since i installed this bios my fans are pulsing from fullspeed to lowest speed all the time. didn't had this with the official 1.0.0.4c bios. already changed the fan settings in bios to quiet and calibrated the fans. temps are as low as they can be.

the readout of >2800 rpm for one of the fans doen't change if the speed is reduced.

[File:bios.PNG]

Very weird, I never had such a problem.

@Screemer I've never experienced that issue. As a starting point I'd recommend you check none of your fans are faulty and that you do a full cmos reset (remove cmos battery, power cord, and set jumper to the "clear" position, press power button 2-3 times to discharge capacitors).

Zitat von barisuraz im Beitrag #28

---

I guess we can't do anything else until Biostar releases a new BIOS. I contacted support for a beta BIOS so if I get a new file I'll share it with you guys.

---

Nah, they told me that the one with 1.0.0.4c on the site is the latest one. Z55086 was my Question ID

i reset the bios but still got the issues with the fans running at full speed, going to low and immediately going to full speed again. tested with one fan on each header as well. same issue. flashed back to X37AG807.BST which i got from biostars tech support and the problem is gone.

looks like an issue with temp readings. as soon as the rpm of my fans drop hwinfo shows 37+ degrees celsius. just a moment later it shows 25 degrees c and the fans run at full speed again.

€dit: lol. i startet prime95 with ffts in place and avx on and the temps jump instantly and the fans went to low rpm. i quit prime, temp trops below a certain temp, which i can't exactly figure out, and boom fans run at full speed.

my workaround right now is to start prime with one thread and small ffts. that occupies the cpu a little and brings the temps to ~40°c and the fans spin on low. little annoying to blow energy for nothing but better then becoming deaf.

i use a custom watercooling loop which cools my rig in idle slightly above ambient. so 25-27°c is normal in idle. maybe that's why no one else is experiencing that issue.

€dit2: looks like the threshold is 33 or 32°C

@Screemer I tested every fan header on the board I have with 3 different fans and a few different AIO pumps and could not reproduce the problem you have. There are some points to note about Biostars smart fan though; firstly the values you input are Hex values which makes things very weird and awkward to start with (check the new descriptions I have put in for these options to clarify much better how the options work). The [Fan Ctrl FULL] option in the modded firmware allows you to easily set a maximum CPU temp at which preset and custom fan settings are ignored, in the default firmware this limit is set to the stupidly low value of 71c, I've increased it to 85c. Its also worth noting that the [Fan Curve Sensitivity] option can be a very fickle option depending on your setup adjust it in very small steps. Theres no reason why you would be experiencing any fan related problems with the modded firmware as that aspect as not been changed outside of superficial ones. @barisuraz is this a problem you have run in to or can reproduce?

In short, due to the absolutely bizarre manner of how smart fan works it is not a straightforward set and forget affair the very nature of how it works forces you to fiddle with it until it works as you intend it to. The values I set things to should be universal and work well for most people but certainly not everyone which is why I added much better descriptions for each setting as the ones provided by Biostar are poor at best.

Lastly that firmware esupport sent you is no different to the one you'll find on their support page :p

Like I said. No problems when I reflash the one I mentioned above, which I got mid of July before they uploaded it to their website.

I'll try to fiddle around with the smartfan settings but I don't think that I can fix it with it. Even if I disable smartfan completely I get the same behavior. Full speed below 33/34°C und lower speeds above that temp. Maybe the problem is that i am using daisychaned fans for my radiators and that way they don't deliver a proper pwm signal.

@ket no I haven't had any problems with the fans but I think anyone who owns this board or any Biostar board should cut their losses ASAP and buy a decent board. I'll be doing so in the near future.

We have a reply from Biostar @Screemer @Lost_N_BIOS @barisuraz. It is as dumb as ever.

Their reply;

Regarding to your questions,please refer to below reply:

1. According to AMD official specification,PBO (Precision Boost Override) function support for X470 Chipset,

our X370 Series do not provide this function.

2. We tested another MSI GTX980 Gaming 4G MS-V317 VGA Card + BIOS Video set to [UEFI] + Dell S2340L + HDMI Cable ==> Can get display,there is no problem.

3.

(1)We test disable Gear Diwn with XMP ==>　Can get display.

(2)

Geardown Mode
This mode allows the DRAM to adjust its internal clock to ½ mode or ¼ mode operation. Not supported for data rates below 2666MT/s.

Gear Down mode is only supported during initialization and self refresh exit.

When DDR4 frequency lower than 2666MHz，Cmd2T can set 1T/2T
When DDR4 frequency above 2666, Cmd2T no function, enter into OS Check as 1T
When Gear Down set to Disable，Cmd2T can set 1T/2T

4. Memory OC: try to adjust SOC voltage to 1V or higher.

5. X370GT7 Specification:

https://www.biostar.com.tw/app/en/mb/int...on.php?S_ID=874

Support 4-DIMM DDR4-3200(OC)/ 2933(OC)/ 2667/ 2400 up to 64G maximum capacity

6. For Others,thanks for your good suggestions,we will pass to our PM and R&D for future design considerations.

My response;

1. I will pass this information on to the people that wanted me to mention it and add it to the modified GT7 firmware being made and get them to more thoroughly test it.

2. I will verify this with another GPU.

3. (1) At what memory frequency did you test disabling Gear Down? As I have already stated the board fails to POST with Gear Down disabled with memory frequencies higher than 2400MHz. Put simply, set any frequency of 2666MHz or higher and the board will not POST when Gear Down is disabled. I do not know how I can say that any clearer for you. I can take a short video of this behaviour if you require.

3. (2) I am fully aware of how Gear Down works, your explanation enforces my above explanation.

4. I already know how much SoC voltage my CPU requires for high frequency DRAM. On a Asus X370-F Strixx and Asrock X470 Fatal1ty K4 the SoC voltage required is 1.05v to attain a stable 3600MHz DRAM frequency @ 2T. If necessary I can provide a screenshot from one of these boards showing this.

5. I am fully aware of the specification of the GT7, I am one of the people developing a modified firmware for it after all ([OFFER] Biostar X370GT7 modified UEFI). The board would be a capable memory overclocker if the issue of it failing to POST with Gear Down disabled and using a DRAM frequency of 2666MHz or higher gets fixed.

6. So by the wording you use are you saying in the forthcoming weeks you have plans to address the current outstanding issues with the GT7 and GT8? I want to be absolutely clear about this before I make any editorial updates to the places I published the GT7 review and make announcements to people who follow me in the large social media groups I am a member of.

Thanks

Now let's wait and see if there is a reply...

1)We test disable Gear Diwn with XMP ==>　Can get display What?!? WAT? I know it's simple copy/paste mistake, but still it's a funny read Made my morning!


What was #5? That's good, you got a #6 reasonable response there at the end, hopefully they do that!

#3 + #3 (2)

Good response at #6 too, I'm sure they will catch that and respond accordingly

@Lost_N_BIOS #5 would seem to just be Biostar trying to make excuses for Gear Down not working properly. Now it has been clarified to them again (this will make it 4-5 times now...) that the board does not POST with Gear Down disabled and a memory frequency of 2666MHz or higher it'll be interesting to see if I get a reply or if all of this magically gets "lost" again.

lol yep when I read "(1)We test disable Gear Diwn with XMP ==>　Can get display." I very much looked at that and thought WTF but also thought "ok lets keep this professional and show I am the better here". I've chastised them quite a bit for their procedures and very slow responses. Even when they were taking a week or more to reply their replies were no better than what I put in my last post. I think it is clear for anyone to see reading that they are clearly not testing following my very precise steps on how to reproduce each issue and / or just pulling half the answers they are giving out of their arse.

@barisuraz have you tested the Precision Boost option I added to the last firmware update? Does it work as it is supposed to?

Hopefully they will read your 4-5th explanation slowly and absorb this time around. I was tired and half asleep, that woke me up laughing so I had to share my thoughts, nothing meant offensively @Biostar all in good fun

@ket the Precision Boost Overdrive is working as intended. Thank you.

@ket ^^ Did you figure out something?