Link to pretty comparison grapghs
Summary Table
| Title | Before @ 22-24c ambient | Graph | After @ 23c ambient | Graph |
|---|---|---|---|---|
| Idle | 71°C CPU |  | 53°C CPU |  |
| Baldurs Gate 3 1080p | 80,4°C CPU 67,2°C GPU |    | 67°C CPU 64°C GPU |   |
| Baldus Gate 3 4k | 80,2°C CPU 66,7°C GPU |    | 71°C CPU 65°C GPU |   |
| Google Chrome 1080p | 73°C CPU |  | 51°C CPU 46°C GPU |   |
| Google Chrome 4k | 60°C CPU |  | 51°C CPU 46°C GPU |   |
| Cinebench Single Core | 92,2°C CPU 57,3°C GPU |  | 66°C CPU 50°C GPU |   |
| Cinebench Multi Core | 91,6°C CPU 63,1°C GPU |  | 73°C CPU 60°C GPU |   |
| 3D Mark Timespy 1080p | 75,6°C CPU 61,6°C GPU |    | 70°C CPU 64°C GPU |   |
| 3D Mark Timespy 4k | 79,5°C CPU 64,8°C GPU |    | 70°C CPU 61°C GPU |   |
Graph order
- CPU temp °C
- CPU clock MHz
- CPU power W
- GPU temp °C
- GPU core clock MHz
- GPU memory clock MHz
- GPU power W
3D Mark Benchmark Results
| Before | After |
| 4 741 Time Spy @ 1920×1200 desktop resolution https://www.3dmark.com/spy/41437423 | 4 893 Time Spy @ 1920×1200 desktop resolution https://www.3dmark.com/spy/41456573 |
| 4 705 Time Spy @ 3456 × 2160 desktop resolution https://www.3dmark.com/spy/41437649 | 4 840 Time Spy @ 3456×2160 desktop resolution http://www.3dmark.com/spy/41458612 |
Parameters Comparison in Baldurs Gate 3
| Resolution | 1920×1200 | Max | Min | Avg | 3456×2160 | Max | Min | Avg |
|---|---|---|---|---|---|---|---|---|
| Before | 80,2°C CPU 66,7°C GPU | 714.9 FPS | 1.2 FPS | 56.3 FPS | 80,2°C CPU 66,7°C GPU | 345.1 FPS | 0.7 FPS | 20.9 FPS |
| After | 67°C CPU 64°C GPU | 687.5 FPS | 0.8 FPS | 53.9 FPS | 71°C CPU 65°C GPU | 414.9 FPS | 1.7 FPS | 17.4 FPS |
Baldur Re-testing
Results right after liquid metal were strange. Temps were down as well as avg performance. Max and Min were higher but overall FPS dropped. It may be because of different benchmark timing. Before, I launched all data recordings in menu while after – when game loaded into the Gauntlet of Shar so there was no in menu 120 or 60 fps for minute or so. That’s why I did it again and will compare data from videocapture.
Afterburner graphs
CPU
Framerate Avg
GPU
Lost temp sensor for some reason….
Afterburner Benchmark
08-09-2023, 11:12:25 bg3.exe benchmark completed, 17237 frames rendered in 358.235 s
Average framerate : 48.1 FPS
Minimum framerate : 0.2 FPS
Maximum framerate : 683.8 FPS
1% low framerate : 0.2 FPS
0.1% low framerate : 0.2 FPS
Conclusions
I even changed nVidia drivers from Studio to Game Ready to give it another chance but with no success. It behaved similarly – temps went down as well as avg. FPS but Max and Low both were higher. What disturbs me even more are 3D Mark scores. They were better after the mods…. I know it’s not a gaming laptop but this behaviour does not make any sense to me. It’s bette but also worse? To make things clear, below are screenshots from game capture at 1920×1200 before and after. I’m not touching 4k because change from 7 to 9 FPS is….not reliable, let’s say. ;d
| Before | After | Change |
|---|---|---|
 |  | -3°C +0,5W CPU framerate from avg(40,79)FPS to avg(39,94)FPS Going out of the room, looking far away, lighting fire and going to the crystal. All before loading to forest.  |
 |  | |
 |  | |
 |  | |
 |  |
I leave these results to your judgement.
Below you can find some more photos from testing phase.
Motherboard temps (hotspot measurement)
External thermocouples
One for the keyboard under the key with number 6.
One for the lid was placed exactly like that. Photo taken from video recording that I will share later.
Benchmark Recordings
Links to products are refferal links with my Amazon Associate Program.