• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Any ASUS Prime X370 Pro owners out there ... besides me??

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
Well I finally got too impatient awaiting an "official" AGESA 1.0.0.6 BIOS and installed the Beta 0803. Have to endorse Mackerel's opinion ..... !Wow!. There are a ton of memory settings to play with now. I don't understand 90% of them. I cheated and copied some of the Stilts settings for my G.Skill F4-3600C16-8GTZ memory. I have it running at 14-14-14-30-52-1T @ 3333 Mhz @ 1.375V. I was able to bump my CPU one more notch closer to 3.9 Ghz. I just can't get 3.9 Ghz stable for more than a couple of hours. I have to increase Vcore too much and that actually impairs the cpu and memory benchmark speed. I have it running 24/7/365 stable at 3.875 Ghz @ 1.381V in the BIOS which droops down to 1.319~1.325V SVi2 in HwInfo at full load on BOINC running mostly AVX computations. I think that this is where my chip peters out. I don't see that much of a performance increase for that last 50 Mhz. I think I will stop where I'm at. I am happy so far with BIOS 0803 and am glad it is working well. I was dreading issues of having to rollback if it didn't go well. I don't think that will be necessary. Wondering what, if any, more things they will add to an "official" 1.0.0.6 BIOS.
 
I'm guessing they released it as beta to check for bugs which might require fixing. I wouldn't expect new functionality in a final version, although there might be other minor updates.
 
I know that with the CHVI, there were a few bugs in 1401 that were addressed in 1403 the latest Beta which came out today or yesterday
 
I swear it was marked at beta initially

Keith Myers said:
Well as usual, my timing is impeccable. Soon as I give up waiting for the official release and install the latest beta, ASUS releases another beta BIOS. Marked as AGESA 1.0.0.6a. No information on what has changed from previous though.

http://dlcdnet.asus.com/pub/ASUS/mb/SocketAM4/PRIME_X370-PRO/PRIME-X370-PRO-ASUS-0805.zip
Click to expand...

Just looked through the download sections again because of all the people reporting it in different platforms that are always different from where I always find it. See that it is now NOT marked as beta. So this I guess means it is the "official" AGESA 1.0.0.6a BIOS.

Sure would like to know the changelog from last revision.
 
I haven't tried out the 805 BIOS yet. It's getting mixed reviews over on OCN Prime X370 thread ... from no change or detriment to can't reach previous memory clocks. I am going to sit on it a bit more until the dust settles. Perfectly happy with the beta 803 BIOS. R7-1700X @ 3.875 Ghz @ 1.381V. 16GB 3333 Mhz - CL14 @ 1.375V. BOINC stable.
 
Flashed the official non-Beta AGESA 1.0.0.6a BIOS for the Prime X370 yesterday. Seems to be about the same as the 803 BIOS though I do think that it needs more CPU Vcore to be stable compared to previous settings on 803. All voltages monitored with SIV and HwInfo64. No apparent changes in memory stability or voltages needed to pass SAT for 1 hour. I had tried 3333 for a while but it wasn't BOINC stable for more than a day and really there isn't that much difference in performance over my optimized CL14 3200 memory which is week long stable. I think I am staying put on 3200 for the memory. I was running 3.9 Ghz on the cores for a day running BOINC and thought I was stable but found the computer black screened eventually. Backing off again to 3.875 Ghz which seems to be BOINC week long stable. I might give 3.9 Ghz another go with another bump to Vcore. I am currently at 1.40V LLC1 Optimized Phase which droops down to a low of 1.331V. I tried LLC2 for a day at 1.40V which drooped down to 1.344V but it black screened eventually. I'm not sure I want to use more Vcore as I am already at worst-case 85° C. Tctl BOINC loaded with my Corsair H110i in Performance Mode. OF course that temp depends on the mix of work on the system. The more typical temps are in the 78-82° C. range. VRM temps seem to be fine however and have never been over 52° C. and that is where the PCH temps seem to run at also.

So no full-stop roadblocks for 805 BIOS in my opinion.
 
Believe that I am done messing with systems. The Ryzen system is BOINC stable at 3.875 Ghz @ 1.381 Vcore. The memory is stable at 3200 Mhz. I did reconfigure the AIO though after reading a more technical review of the Corsair H110i than the usual review. It went into details about how little coolant volume the pump moves and how it has trouble with as little as a 12" gravity head. That information led me to move the radiator out of the roof of the Corsair Air 740 case to the front with fans pushing outside air through the radiator. That gained me about 2-4 ° C of CPU temperature improvement. That little experiment also prompted me to move my Corsair H105 radiators from the roof to the front in my FX systems in Corsair Air 540 cases. That has led to a massive 10-15° C. CPU temperature drop in those systems. I had no idea that would improve things that much. The move of the radiators to the front of the cases has caused about a 2-4° C. increase in the dual GTX 1070s GPUs in all systems though but nothing that is alarming. They all run around 60-65° C. fully loaded with BOINC jobs and are still able to pull off about 300 Mhz GPU Boost core speeds over stock clocks. I am very happy with all systems now.

Now I just have to wait for news about Threadripper shipping dates, prices and performance for the eventual upgrade of the FX systems. Or go with more Ryzen CPUs.
 
I see there is another official BIOS for the Prime X370 Pro. Version 0807 is out. It has enabled all the CBS settings that the Crosshair VI Hero motherboard owners have had from day one. A mixed commentary from trial users so far with some bricked boards and others reporting that XMP memory profiles working out of the box with no BIOS intervention. A lot of new BIOS options to play with but unsure whether any had made it easier to overclock your memory or whether memory overclocks are more stable. A few reports that the higher overclocks, like Stilts Fast 3200, Fast 3300 and Fast 3466 Mhz memory timings are working without too much fiddling.

I am not going to try it out until I have finished the WOW contest which starts up on Monday. Not going to chance losing one of my fastest and most productive crunchers the weekend before the start of the contest. I will play with the 0807 BIOS afterwards and have a couple of weeks of others testing it out before I make the leap of faith. Am stable at 3.875 Ghz for the CPU and 3200 Mhz and CL14 Fast settings for the memory.
 
I just put 0807 on mine, and some new ram at the same time. The new ram is Corsair LPX 2800c16 sold as a 4x4 kit. The interesting thing is, when I looked at the SPD, it showed two XMP profiles. 2800 at 1.2v, and also 3000 at 1.35v. Win!

The intention was to split the kit into two pairs, one each for my Ryzen systems as neither currently has any as I had to rearrange my old ram for my 7800X system. Since it was nearer, I first put it in my Asus B350M-A system with 1600 in it. Set it to 2800 profile, and it booted fine. Compared a quick P95 bench compared to 2666, no significant difference. The CPU wasn't saturating the ram so no point setting 3000 and I left it at 2800.

I put the other pair in the X370, still on bios 0805 at this point. It booted with ram at 2133, and I updated the bios. So far so good. I looked at CBS... that looks a bit more than I want to tinker with so I left it alone. I set ram to DOCP 2800 and... failed boot. Ok, let's try the 3000 profile. Still no boot. Back to 2800 profile but back off to 2666, booted fine. Tried 2800 once more, no boot. It isn't happening at default settings at least. This is annoying, as my X370 has the 1700 in it, and with more cores it might use the ram bandwidth to help feed it. Think I'm not going to spend any more time with this ram in this system. I have LPX 3000 2x4 elsewhere which I know runs at 2933 on the X370 on 0805. So I'll swap that over, and will see if the LPX 2800 works at either speed on the other system.

Even with improvements, high speed ram compatibility is still a bit hit and miss. Curious it worked on the B350M system but not X370.
 
mackerel said:
I just put 0807 on mine, and some new ram at the same time. The new ram is Corsair LPX 2800c16 sold as a 4x4 kit. The interesting thing is, when I looked at the SPD, it showed two XMP profiles. 2800 at 1.2v, and also 3000 at 1.35v. Win!

The intention was to split the kit into two pairs, one each for my Ryzen systems as neither currently has any as I had to rearrange my old ram for my 7800X system. Since it was nearer, I first put it in my Asus B350M-A system with 1600 in it. Set it to 2800 profile, and it booted fine. Compared a quick P95 bench compared to 2666, no significant difference. The CPU wasn't saturating the ram so no point setting 3000 and I left it at 2800.

I put the other pair in the X370, still on bios 0805 at this point. It booted with ram at 2133, and I updated the bios. So far so good. I looked at CBS... that looks a bit more than I want to tinker with so I left it alone. I set ram to DOCP 2800 and... failed boot. Ok, let's try the 3000 profile. Still no boot. Back to 2800 profile but back off to 2666, booted fine. Tried 2800 once more, no boot. It isn't happening at default settings at least. This is annoying, as my X370 has the 1700 in it, and with more cores it might use the ram bandwidth to help feed it. Think I'm not going to spend any more time with this ram in this system. I have LPX 3000 2x4 elsewhere which I know runs at 2933 on the X370 on 0805. So I'll swap that over, and will see if the LPX 2800 works at either speed on the other system.

Even with improvements, high speed ram compatibility is still a bit hit and miss. Curious it worked on the B350M system but not X370.
Click to expand...

Did you set the X370 system back to at least F5 defaults or clear CMOS entirely first before updating to 0807? I am going to the full clear when I update. There is so much more being put in play with all the new CBS options that I don't want anything left over in the BIOS from 0805. I've always gone back to F5 defaults at the least when updating the BIOS. When there was a big change like from AGESA 1.0.0.4 in 0604 to AGESA 1.0.0.6a in 0805, I've felt more comfortable doing the full CMOS clear to make sure nothing is left over.
 
I'm still convinced that AMD is still clueless about DDR4 and that leaves the motherboard manufacturers groping around in the dark making random changes. So I kinda got fed up and dumped all my Ryzen 5 and 7 parts for now. Still interested in a Ryzen mini ITX at some point though.
 
Finally got to the newest Prime Pro BIOS 0810. Wasn't going to chance anything while I was in the middle of the WOW contest. Numbskull did a bang up job and got to #24 in the Top 100 Hosts list at SETI. My old FX-8300 running Linux and the special app is currently at #8 in the list. Anyway, got 0810 installed and am doing better than I was before on the old 0805 BIOS. I never bothered with the interim 0807 BIOS. I am now running at a full 3.9 Ghz on the cores fully loaded with only a 0.01875 V offset added. This is down from the 0.3750 offset I used to run for my stable 3.875 Ghz clocked cores. I am also running The Stilts Fast 3333 memory settings at CL14 at 1.38V for the memory. I never could get 3333 or 3466 stable before. I still had no luck with 3466 on the memory and caused some system corruption so I backed down to my new stable 3333 Mhz CL14 settings. I ran the memory up to 1.40V for the 3466 trial but maybe it needed a lot more. The 0810 BIOS opened up a ton more settings in the AGESA 1.0.0.6 profile and now the Prime Pro probably has 90% of the BIOS configuration options that the Crosshair VI Hero has had from day one. I only availed myself of the BankGroupSwap setting though and kept most everything else on Auto. So, I would say that the 0810 is another incremental improvement in stability and DDR4 compatibility.
 
With changes in each bios I wonder if I should re-validate my OC, or even try pushing further. I had put the bios on shortly before I received the Vega 56, and had stability problems from unknown sources. That it happened once when not under GPU load made me question if it was the system not the GPU. In short, un-did CPU/ram OC, everything was fine. Re-did OC for bench, everything remained fine. I wonder if there was some background setting that wasn't updating properly between updates, but this is just speculation for now. I've only seen it happen with SMT before during a previous update, which has options Auto and Disabled. It was Disabled before update, Auto selected after, was not giving me SMT until I set to Disabled, reboot, then back to Auto.
 
I still maintain that with such big changes in the BIOS from previous, that it is good practice to blow the BIOS back to default cleared status. You are writing so many more options with the 0810 BIOS that I figure you have to be re-using old memory addresses within the BIOS and a simple rewrite may not completely erase previous settings and you might end up with a mish-mash of defaults of the new options. I have no prove of that happening myself but I have read anecdotal stories in the Prime Pro thread over on Overclock.net of cases that suggest incomplete BIOS setup on a new BIOS. I have alway followed this practice and haven't been bitten yet. 'Knock on wood' Flashed the BIOS, waited for the message it completed and let it reboot the system. One memory training cycle and it popped right into Windows. Went back into the BIOS to see all the new settings and set back my previous stable 3.875 and 3200 CL14 settings and things tested out the same. So it was time to see if I could OC a wee bit more. Happy I could get a little bit more out of the overclock. I don't think I want to try pushing any harder since the room temp is north of 29° C all the time with the special cruncher running there too. The A/C has been cranking non-stop mostly for the past week since it has been 43-45° C outside for the past week. I will attempt 4.0 Ghz once winter arrives again and the ambient gets colder. My 280mm AIO is just not up to the job with such high ambients. With the A/C off, I have seen 36-37° C. on the radiator loop with fans at 100%. For a full time, fully loaded cruncher, I can see that nothing less than a 360mm radiator is really needed in the summer.
 
mackerel said:
With changes in each bios I wonder if I should re-validate my OC, or even try pushing further. I had put the bios on shortly before I received the Vega 56, and had stability problems from unknown sources. That it happened once when not under GPU load made me question if it was the system not the GPU. In short, un-did CPU/ram OC, everything was fine. Re-did OC for bench, everything remained fine. I wonder if there was some background setting that wasn't updating properly between updates, but this is just speculation for now. I've only seen it happen with SMT before during a previous update, which has options Auto and Disabled. It was Disabled before update, Auto selected after, was not giving me SMT until I set to Disabled, reboot, then back to Auto.
Click to expand...

I think I have seen lots of posts that the 0810 allows for higher clocks or equal clocks with lesser voltage. And a LOT more posts of people almost always being able to hit 2933 and usually 3200 memory with tightened timings. That's even with Hynix memory or Samsung D-dies and double ranked. There is a lot of recommendation of running lesser core voltage with LLC3 - LLC5 with no concerns now that there is any damaging overshoot. And the recommendation is to use Extreme on the phase control for better voltage control with minimal droop. I am using LLC3 and Extreme now for the CPU and the droop is minimal, still some, but a lot less than before. I'm being a little careful first before trying LLC5. And was able to use BankGroupSwap finally with the 0810 BIOS, something that I have been envious for a long time reading the CHVI thread.
 
Keith Myers said:
I still maintain that with such big changes in the BIOS from previous, that it is good practice to blow the BIOS back to default cleared status.
Click to expand...

I think you or someone else had mentioned that previously. I thought I wouldn't need it as at the time the system was practically stock anyway. I had undid the CPU OC and also turned off XMP so ram was also at safe defaults. That's a point, I would have had to re-apply the OC after the bios update, so that kinda blows a hole in my theory. I'll try to remember to keep it in mind for future updates as everything seems ok for now.

When I get some free time I'll revisit overclocking on it, and see if much has changed. Note personally I go for low voltage overclocks for 24/7 running, so I don't expect to see much change there. I occasionally go for max clock just to see what I can do, and there it might get more interesting.
 
I skipped the interim 0812 BIOS release because of all the negativity about ASUS removing all the new cool CBS menu options. Then 0902 showed up with with the same removal of CBS options but a few lingering ones down in the memory settings that still prove useful. I figured I would give it a try because all I ever used in the CBS menu in the 0810 BIOS was the BGS/BGSA option which I never found to be helpful/hurtful and probably wouldn't miss. I didn't and am still able to run my Stilt's Fast 3200 CL14 memory settings. One large benefit with 0902 is the apparent lowering of necessary Vcore for a desired clock speed. Comments have been about 50-50% mixed that matched my observation of lowered CPU voltage to no change observed. I have now lowered my Vcore from 1.369V at 3.875 Ghz to 1.344V at 3.9 Ghz with the same full load for 24/7 distributed computing. The reduction in volts also lowered my temps overall on all parts by a few degrees.
 
Hexxer said:
Found it out yeterday. It dosent work on 2400. Wonderfull....not. Set the values by hand even dosent work. It stops booting or give some strange failures.
Click to expand...

I have the exact same model number of the RAM you are using. It runs fine with DOCP settings, even got 2800 stable with 16-18-18-42 timings... Problem is your timings are off. Ryzen hates odd numbers in where CL is concerned... the timings should be 14-16-16-31. Those are the DOCP timings and should work just fine. the 15-15-15-36 are the standard timings of the RAM, not the DOCP or 'XMP' profile.

I get this RAM stable with 1.37v @ 2800. I can almost get 2933 to work, but I have to push 1.4+ volts to do so, and that's just a tad much for me.. I know it's suppose to be safe to push that much, but even with 1.2v the RAM ends up getting pretty warm during gaming. That seems to stress the RAM the most. Not even IBT gets the DIMMs that warm, even when doing a custom 12024MB run that takes almost an hour.

Main problem I'm having with this board is the clock. It loses time. Even Asus is confused after I recorded the time stuttering in the BIOS screen. It would freeze every 6 to 7 seconds. They have no idea why. It even continues after updating to BIOS 0902. I may end up having to RMA it, if they can't figure this out.
 
You must log in or register to reply here.

Similar threads

A
ASUS Prime X370-Pro
Replies
7
Views
1K
trents
trents
D
Overclocking problem on Asus Prime X370-Pro
Replies
1
Views
2K
Mandrake4565
Mandrake4565
JrClocker
Planning New AMD Threadripper Build
2
Replies
36
Views
3K
JrClocker
JrClocker
D
Asus Prime ZA270 doesn't boot with error USB device over current status detected
Replies
2
Views
1K
trents
trents
Lochekey
Vostok V1 Thermal Interface Material
Replies
9
Views
1K
Nebulous
Nebulous
Back