BIOS updates for AM4 motherboards

Pieter3dnow said:
Okay but how much of the hidden stuff is actually working , it is nice you can put values to things you want to change but if nothing happens ;) .
Click to expand...
Prior to 1.72, the things that have been there, but non-functional that I've found, have been: CPU VDDP, and SubTimings. BCLK works, albeit in the limited capacity it is able due to a lack of hardware. Meaning you have: 100, 101, 102, 103 and that's it. 100 seems to run much closer to 100MHz than "Auto" does (99.8MHz), where as 103 for me has seemed to result in the SATA controller flaking out which causes a Windows BootLoader error OR no POST. 101 and 102 have worked just fine though, which while not much, at least bumps the memory up a tiny bit more to just shy of 3300MHz.

BUT on the flip side... beings we're months into Ryzen, new settings even if unavailable w/o modding, is a good sign. It means we'll eventually (hopefully) see them added, or at the very least it shows us they're TRYING to add new features. :)

Nevertheless, I'm too eager to flash this thing, so I'm just going to change all the SubTiming's "View Access" to USER and dig into this puppy!

UPDATE: Alright so first things first... Flashed, rebooted, configured fans (I suspect my modding BIOSes results in the OC Profiles failing to apply, no biggie). Set DDR4-3200 (it does indeed list DDR4-4000 now!), set timing to 15-15-15-35 1T, set DRAM to 1.35V, saved and booted to Windows. (Not abnormal for my system to post perfectly fine at 3200 the first change after a BIOS Flash).

Right out of the gate, what's worth mentioning? I set the timings to 15-15-15-35 and CPU-Z reports it's RUNNING at 15-15-15-35! :D No more "Evens-Only" on the tCL (y)

Time to run some tests...
 
Last edited:
gigaxtreme1 said:
Gigabyte has some beta and test BIOS with AGESA 1005 code
http://forum.gigabyte.us/thread/886/am4-beta-bios-thread
Click to expand...
The AMD Twitch stream I watched earlier (was on VideoCardz.com or .net, whatever it is), they said the new AGESA is 1.0.0.6 for the latest that is rolling out.
My Titanium's 1.72 reads "8001126", whatever that translates into. Previously it was "1111" I think. That one was using 1.0.0.4a. The "dot6" (as the one AMD dude called it), is the one with the unlocked memory speeds and timings, so it's definitely what this 1.72 BETA BIOS has, despite AIDA64 reporting that I"m running 1.0.0.4a heh


Anyways, my REAL reason for posting was to relay some very welcomed news with this BIOS. I mean, first I guess would be that it woke up from sleep just fine, and performance was still there from all I could tell, which is always a plus lol
However, the bigger deal for ME is... no Cold OR Warm reboot issue! Every BIOS I've tried after Shipping ver 1.10 had resulted in restart issues; however, this time I was able to restart from Windows and had no troubles! I'm back in the BIOS on the first pass :D

Also of potential interest interest to Pieter3dnow, those two additional DRAM Voltage settings I had unhidden that you asked me about (see my image in post #318), they read 0, so I don't know what that exactly means. It might be that it's implying "Auto", or it might be that it flat ut doesn't work, I dunno. Given that my system POSTs and restarts just fine now withOUT monkeying around, I had no reason to touch it :p The CLDO_VDDP is set at "Auto" by default and has no reading next to it's setting (I had also unhidden the CPU_VDDP in this one as well just to see if anything changed, and probably hasn't. It's "reading" still reports 0.000V)
UPDATE: Alright I tried inputting a value on the BIOS and TRAINING, but was met with them being unable to have decimal values. It'll take whole numbers just fine, 1, 2, 3, 5, 10 (placed, I didn't restart with it that way!). I'm going to chalk them up as settings that'll be worked on and available a bit later. Bummer, but, progress is being made so that's good as well :)

Going to see what sort of overclocks I can get on memory now. I don't know if it's MSI's doing, or AMD's, but I can note that setting the DRAM's speed in the dropdown does change other settings (this isn't a "new" thing, it's been something I've noticed since at least BIOS v1.5). The "CPU NB/SoC Voltage" increases from the 0.900V up to 1.150V. Also, the fans ramp up for no reason as well. There's no temp change, and but they definitely seem to be increased by at least 1V based on what it took to bring them back down to what I HAD them running at. I get the impression that changing the DRAM above the AMD officially supported speed of 2666 might ALSO flip the "0C" Overclock Mode bit, except... if that was happening, then there'd be an obvious loss of Boost Mode in Windows (I run default Balanced plan [not AMD's]), which I can confirm is not happening, since it still ramps up to 3.9GHz as rated for.

Anywho, time to see what sort of clocks I can get out of this memory ^_^

EDIT: I tried MSI's goofily-named "Memory Try It !" option, with pre-defined overclocks and timings first. DDR4-3333 16-18-18-36 1T with its applied 1.36V worked. DDR4-3466 same timings did NOT work (it applies 1.4V but I knocked it down and re-tried up to that, still didn't help.)
They've provided presets up to 4000, with timings of 20-22-22-39 and 1.45V. The one site Pieter had linked back in Feb had taken these B-Die up to 1.8 or 1.9V in order to get >4000 out of 3200 TridentZs, but I don't think I'll attempt that with mine. Was said 1.5V 1.45V would be the max they'd use/suggest for daily operation, but I have a feeling mine probably won't even pull off 4000 at that voltage and even more relaxed timings. In time, we'll see :p 3466 comes first!

EDIT2: So 3466 17-17-17-38 1.4V was doable... at CR 2T, but not at 1T. UPDATE: Guess that was a one-time-only deal. I did everything I could, even tried recreating the steps I took to achieve it, but to no avail am I able to get 3466 to POST at ANY primary timings, voltage, or Command Rate. (Mind you I'm leaving every advanced Timing option on Auto, so I'm sure once I get familiar with DDR4's finer tunings, I'll manage to get it back to 1T). And who knows, maybe 1T isn't that big of a killer of performance like it was once upon time? I mean look at the way 3200 performed at 22-22-22 (I think tht was in the memory speeds thread that I posted it). Time to see how high 2T will take me for now. :)

EDIT3: INTEResting... I disabled the Memory Try It! preset and as such, it put everything back to Auto. I manually changed the timings to 18-18-18-40 (am attempting 3600), and was going to set the voltage when I noticed the DRAM voltage for the "Voltage(BIOS)" and "Voltage(Training)" now had values! Seems like the reason there isn't support for decimals is because they're rated in mV! The (BIOS) option shows 1360, which is in line with the auto 1.36V it applies when I select a DRAM speed that's fairly high (3200 I know for sure). The (Training) option now shows 1500, which I'd assume means 1.5V. Not sure why they only now decided to receive values, but it is indeed curious and merits further exploring :D Thanks, random BIOS anomaly!! lol
 
Last edited:
Pieter3dnow said:
Make that 1.45V for daily use and 1.5V for benching.
Click to expand...
Thanks. I've corrected my post :)

Settled on DDR4-3333 16-15-15-35 1T 1.376V (aka 100% Auto everything, which the voltage is just the result of being set to 1.36V and running higher).

Will throw these speeds through some benches quick and see how they fair, then play a game and knock out some birthday related stuff for a friend (which will require the system be functional heh)
 
gigaxtreme1 said:
Formula 350, I ain't gonna quote your post, but you could wrtite books. Nice to see progress on the platform!
Click to expand...
lmao I know, I talk type a lot... :shame:
It honestly comes from all my days on the internet, searching for info, coming across a forum post only....for.... it to be devoid of any useful details :cry: heh
THAT, or someone who responds and says "I'll let you know my results" or "I'll post my settings when I get home" and there's never a follow-up... lol

I could compile all my forum posts into a book though...
"The nigh-incoherent memoirs of a computer junkie..." :pompous:
 
Formula.350 said:
lmao I know, I talk type a lot... :shame:
It honestly comes from all my days on the internet, searching for info, coming across a forum post only....for.... it to be devoid of any useful details :cry: heh
THAT, or someone who responds and says "I'll let you know my results" or "I'll post my settings when I get home" and there's never a follow-up... lol

I could compile all my forum posts into a book though...
"The nigh-incoherent memoirs of a computer junkie..." :pompous:
Click to expand...

By the way this bios is still Agesa code 1.04a according to HWinfo644 which I trust implicitly. My guess is many of the new settings shown are not fully functional and will be enabled in a followup beta. That would explain why I have not mmade significant progresds in raising 4 dimm FlareX DDR4 3200 above my old 2667mhz from bios 1.6. I guess it will take another 2 weeks before a beta based on Agesa 1.06 appears for this board.
 
os2wiz said:
By the way this bios is still Agesa code 1.04a according to HWinfo644 which I trust implicitly. My guess is many of the new settings shown are not fully functional and will be enabled in a followup beta. That would explain why I have not mmade significant progresds in raising 4 dimm FlareX DDR4 3200 above my old 2667mhz from bios 1.6. I guess it will take another 2 weeks before a beta based on Agesa 1.06 appears for this board.
Click to expand...
Yea, but remember, just after 1.0.0.4 came out there were some BIOSes mis-reporting it through software because the Version String wasn't updated, even though it WAS still the actual code. Reason was due to the AGESA being buggy so they took the changes and manually applied them to the previous AGESA. That could indeed be the case here again, but I do know that the 8001174 string I mentioned is a fair bit higher in numbers than what it had been.

Where does HWiNFO say that actual version though? All I was able to find was the same hex string that is reported inside the BIOS.
 
Asus Prime X370 new BIOS 0612 is out, still AGESA 1.0.0.4.

-I can now cold boot at 3200 C14 (previously only 3200 C16). If I do a complete cold boot (cutting power to the system for a minute), it still power cycles itself 3 times on the first boot, but it no longer reverts to 2133 MHz RAM or displays "Overclock failed".
Normal shutdown/reboot works fine, no power cycling, delays or errors.

-SOC voltage automatically sets itself to 1.1V when you enable DOCP now. I guess this is how Asus "improve stability", but it seems a bit excessive. It was at ~0.937V before. I dialed it back to 1.0V manually because I see no reason to degrade the CPU for no benefit.
 
Formula.350 said:
Yea, but remember, just after 1.0.0.4 came out there were some BIOSes mis-reporting it through software because the Version String wasn't updated, even though it WAS still the actual code. Reason was due to the AGESA being buggy so they took the changes and manually applied them to the previous AGESA. That could indeed be the case here again, but I do know that the 8001174 string I mentioned is a fair bit higher in numbers than what it had been.

Where does HWiNFO say that actual version though? All I was able to find was the same hex string that is reported inside the BIOS.
Click to expand...


No it is confirmed by others that even the latest Asus bioses are all Agesa 1.04a. So this roll out is a dress rehearsal for Agesa 1.06 in a couple of weeks.
 
JimmiG said:
-I can now cold boot at 3200 C14 (previously only 3200 C16). If I do a complete cold boot (cutting power to the system for a minute), it still power cycles itself 3 times on the first boot, but it no longer reverts to 2133 MHz RAM or displays "Overclock failed".
Normal shutdown/reboot works fine, no power cycling, delays or errors.
Click to expand...
That's what my Titanium would do on the previous Beta from a few weeks ago. I thought maybe that was just MSI getting things wored out better with their BIOS, but sounds like maybe it was a minor update AMD pushed out in AGESA... Cuz that latter part of "cycles, but eventually boots with the correct settings" was exactly the same thing happening to me with v1.61 Beta. Every update after shipping BIOS, I had issue with warm-reboots with RAM at 3200, which up until 1.61 it meant I couldn't run that. Then it would do just as you said... Fail... Fail... Fail... POST. I'd go into BIOS and instead of being greeted with "2133", it was 3200. Even in Windows it'd report 3200, and benchmark results indicated that as well.

On this brand new one from a couple days ago, that no longer is the case, and it's totally fixed, no cycling. It finally, once again, "Just Works" as you'd expect :D


-SOC voltage automatically sets itself to 1.1V when you enable DOCP now. I guess this is how Asus "improve stability", but it seems a bit excessive. It was at ~0.937V before. I dialed it back to 1.0V manually because I see no reason to degrade the CPU for no benefit.
Click to expand...
Again, same thing sorta with the Titanium for the last month or so. Select DDR4-3200 (not A-XMP or XMP profile, just flat out picking that Memory Speed) and it'd ramp stuff up in the background. Reboot and the NB voltage would read 1.1V to 1.15V. So once more, I can't help but think that it's maybe AGESA related, beings we're experiencing the exact same things, but you're on an ASUS and I'm on an MSI (we both have an X370).

os2wiz said:
No it is confirmed by others that even the latest Asus bioses are all Agesa 1.04a. So this roll out is a dress rehearsal for Agesa 1.06 in a couple of weeks.
Click to expand...
I suppose it's plausible. Just seems to weird that they'd continue calling it 1.0.0.4 despite being such a night-and-day difference in system behavior.
Personally, I would be 100% on board with that theory, but what puts me off, is that not only are we given the huge array of DRAM speeds above 3200 now... but also are given SubTiming control. Those are the exact features of 1.0.0.6, which even that vid from the AMD employees said "I believe ASUS released a Beta for the Crosshair yesterday [which would've been like 5-6 days ago now at time of writing] that uses the dot-6 AGESA".

At the end of the day, I'm not going to split hairs! lol It has provided many new features. It has fixed problems for me. It hasn't created any new ones. The performance is where you'd expect it.
I. AM. HAPPY!!! :D(y)
 
It could be that AMD have gone out with common recommendations for all mobo manufacturers, which is why the behavior is similar across different boards, maybe? It's definitely still the same 1.0.0.4 AGESA code from what I can tell.

Anyway I'm happy with the changes too. Having the RAM randomly revert to 2133 when booting was a show stopper, but I can live with this minor inconvenience of the occasional automatic power cycling since it just takes a few seconds and doesn't require any input from me (like having to go into BIOS and double check the RAM speed and timings).
 
Last edited:
Formula.350 said:
That's what my Titanium would do on the previous Beta from a few weeks ago. I thought maybe that was just MSI getting things wored out better with their BIOS, but sounds like maybe it was a minor update AMD pushed out in AGESA... Cuz that latter part of "cycles, but eventually boots with the correct settings" was exactly the same thing happening to me with v1.61 Beta. Every update after shipping BIOS, I had issue with warm-reboots with RAM at 3200, which up until 1.61 it meant I couldn't run that. Then it would do just as you said... Fail... Fail... Fail... POST. I'd go into BIOS and instead of being greeted with "2133", it was 3200. Even in Windows it'd report 3200, and benchmark results indicated that as well.

On this brand new one from a couple days ago, that no longer is the case, and it's totally fixed, no cycling. It finally, once again, "Just Works" as you'd expect :D



Again, same thing sorta with the Titanium for the last month or so. Select DDR4-3200 (not A-XMP or XMP profile, just flat out picking that Memory Speed) and it'd ramp stuff up in the background. Reboot and the NB voltage would read 1.1V to 1.15V. So once more, I can't help but think that it's maybe AGESA related, beings we're experiencing the exact same things, but you're on an ASUS and I'm on an MSI (we both have an X370).


I suppose it's plausible. Just seems to weird that they'd continue calling it 1.0.0.4 despite being such a night-and-day difference in system behavior.
Personally, I would be 100% on board with that theory, but what puts me off, is that not only are we given the huge array of DRAM speeds above 3200 now... but also are given SubTiming control. Those are the exact features of 1.0.0.6, which even that vid from the AMD employees said "I believe ASUS released a Beta for the Crosshair yesterday [which would've been like 5-6 days ago now at time of writing] that uses the dot-6 AGESA".

At the end of the day, I'm not going to split hairs! lol It has provided many new features. It has fixed problems for me. It hasn't created any new ones. The performance is where you'd expect it.
I. AM. HAPPY!!! :D(y)
Click to expand...


We have been given the options but few are having success in achieving the desired speed. So that tells me we need Agesa 1.06 to deal the deal. I had zero improvement with my new Titanium 1.72 bios. I have found a bug in the 1.72 bios as well. Sometimes the bios will revert to 3.6 GHZ although it is set to 3.85. This will happen on boot and with no indication that there was any issue with instability. I can reboot again and it will stay at 3.6. Now if I change the setting to 3.825 it will boot at 3.825. If I go to bios again and set it back to my original overclock of 3.85GHZ it will then boot to it with no problem. So there may be a specific problem associated with the 3.85 multiplier.
 
Formula.350 said:
That's what my Titanium would do on the previous Beta from a few weeks ago. I thought maybe that was just MSI getting things wored out better with their BIOS, but sounds like maybe it was a minor update AMD pushed out in AGESA... Cuz that latter part of "cycles, but eventually boots with the correct settings" was exactly the same thing happening to me with v1.61 Beta. Every update after shipping BIOS, I had issue with warm-reboots with RAM at 3200, which up until 1.61 it meant I couldn't run that. Then it would do just as you said... Fail... Fail... Fail... POST. I'd go into BIOS and instead of being greeted with "2133", it was 3200. Even in Windows it'd report 3200, and benchmark results indicated that as well.

On this brand new one from a couple days ago, that no longer is the case, and it's totally fixed, no cycling. It finally, once again, "Just Works" as you'd expect :D



Again, same thing sorta with the Titanium for the last month or so. Select DDR4-3200 (not A-XMP or XMP profile, just flat out picking that Memory Speed) and it'd ramp stuff up in the background. Reboot and the NB voltage would read 1.1V to 1.15V. So once more, I can't help but think that it's maybe AGESA related, beings we're experiencing the exact same things, but you're on an ASUS and I'm on an MSI (we both have an X370).


I suppose it's plausible. Just seems to weird that they'd continue calling it 1.0.0.4 despite being such a night-and-day difference in system behavior.
Personally, I would be 100% on board with that theory, but what puts me off, is that not only are we given the huge array of DRAM speeds above 3200 now... but also are given SubTiming control. Those are the exact features of 1.0.0.6, which even that vid from the AMD employees said "I believe ASUS released a Beta for the Crosshair yesterday [which would've been like 5-6 days ago now at time of writing] that uses the dot-6 AGESA".

At the end of the day, I'm not going to split hairs! lol It has provided many new features. It has fixed problems for me. It hasn't created any new ones. The performance is where you'd expect it.
I. AM. HAPPY!!! :D(y)
Click to expand...

Your bugs were fixed but I see no improved ability for my 4 Flare-X dimms to achieve anything above the 2667mhz I achieved with bios 1.6. I have tried everything including loosening timings, adding 43 ohms, upping dram voltage to 1.40 v and soc/cpunb vottage to 1.10 v. No luck. By the way there is a bug in this new beta bios . It does not handle 38.5 multiplier properly.
 
os2wiz said:
Your bugs were fixed but I see no improved ability for my 4 Flare-X dimms to achieve anything above the 2667mhz I achieved with bios 1.6. I have tried everything including loosening timings, adding 43 ohms, upping dram voltage to 1.40 v and soc/cpunb vottage to 1.10 v. No luck. By the way there is a bug in this new beta bios . It does not handle 38.5 multiplier properly.
Click to expand...
Does 2 sticks run at 3200 at least?
 
os2wiz said:
We have been given the options but few are having success in achieving the desired speed. So that tells me we need Agesa 1.06 to deal the deal. I had zero improvement with my new Titanium 1.72 bios. I have found a bug in the 1.72 bios as well. Sometimes the bios will revert to 3.6 GHZ although it is set to 3.85. This will happen on boot and with no indication that there was any issue with instability. I can reboot again and it will stay at 3.6. Now if I change the setting to 3.825 it will boot at 3.825. If I go to bios again and set it back to my original overclock of 3.85GHZ it will then boot to it with no problem. So there may be a specific problem associated with the 3.85 multiplier.

Your bugs were fixed but I see no improved ability for my 4 Flare-X dimms to achieve anything above the 2667mhz I achieved with bios 1.6. I have tried everything including loosening timings, adding 43 ohms, upping dram voltage to 1.40 v and soc/cpunb vottage to 1.10 v. No luck. By the way there is a bug in this new beta bios . It does not handle 38.5 multiplier properly.
Click to expand...

Do you have the "Memory Retries" set to 0? If so, that'd likely explain something. I'm wondering if your issue isn't down to your 4 sticks of RAM causing the issues at start, and without a "Retry" set, it drops into Failsafe Mode right away. Beings you're using an 1800X, default clock is.... 3.6GHz! Does this happen with 2 DIMMs, or with memory set to 2133?

As far as bugs being fixed for me... what irked me was that v1.10 BIOS was very functional (for me), besides the Sleep Bug of course. I had no issues with 3200, or reboots not applying it. It's every BIOS afterwards that introduced problems. Granted, it fixed or improved Ryzen for other's, so that is good, considering I could (and did) indeed 'deal' by staying on 1.10. I'm sure you can appreciate the annoyance of things working out of the box and then ceasing to, given it put me into the same situation as you for the most part. We knew beforehand that there'd be issues with 4x DIMMs and/or Double-Rank DIMMs. So while that does suck and I empathize, it was still expected to be a potential issue :\ I will admit that, in this day and age, it shouldn't have been. That does worry me a bit about how Threadripper and EPYC will function, or if they have a different memory controller. Quad-Channel on TR with only 2667 for high density sticks would be a show stopper. Only thing that would save them slightly would be there being 8 DIMM slots (which was how the last AMD Workstation boards were, so it's quite possible). The fact we, with Ryzen, are capped at 64GB is kinda dumb. Sure, that IS a lot for a desktop platform, but... *shrug* I dunno, maybe they planned to take a page out of Intel's book of "Customer Shafting" and didn't intend for Ryzen to be used by folks for any kind of Workstation/CAD/Rendering, even though it's very good at it. Enter a 12 Core (maybe 10 even?) TR for maybe $600 with Quad-Channel, 8 DIMMs, and 256GB capacity... Time will tell!
 
Formula.350 said:
Do you have the "Memory Retries" set to 0? If so, that'd likely explain something. I'm wondering if your issue isn't down to your 4 sticks of RAM causing the issues at start, and without a "Retry" set, it drops into Failsafe Mode right away. Beings you're using an 1800X, default clock is.... 3.6GHz! Does this happen with 2 DIMMs, or with memory set to 2133?

As far as bugs being fixed for me... what irked me was that v1.10 BIOS was very functional (for me), besides the Sleep Bug of course. I had no issues with 3200, or reboots not applying it. It's every BIOS afterwards that introduced problems. Granted, it fixed or improved Ryzen for other's, so that is good, considering I could (and did) indeed 'deal' by staying on 1.10. I'm sure you can appreciate the annoyance of things working out of the box and then ceasing to, given it put me into the same situation as you for the most part. We knew beforehand that there'd be issues with 4x DIMMs and/or Double-Rank DIMMs. So while that does suck and I empathize, it was still expected to be a potential issue :\ I will admit that, in this day and age, it shouldn't have been. That does worry me a bit about how Threadripper and EPYC will function, or if they have a different memory controller. Quad-Channel on TR with only 2667 for high density sticks would be a show stopper. Only thing that would save them slightly would be there being 8 DIMM slots (which was how the last AMD Workstation boards were, so it's quite possible). The fact we, with Ryzen, are capped at 64GB is kinda dumb. Sure, that IS a lot for a desktop platform, but... *shrug* I dunno, maybe they planned to take a page out of Intel's book of "Customer Shafting" and didn't intend for Ryzen to be used by folks for any kind of Workstation/CAD/Rendering, even though it's very good at it. Enter a 12 Core (maybe 10 even?) TR for maybe $600 with Quad-Channel, 8 DIMMs, and 256GB capacity... Time will tell!
Click to expand...


Memory retries set to 5. I only keep the extra 2 dimms in when I am testing for higher mhz. This problem has occurred with only 2 dimms. Chew tells me it is an F9 boot loop.
 
AMD has published another Community Update.

Welcome to the fourth installment of the Ryzen Community Updates! If you’re checking into this series for the first time, this is where we let our community know about all the exciting updates that are on their way for the AMD Ryzen™ processor. We’ve covered a lot of ground in a short amount of time: game updates, new Windows® power plans, stability improvements, feature additions and much more. Today’s update is the one many of you have been most excited about: the AGESA that focuses on overclocked memory. There’s some great news for IOMMU/ACS users, too!



What is an “AGESA”?
AGESA is an acronym that stands for “AMD Generic Encapsulated System Architecture.” As a brief primer, the AGESA is responsible for initializing AMD x86-64 processors during boot time, acting as something of a “nucleus” for the BIOS for your motherboard. Motherboard vendors take the core capabilities of our AGESA updates and build on them with their own “secret sauce” to create the BIOS files you download and flash. Today, the BIOS files for AMD AM4 motherboards are largely based on AGESA version 1.0.0.4.



Beginning this month, as we promised to you, we began beta testing a new AGESA (v1.0.0.6) that is largely focused on aiding the stability of overclocked DRAM (>DDR4-2667). We are now at the point where that testing can begin transitioning into release candidate and/or production BIOSes for you to download. Depending on the QA/testing practices of your motherboard vendor, full BIOSes based on this code could be available for your motherboard starting in mid to late June. Some customers may already be in luck, however, as there are motherboards—like my Gigabyte GA-AX370-Gaming5 and ASUS Crosshair VI—that already have public betas.



Starting With Virtualization
If you’re the kind of user that just needs (or loves!) virtualization every day, then AGESA 1.0.0.6-based firmware will be a blessing for you thanks to fresh support for PCI Express® Access Control Services (ACS). ACS primarily enables support for manual assignment of PCIe® graphics cards within logical containers called “IOMMU groups.” The hardware resources of an IOMMU group can then be dedicated to a virtual machine.



This capability is especially useful for users that want 3D-accelerated graphics inside a virtual machine. With ACS support, it is possible to split a 2-GPU system such that a host Linux® OS and a Windows VM both have a dedicated graphics cards. The virtual machine can access all the capabilities of the dedicated GPU, and run games inside the virtual machine at near-native performance.



This is certainly a complicated setup for most users, but I have no doubt that there will be a whole lot of you enthusiastically nodding at this news. We’re grateful for your feedback and your patience, and we hope the new support for ACS serves you well.



What's Next For Memory
AGESA 1.0.0.6 officially adds 26 new parameters that can improve the compatibility and reliability of DRAM, especially for memory that does not follow the industry-standard JEDEC specifications (e.g. faster than 2667, manual overclocking, or XMP2 profiles).



The following table spells out all the new parameters, and provides a few words on what they do. Keeping in mind that this is overclocking territory, manual or automated control of these parameters should nevertheless make it a little more straightforward to use DDR4-3200 modules—or faster if you have the talent!1
Click to expand...

Not included in the quote is a lengthy table.
 
os2wiz said:
Memory retries set to 5. I only keep the extra 2 dimms in when I am testing for higher mhz. This problem has occurred with only 2 dimms. Chew tells me it is an F9 boot loop.
Click to expand...
Yea, that's the same error I get in older versions on my cold/warm restarts, as well as now on a failed higher RAM speed attempt. Though, curiously, I'm pretty sure that a 9 on these 7-segment displays do not carry the tail loop (like in this font)... and curiously, that's the exact way the "F9" is that I'm seeing. To me it looks more like Fg, but... beings that is outside of range of Hex values, I am doubting it :p Weirder still, is the F9 is listed (in the manual) as being "Reserved for future AMI Error codes".

I wouldn't be surprised if changing the ODT would help.... but I unfortunately have absolutely zero knowledge in those matters :( My last system had only like 60, 80, 120Ohm lol Before that it would've been "What's this ODT you speak of?" (then again I jumped from A64 x2 era to Ph II era).
 
Formula.350 said:
Yea, that's the same error I get in older versions on my cold/warm restarts, as well as now on a failed higher RAM speed attempt. Though, curiously, I'm pretty sure that a 9 on these 7-segment displays do not carry the tail loop (like in this font)... and curiously, that's the exact way the "F9" is that I'm seeing. To me it looks more like Fg, but... beings that is outside of range of Hex values, I am doubting it :p Weirder still, is the F9 is listed (in the manual) as being "Reserved for future AMI Error codes".

I wouldn't be surprised if changing the ODT would help.... but I unfortunately have absolutely zero knowledge in those matters :( My last system had only like 60, 80, 120Ohm lol Before that it would've been "What's this ODT you speak of?" (then again I jumped from A64 x2 era to Ph II era).
Click to expand...
Had my odt at 48 ohms and then 60ohms. I did once get to where the windows interface came up at 2933mhz but before entering it the system went black and I had to do a reset.
 
This may be a silly question, but can anyone say if the P2.5 BIOS for the ASRock AB350 Pro4 is stable? I'm planning on flashing but I've seen a lot of bricking issues with previous versions (they may have been beta builds so they may not be comparable)
 
I would wait for the new AGESA version coming soon. They just released the base code.
 
Last edited:
gigaxtreme1 said:
I would wat for the new AGESA version coming soon. They just released the base code.
Click to expand...

Will do. Besides the memory compatability, have the AGESA updates really improved overall system performance or has it been minor things here and there?
 
zach1928 said:
Will do. Besides the memory compatability, have the AGESA updates really improved overall system performance or has it been minor things here and there?
Click to expand...
1004a improved performance in a few games by 2-10% (depending on the game), but made some systems more difficult to OC. Not sure about later versions, but I know people are generally achieving higher ram speeds with them, which'll net you a few fps due to ryzen's design.
 
Nobu said:
1004a improved performance in a few games by 2-10% (depending on the game), but made some systems more difficult to OC. Not sure about later versions, but I know people are generally achieving higher ram speeds with them, which'll net you a few fps due to ryzen's design.
Click to expand...

Well thats good. Hopefully this will let my Team Vulcan sticks run at 2666. Also hope ASRock doesn't gloss over updating my board.
 
zach1928 said:
Well thats good. Hopefully this will let my Team Vulcan sticks run at 2666. Also hope ASRock doesn't gloss over updating my board.
Click to expand...
I dont think they will, I have that board in my plex server and i have the x370 killer and they pretty much follow the same bios updates. This one is too important for any mobo makers to pass up.
 
1006 beta is rocking pretty hard on my GA AB350 Gaming 3. It fixed the command rate bug (higher speeds got forced to 1T and it was not configurable) and added a bunch more options

Finally have my Mushkin dual rank 2x16GB at 2400 using XMP and voltages on auto

Had Forza crash a time or two but no system hang so not sure if it was the ram or not. Windows memory diagnostic was clean
 
luke51087 said:
I dont think they will, I have that board in my plex server and i have the x370 killer and they pretty much follow the same bios updates. This one is too important for any mobo makers to pass up.
Click to expand...
Have you had any success with RAM speeds above 2133 with the AB350?
 
hondaman said:
For you daring types, there appears to be BETA bios updates for the Taichi that aren't posted on their website. Discussion of that is talked about here https://forums.overclockers.co.uk/threads/the-asrock-thread-support-feedback-etc.18760683/page-30
Click to expand...


It is only daring for people who don't read up on changes. On Asus side of things they sometimes have notes on things that change and what is Beta now might be the same thing you download from the official website a few weeks later...
 
Pieter3dnow said:
It is only daring for people who don't read up on changes. On Asus side of things they sometimes have notes on things that change and what is Beta now might be the same thing you download from the official website a few weeks later...
Click to expand...
Just because the release notes are the same doesn't mean they didn't tweak a few user non-visible things...but if the file's checksum is the same that's a different story. ;)
 
You must log in or register to reply here.
Back
Top