bios, so I just backed down to 2666, and I figured out how to set 2T. Its not the most logical. You have to just go to the line that says something about "Gear Down" and pick disabled. Don't also pick 2T because then it will set the RAM to 1T. Yeah, took me a while to figure that out. Not logical at all. Again, just disable Gear Down only for 2T.
Anyway, still a little speed upgrade from before, and I may be able to run 2933, I just didn't try yet. There was also a 2800 multiplier I think, so there are other options for me to play with.

edit: Just changed the speed to 2933 and it booted into windows just fine. 2933 vs 3200 isn't bad at all so I'm pretty happy for now.


F4-3000C16D-16GISB - Hynix MFR, single rank. When I first got this sometimes it would run at 2933, but over time it would lock up and have issues so I had to go back to 2133. With the new bios it seems like 2933 might be fine. It booted into windows no problem, but I haven't done any further testing. The same 1T vs 2T issues apply to this stuff as above.
 
Last edited:
looks like the asrock x370 gaming K4 got the official AGESA 1.0.0.6 bios update.. lets see how long it takes for the other boards.

rest of them are slowly being added now to their website.
 
Last edited:
Finale BIOSe AGESA 1006

X370 Professional Gaming - BIOS 2.40
X370 Taichi - BIOS 2.40
X370 Gaming K4 - BIOS 2.50
X370 Killer SLI - BIOS 2.50
AB350 Pro4 - BIOS 2.60
AB350 Gaming K4 - BIOS 2.60
AB350M Pro4 - BIOS 2.50
 
cybereality said:
I installed the MSI Titanium 1.6 BIOS and boot times are much better, in the 10 - 15 second range which is great.

Unfortunately, I had an issue where the machine wouldn't boot, just a black screen forever. I reset BIOS and rebooted a bunch of times and finally got into Windows. But that makes me a little nervous.
Click to expand...
Yea I had the same problems with 1.6 Beta (I think it was 1.6?), so I ditched it promptly.
It's a black screen with a flashing cursor in upper left corner? Don't remember which error code was present but ended up being at the Video Initialization stage.
Personally, 1.7x has been a godsend, finally back to the "just works" point that I had with v1.10
 
OK. Good to know it's not just me. I'll probably just leave the computer on until the next official BIOS is out.
 
Formula.350 said:
Yea I had the same problems with 1.6 Beta (I think it was 1.6?), so I ditched it promptly.
It's a black screen with a flashing cursor in upper left corner? Don't remember which error code was present but ended up being at the Video Initialization stage.
Personally, 1.7x has been a godsend, finally back to the "just works" point that I had with v1.10
Click to expand...

Yes since 1.73 and 1.74 the ability for two dual rank B-die dimms to just A-XMp to 3200mhz is a godsend. But achieving STABILITY on the memory is anything but easy, II stretched every tool to its limits to stabilize at 3200mhz and in the end rolled a bit back to 3067mhz now the instability is history. I'll stick here for a while until something changes in the next bios revisions that will improve stability further.
 
cybereality said:
OK. Good to know it's not just me. I'll probably just leave the computer on until the next official BIOS is out.
Click to expand...
I mean, I do. :) Sorta lol I sleep it, but yea, it doesn't get shut off (restarted) unless there's a crash, a BIOS to flash, or I've gotten an itch for SubTiming tweaking. lol

Which on that note... 3200 at 12-12-12-26 didn't work lol Not that I figured it would.


os2wiz said:
Yes since 1.73 and 1.74 the ability for two dual rank B-die dimms to just A-XMp to 3200mhz is a godsend. But achieving STABILITY on the memory is anything but easy, II stretched every tool to its limits to stabilize at 3200mhz and in the end rolled a bit back to 3067mhz now the instability is history. I'll stick here for a while until something changes in the next bios revisions that will improve stability further.
Click to expand...
Your Flare-X are DR? o_0 Why would G.Skill make 8GB DIM--- oh, wait, you're talking about your new 2x16GB Trident Z kit aren't you? If that's the case, 3066 isn't too shabby :)
 
Just updated my X370 Professional Gaming, and just wow, I am running 3200 on 64gig of tridentZ memory now, cinebench finishes without issue. no other overclocking otherwise, yet... perhaps I will try that tomorrow.
 
Wow, that's good you got dual rank to run that high. These new bios are starting to perform.
 
Last edited:
Formula.350 said:
I mean, I do. :) Sorta lol I sleep it, but yea, it doesn't get shut off (restarted) unless there's a crash, a BIOS to flash, or I've gotten an itch for SubTiming tweaking. lol

Which on that note... 3200 at 12-12-12-26 didn't work lol Not that I figured it would.



Your Flare-X are DR? o_0 Why would G.Skill make 8GB DIM--- oh, wait, you're talking about your new 2x16GB Trident Z kit aren't you? If that's the case, 3066 isn't too shabby :)
Click to expand...


Yes the Trident Z 32GB kit. Yes I can live with it. Perhaps I will get to a stabel 3200mhz soon. The fun ny thing is ifeverything is to default before I load A-XMP , I have zero proble of booting into 3200mhz. It is just remaining there. With SOC at 1.175 and dram voltage at 1.40 volts nad ProcODT at 60 ohms, I have no intention of pushing these voltages higher for 24/7 use.
 
os2wiz said:
Yes the Trident Z 32GB kit. Yes I can live with it. Perhaps I will get to a stabel 3200mhz soon. The fun ny thing is ifeverything is to default before I load A-XMP , I have zero proble of booting into 3200mhz. It is just remaining there. With SOC at 1.175 and dram voltage at 1.40 volts nad ProcODT at 60 ohms, I have no intention of pushing these voltages higher for 24/7 use.
Click to expand...
If you're not running the 1.7X Beta BIOSes, then I really do suggest trying them!

I managed to get 3466 into Windows today, but it required 1.47V on DRAM, which was more than I wanted to bother with beyond testing. Unfortunately it BSoDed before I could get a Cache/Mem run in lol I'm on 3200 with Chew's tightened 3333 timings. Didn't last but 25mins in AIDA Cache and RAM stress, but it did make it through Pi 32M.
Substantial gains to be had from just default CPU speed, bandwidth wise, and I'm pretty certain that they're STILL tighter than what 3066's SubTimings are. He uses an x370 Taichi, but they still worked on my Titanium.

Just be sure to set them ALL AT ONCE. Some of these are reliant on others and caused mine to not POST until I set them all manually before saving. (Everything after tCKE are not available to change on the Titanium)

Chew-3333-User-Timings.jpg


Be sure to read his post though, as he outlines voltages and makes some other tips for it.
I can't really comment on his LLC settings. Only thing I've changed on mine is the Power options, changing them to "Current Balanced". (no clue if that's needed, or helps, but meh)
 
It sounds as if the bios is getting slightly better each time allowing bigger step to be made. Lets hope they keep making this progress each month we might end up with something beautiful before the end of the year ;) .
 
You must log in or register to reply here.
Back
Top