• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Astral 5090

I've just run a quick test with the new release of HWMonitor that now shows hostpot temps. This is on a 5090FE.

Only done a 10 minute loop of Heaven bench (Which is old but still a good basic medium-high GPU workout) and on my 5090FE it seems these hotspot temps are absolutely nothing to worry about in this case. Unless of course the hotspot readings are all lies? :rolleyes:

Hotspot #1 and #2 track almost exactly the same, or often slightly under my GPU Memory junction temp (which is always about 10-15 degrees higher than the core temp).

Core: 66 degrees
Memory: 80 degrees
Hotspot #1: 78 degrees
Hotspot #2: 78 degrees

This is at my lowest undervolt profile (so card was only pulling ~320W) but I assume temps will just track upwards and in alignment with more power and volts, as you would expect.
 
Last edited:
Tetras said:
That's not related to your issue.

I don't know how nvidia reports the hotspot, but normally they would have several sensors available across the die and the hotspot is the uhh, hottest spot. That's not something you can feel with your hands, or in your room, since it is highly localised.

With nvidia 50 series cards nvidia removed access to this, you can't see it with GPU-Z or hwinfo anymore, but like I said, that doesn't mean anything about how hot your card is generally.
Click to expand...
i dont have an issue, what you talking about.
 
simon blakemore said:
well why read it,,what a dumb statement lol
Click to expand...
It's a discussion forum, came to see what the issue was, saw you had many good replies which you refused to listen to whilst demonstrating a total lack of understanding of GSCE level physics, gave up.
 
Last edited:
Enough you missed the point. Try reading up. The discussion is about the falsifying of the temps in the software. This has coincidentally been raised on video cards website where the hot spot temps have been enabled. These were turned off by nvidia for good reason. The 5070ti was 38 degrees hotter in relation to the hotspot. My point is my 5090 is no way running at 70 degrees given the heat it pumps out, thats it, nothing more. I dont believe the software so stop trying to be clever because its boring
 
The GPU temperature reading is an average across sensors isn't it? (excluding the hotspot reading issue)

I think the point some people are trying to make is that the temperature reading of 70 could still be correct, not because the reading is wrong or it's false, but because of physics in the sense that for example -

Card A (lets say a 4090) is running at 450w. the heatsink and fans are built to cater for that 450w load, and the temp sensor for the core reads 70 degrees. 450W of heat is still being transferred into your room, that will take longer to heat the room and the maximum temp your room reaches will be lower overall.

Card B (lets say a 5090) is running at 600w. the heatsink and fans are built to cater for the 600w load, the core of the GPU can still be 70 degrees, the same as Card A, but 600w of heat is being transferred into your room, your room will heat up faster, and the overall room temperature will be higher because it has an extra 150w of heat being pumped into the room.

Hopefully that made sense.
 
Last edited:
simon blakemore said:
Enough you missed the point. Try reading up. The discussion is about the falsifying of the temps in the software. This has coincidentally been raised on video cards website where the hot spot temps have been enabled. These were turned off by nvidia for good reason. The 5070ti was 38 degrees hotter in relation to the hotspot. My point is my 5090 is no way running at 70 degrees given the heat it pumps out, thats it, nothing more. I dont believe the software so stop trying to be clever because its boring
Click to expand...

Show your results, from hwmonitor with the hotspot temps, let's see the fake numbers.
 
simon blakemore said:
i dont have an issue, what you talking about.
Click to expand...
You keep us telling us the software readings are wrong, but like I said to you, the hotspot articles are not relevant.

Your card can have a much higher (e.g. 20+ degrees) hotspot than the reported GPU temp and the reading still be broadly accurate.

Your software readings can be wrong or right, but the card's hotspot being hot proves nothing.

If you want to prove they're wrong, then you need to measure it yourself.
 
I think it's pretty obvious that 'Mr Blakemore' is on the wind up now, but I hope he carries on. Not enough laughs around these days.
 
what? lol...whats the point of having them so you if wnat to observe that there maybe some kind of idea when thermal throttling taking place on a CPU or when you temps look dangersouly high on a GPU and it shuts down, its ok dont worry about the temps as arent accurate? What garbage. Well why bother having them then. Can we just leave it here and end this ridiculous conversation. The cards cooking and thats that....
 
If you're convinced the software is wrong then try an earlier version or another program like Afterburner or hardwaremonitor to confirm it and if its wrong contact Asus.

If it's an issue with the the sensor(s) I would think you would have a whole host of system problems as soon as you hit a certain threshold under extreme load if the sensors where under reading ex 70-80 degrees and continuing to use the card may cause damage so RMA it.
 
You must log in or register to reply here.
Back
Top Bottom