I am doing some exploring and as part of that I am running Prime95 for stress testing. I just want to confirm that it is normal to see thermal throttling during stress testing.
I see 100C while on AC power going down to 90C while running on battery power. On battery the computer is using 75W during the stress testing.
The fan(s) are operating and I can feel warm-ish air exiting the vents at either side. It isn’t an especially strong air flow. Core clock maximum is 4100 MHz with an average of around 3900.
I am using HWinfo64 to see these figures.
Does this sound reasonable ?
BTW, I started the stress testing because the computer was always running cool and te fans never turned on. I wanted to make sure the fans actually worked.
1 Like
I stress tested using xmrig (randomx algo) - I have the Ryzen 7 and have the dGPU.
On AC, it did get up to 90c for me, getting about 4200 MHz. Using lm-sensors, it was showing 54 W for the CPU on balanced.
On DC, it dropped to 74c, got around 3800 MHz, and was showing 30 W draw on balanced.
Using Fedora 39.
I’m surprised your saw 75 W. Did you mean at the wall? Because the CPU should have a limit at 54 or 60 W?
The 75W was the draw from the battery as per HWinfo64. I am not sure how accurate that is though because sometimes it shows zero in or out even though the battery is charging or discharging. I also only see a maximum of 80W for battery charging with a 180W supply. My next experiment will be to stick a power monitor on the input of the AC adapter to see what it reports.
That checks out.
It was shared in another thread that the battery peaks at a charging speed of 1C. C-rate is the ratio of charging speed (in W) to battery capacity (in Wh). So with an 85 Wh battery that means the battery charges at a max of 85 W.
However that is the peak, which can only be achieved under specific conditions (including temperature, battery state of charge, and more). Practically I usually see around 60-80w on my FW16.
Hmmmmm … very interesting. So in theory we can achieve close to 180W for the combination of battery charging and operating the computer. A quick test showed I could bring the supplied power to around 125W but that was with an 87% charge in the battery so maybe this was too much battery charge to draw 1C. (80W to operate the computer and 45W for battery charging)
Correct, the C-rate of this battery is 1, which means the maximum power is 85W. 75W is reasonable as DC-DC conversion loss and internal resistance are present.
The safe charging temperature is much narrower compared to discharging, thus 85W charging can be achieved much less frequently.