The MSI GE66 gaming laptop performs well in games, but just how hot does it get and what performance improvements can we make?
To find out, I’ve tested thermals with the highest specced configuration, so we should be looking at more of a worst case here.
The 2070 models aren’t Max-Q though, so the GPU would use a little more wattage.
Air comes in through the vents underneath towards the back
and is exhausted out of the back corners
and on the left and right sides towards the back.
Inside there are a couple of fans, as well as two heatpipes that cross over to the other side.
They’re not strictly shared in the sense that the pipes don’t touch both CPU and GPU, and it looks like we’ve got some nice VRM coverage. Too nice perhaps, my unit appears to have an overflow of paste
it doesn’t look like a factory application so might just be my review unit, I’m not sure, but it could potentially affect the results a little.
The MSI Dragon Center software lets you select between different performance modes, which from lowest to highest are silent, balanced and extreme modes. You’ve got the option of overclocking the GPU in extreme performance mode, however no overclock was applied by default. You can also toggle coolerboost here, which sets the fan to max speed, however there is some manual customization that can be done to CPU or GPU fan.
There’s also no undervolting done out of the box, and by default it’s disabled, however if you boot into the BIOS and then press this epic cheat code you’ll be able to enable undervolting
as well as a ton of other options, so be careful and only change what you understand.
Although the RTX 2080 Super is Max-Q, it’s using Nvidia’s new Dynamic Boost, which means the power limit can boost higher depending on if there is power available. I saw a 105 watt average in a GPU only stress test, but it will be less with the CPU active, closer to its regular 90 watt limit which is what happens if dynamic boost were to be disabled.
Thermals were tested with a 21 degree Celsius ambient room temperature. Idle results down the bottom were ok.
Worst case stress tests were done with the Aida64 CPU stress test with CPU only checked and the Heaven benchmark at max settings at the same time, while gaming was tested with Watch Dogs 2 as I find it to use a good combination of processor and graphics. Temperatures were the warmest in silent mode when under either workload, which makes sense as this mode is designed to run the fans quieter, and the GPU was thermal throttling. In the stress test, the CPU was thermal throttling in balanced mode at 95 degrees Celsius, but the increased fan speed from extreme mode was enough to remove this, meanwhile in this particular game CPU thermal throttling was less of an issue long term, there were just some spikes.
When we enable coolerboost, so set the fans to max speed, there’s a decent improvement to thermals. Undervolting didn’t affect temperatures with the stress test running but helped the game a little, then a cooling pad helped a fair bit more.
These are the average clock speeds while running the same tests.
Basically performance increases as we step up through the modes, however we’re not hitting the full all core turboboost speed of the i9. The best we’re able to do is 4.0 to 4.1GHz over all 8 cores with undervolting, still an ok result though, and I’d expect better results with the 6 core option as there would be fewer cores to power.
This is a result of the power limits.
With the stress test running, the CPU would not pass 45 watts, despite PL1 being set higher. The GPU would often run at 90 watts, though in a GPU only workload Max-Q dynamic boost was able to push this up to 105 watts. As both of these workloads hit the processor too, we’re only really seeing the GPU boost up with the game running in silent mode. I’m assuming this is due to the game not being as heavy as the stress test, but the CPU still uses less power so the GPU is able to take it instead.
Interestingly with this game, the CPU power limit was also able to boost higher compared to the 45 watts noted in the stress tests, so it seems to depend on the workload, but again that could be part of dynamic boost.
Anyway this explains why the temperatures in extreme modes were quite good for the most part, we’re being hit with that processor power limit. 
In a CPU only stress test with the GPU now idle, silent mode still has the same 30 watt limit, however this increases with any of the other modes. 
Basically thermal throttling was being hit in this stress test any time we increase from silent mode, but as the fan speed also increases throughout the modes we still observe a steady improvement to the clock speeds.
I’ve used Cinebench to show you how these different modes perform
and surprise surprise, the multicore score closely follows the clock speeds that we just saw. 
When we look at how this stacks up against others, the single core result is one of the best tested, not far off the 9900K in the GT76. The multicore score is decent, the second best 10980HK I’ve tested so far, though it’s worth noting the cheaper RP-15 with Ryzen 7 4800H is doing far better there, especially considering how much cheaper it is.
When we look at how an actual game performs in these different modes, it’s not all that different
silent mode is still giving excellent performance as we’ve still got high GPU power levels here, as noted earlier. For a CPU heavier game performance differences may be more pronounced, but when I was testing watch dogs 2 it was still running smoothly. We can also get a nice 6% boost by overclocking the GPU, undervolting the CPU, and using a cooling pad which is more than what I typically see with these changes.
If you want to see more gaming benchmarks from the GE66, check out MSI GE66 - A Beast In Games where I’ve tested 21 games at all setting levels.
As for the external temperatures where you’ll actually be putting your hands, at idle in silent mode it was in the low 30s, pretty standard.
With the stress tests going the wrist rests are cool, it’s quite warm in the center and hot to the touch up the back, though you don’t need to touch there.
It’s a little cooler in balanced mode as the fans increase here
perhaps a little cooler still with extreme mode, the middle was a bit warm but not bad.
With coolerboost enabled it gets a fair bit cooler to the touch and the wrist rests are still cool.
At idle the fans were just audible, but from time to time they would boost up a bit louder, and this was to a similar level as just watching a YouTube video and browsing Chrome.
With the stress tests going in silent mode it’s still on the quieter side, and if you recall gaming performance was still quite good there, so gaming with a quieter machine is definitely possible at the expense of higher temperatures. Balanced mode was louder, but still lower compared to most gaming laptops.
Extreme mode was more in line with other gaming laptops at max, then cooler boost is quite loud, you’ll definitely want headphones or something, but I think this is a good thing. As there is some fan control and the different modes it gives you the choice of running it how you like, cooler and louder, or warmer and quieter.
Overall I think the performance from the GE66 is quite good. There are a number of different performance modes to choose from, you’ve got the option of running quieter while still getting great performance. You can boost the fans if you want to run much cooler, and unlike the Lenovo 7i you can manually boost them, though at the end of the day the clock speeds are only a little ahead of the 7i, but the GE66 runs a fair bit cooler.
Thermals are generally kept in check due to the lower CPU power limit when the GPU is active. This does unfortunately mean that when both the CPU and GPU are being hit hard, some performance is lost, but as we saw in actual games the performance is excellent and it’s good to have plenty of options.
What MSI is offering in their advanced BIOS such as the option to unlock undervolting is great, many other laptops are simply not doing this. In any case, the performance being lost clearly isn’t really making a difference in games. If you caught my game benchmark video, link in the description, the GE66 with this hardware is performing extremely well, basically only being beaten by thicker more expensive models with a desktop processor, so yeah quite impressive all things considered.
Let me know what you thought about the thermals from the MSI GE66 gaming laptop down in the comments.
No comments yet