[GPU-Z Test Build] Need large-scale beta testing

[W1zzard]

Build in first post updated with new version which should fix the boost clock misreadings on older NVIDIA cards
@Fluffmeister @Maban @tigger

 

[Hugis]

780ti looks good, was also able to save the bios as well.

well spotted wiz ,Edit: dont know whats happening to VDDC it drops to -156.500v and my 12v can drop to minus values ,this is non overclocked, havent had any instability issues my psu is quite old now...

 

[Maban]

Boost clock is properly shown now.

 

[dorsetknob]

win 7 x 64 no problem with this lp OEM HD 8490 1gig

@W1zzard are you going to leave this thread unlocked so people can edit their posts if they add over time

 

[W1zzard]

are you going to leave this thread unlocked so people can edit their posts if they add over time

no need, after next public release this thread will be useless

 

[W1zzard]

780ti looks good

VDDC sensor doesn't look right

 

[SoNic67]

I have submitted my result. Strangely the RX580 with the January drivers (19.2.2) didn't show OpenCL active. I am installing the February 25 drivers now.

Well, with those drivers (19.2.3) everything seem to be OK, except the title of the window that says the same 2.16.0 as nonbeta. Validated again. BIOS read worked.

 

[MihaStar]

Checked the GTX960 + Win8.1 x64 Pro on my workstation machine. Everything looks good (incl. BIOS saving), except the border capture bug I've already described earlier.
The sensor tab must be longer (in vertical dimension) than the main window - and then the bug takes place.

 

[MihaStar]

As I see, nobody uses the internal screenshot capture function They are wasting time with Copy-Paste in MS Paint instead

 

[Voluman]

Ver 1 (yesterday)
Win 7: gtx 1070 - ok, hd graphics 2500 - ok
Win 8.1: gtx 1050ti - ok, gt730 - double boost, gtx750 - double boost, gtx 760ti (670) - ok, dell hd7570 - ok, rx580 - ok, hd graph 610 - ok

Ver 2 (today)
Win 8.1: gtx 750 - ok

Later check gt 730 too with the second version.

Edit: with gt 730 is working normally now

 

[W1zzard]

says the same 2.16.0 as nonbeta

yup, version number hasn't been increased for the beta build to avoid confusion with final public 2.17

 

[xkm1948]

Everything works on EVGA XC 2080 Ti.

 

[SoNic67]

They are wasting time with Copy-Paste in MS Paint instead

Not on Windows 10. We have the snipping tool, useful (in my case) to take one picture of multiple instances of GPU-Z at once.

 

[Vayra86]

Looks as usual on my GTX 1080. Data is correct and 100% complete. Will report once I find quirky behaviour.

Fun fact and (very) minor bug: clicking the Lookup button directs me to this page:

While I have a regular MSI Gaming X 10Gbps, with memory clocked to 11Gbps manually.

 

[Ramo1203]

Everything works fine for me. (Aorus RX580 8GB and MSI GTX 1080TI Gaming X)

 

[ne6togadno]

looks like there is a problem with vega 64 readings

during the bench unigen reported clocks around 1500 mhz gpu 945mhz mem. max values on sensor tab are from gpuz render test i ran before superposition

 

[W1zzard]

looks like there is a problem with vega 64 readings
View attachment 117467View attachment 117468
during the bench unigen reported clocks around 1500 mhz gpu 945mhz mem. max values on sensor tab are from gpuz render test i ran before superposition

Please elaborate

 

[GamerGuy]

Do the missing sensor values appear when you put load on both cards? It seems the second card is sleeping in ULPS?

When I was playing Strange Brigade (has CF support), according to MSI AB, the secondary card was working in tandem with the primary card in that game. But looking at the 1st screenshot of my cards, it'd seem that the secondary card was idle/unseen all the time (there's no idle clock value/temps/fan speed (RPM)/etc).

 

[MrGenius]

looks like there is a problem with vega 64 readings...during the bench unigen reported clocks around 1500 mhz gpu 945mhz mem. max values on sensor tab are from gpuz render test i ran before superposition

There does appear to be something going on there with my Vega 64 too. Superposition says 1614MHz max core and 1165MHz max memory. GPU-Z 2.16.0 also says 1614MHz max core and 1165MHz max memory.

This beta build of GPU-Z says something MUCH different than 2.16.0. As far as the core max anyway. 1679MHz is WAY higher than 1614MHz(what Superposition and GPU-Z 2.16.0 says). I don't know who's right there. 1679MHz doesn't sound unreasonable(it might actually hit that for a second or 2). And 1614MHz sounds too low to me(I doubt it's really running that slow, I'll check with some other software when I have time...done, see EDIT 2). It's only off by 1MHz on the max memory clock though(1164MHz vs. 1165MHz)

EDIT: Just noticed the temp differences. Seems like the beta build is much more accurate with those. There's no way in hell the HBM ever reached 99°C. Don't know about the rest. But I doubt the VR SOC got up to 99°C either.

EDIT 2: Afterburner also says 1614MHz max core too. I still think that's wrong though. Gonna try HWiNFO next...done. HWiNFO says 1614MHz as well. Maybe that is right? Next question is why would Superposition run so slow on my card? Heaven and Valley run between 1653MHz to 1664MHz. This makes no sense...more testing...done. Appears to be right after all. I set the core clock to 1675MHz and was able to hit 1639MHz in Superposition(1080P HIGH). According to Superposition and GPU-Z 2.16.0. Which does make sense. It should be able to reach about that or better. Probably not much better though it wouldn't appear. Setting it higher than 1675MHz crashed.

So...bottom line...beta build wrong...everything else right...IMO. Totally weird how the beta build was saying his was so much lower than it should be, and mine was so much higher. What's with that?

 

[neatfeatguy]

Looks good on my all powerful work computer's GT 730! Confirmed boost clocks are reading correctly

 

[Ruru]

Gigabyte 780, everything seems to work perfectly, also bios dumping. GPU is on h2o so temps are correct there. Also it is on 2.0 x4 slot because Ryzen 2000 compatibility issues and I don't have an option to change PCIE gen in bios, so it's on the chipset's 2.0 x4 slot.

 

[W1zzard]

his beta build of GPU-Z says something MUCH different than 2.16.0.

I'm using a new method to get sensor readings. Can you compare with what Wattman shows?

 

[MrGenius]

I'm using a new method to get sensor readings. Can you compare with what Wattman shows?

I gave WattMan a try and it says what Superposition says. Which would be what 2.16.0 or HWiNFO would be saying too(I have to presume, since they matched Superposition in my previous tests).

I was kinda hoping the beta was right. But it doesn't appear that it is. That would be a pretty big shocker. The news would have to read "Vega's core clocks actually way higher than reported...up to +65MHz!!!". But nope...I don't think so. Maybe I'm wrong though. Anything's possible I suppose.

 

[SoNic67]

When I was playing Strange Brigade (has CF support), according to MSI AB, the secondary card was working in tandem with the primary card in that game. But looking at the 1st screenshot of my cards, it'd seem that the secondary card was idle/unseen all the time (there's no idle clock value/temps/fan speed (RPM)/etc).

Strange Brigade in DX12 mode does not use multiple GPU. It seems to want to use multiple GPU when launched in Vulkan mode.
But then, my case, with one AMG and one nVidia, it freezes when wanting to start the game.

 

[spectatorx]

I will not help you much by saying: radeon r9 380 4GB, windows 10 enterprise 1809 with latest updates, gpu-z is working fine in therms of readings, saving screenshots and exporting bios files.