I wanted to test the thermal performance to see what amount of throttling would be experienced, so let's take a closer look and find out if we can improve things with some under bolting. This laptop has Intel's new eighth generation coffee, like I 78758 CPU, which has six cores and twelve threads it's clocked at 2.2 gigahertz. But the single core turbo boost speed is 4.1 gigahertz, while the old core boost speed is 3.9 gigahertz, so we'll get quite hot with all that power, especially in a laptop this thin, given I've seen it throttling and even larger laptops that I've tested I've also got Nvidia 1070 max Q, graphics in this model, so expect different results. With the 1060 max Q version, I've tested all games with gaming mode enabled basically raises sign up. Software lets you toggle between balanced mode and gaming mode, where gaming mode will increase your CPU power limit under vault, the graphics and overclock the graphics for a bit better performance, which also results in more heat. Thermal testing was completed with an ambient room temperature of 18 degrees Celsius, it's cold here at the moment, as it's winter in Australia, so expect warmer temperatures in a warmer environment at idle. The CPU and GPU were on the warmer side, around 50 degrees Celsius, while playing pub G at high settings. We can see that the temperatures worse, as shown in green by under bolting the CPU by minus 0.1, for our faults, as shown in yellow, we can reduce the temperature of the CPU a little.

The graphics will maxed out at 90 degrees, regardless of the test. This was a combination of PAL limit and thermal throttling. The full load stress tests would run using a 264 and heaven benchmark at the same time and I've tested both the balanced mode and gaming mode. Here in balanced mode, the CPU gets a lower power limit, so performs less as we'll see in the next graph, but temperatures improve a little with the under multiplied in red. These are the average clock speeds while running the same tests for the temperatures just shown it wasn't possible to reach the full 3.9 gigahertz, all core turbo speed in any of these tests, which, in the case of distress, tests was due to power limit or thermal throttling. Granted in pub GE, I don't think the CPU needs to max out as it's, not a bottleneck, which probably explains the lower clock speeds while under bolted in yellow it just doesn't really matter. Maybe I should be using a different game for this test, but these are the speeds that I saw with the gaming mode enabled the TDP increased. So it seems like this affects the power limit provided to the CPU I wasn't able to otherwise tweak it in Intel X to you. This explains why the clock speeds rise with gaming mode, enabled giving us more power, but also a hotter system, as we saw in the last graph under bolting in red and dark blue, did improve things too, but still not quite enough to fully remove throttling in this Particular stress test, but this will of course differ based on the specific workload great.

So how does this performance boost actually translate to games and applications in the games tested? The exact same windows updates, game updates and Nvidia drivers were installed, so there shouldn't be any changes. Other than the CPU under bolting in csgo for the most part were actually seeing slightly lower 1 loads with the under fault applied, but slightly higher average frame rates, Ghost Recon was tested with the built in benchmark and the results are pretty close together. For the most part, but the difference becomes slightly larger at the highest setting levels. Far cry 5 was also tested with the built in benchmark and the results are a mixed bag. Lower 1 lows at all settings, while the average is very a little ashes of the singularity, was tested as a more CPU demanding game and we're, seeing a larger difference at the lowest settings on average. But overall there is a small improvement as showing the improvement in CPU, thermals and power does seem to translate to slightly better gaming performance. Nothing major though it of course greatly depends on the game. Some games, like Far Cry 5, for instance, didn't, seem to make a difference, whereas more CPU, intensive ones like ashes of the singularity received a clear boost. Typically, most games depend on the graphics more than the CPU, though so realistically, in many cases I don't think you'll see big gains from CPU under vaulting here. For that reason, in general we're not seeing much of a difference in the games tested, whereas CPU intensive applications should see a lunch again.

These are the clock speeds I got while just running CPU only stress tests without any GPU load and even with no GPU load was still sink. Parliament throttling preventing the full 3.9 gigahertz speed, but gaming mode and under bolting did help raise. This I've got some Cinebench CPU benchmarks here, and we can see that there's a fair difference in performance just by going from balance to gaming mode, and we can improve this further 0.14 Revolt. Wonderful, despite this sub was still sing. Power limit throttling, as discussed earlier in the CPU and the stress test. The 87 58 should be able to pass 1200 in ideal circumstances, but it's not too far behind here so we're almost able to get full CPU performance in this CPU only workload. As for the external temperatures, where you'll actually be putting your hand, sat idle, the body of the laptop is sitting in the mid 30s in the center and getting warmer went to the back while gaming. This increases to the 50s in the center and was a little warm to the touch, although the left and right sights were quite cool in comparison with very similar results in the full load stress test. As for the fan noise produced by the laptop bowler, you ever listened to some of these tests at idle. That was still audible, which was expected based on the warmer idle temperatures and then, while gaming and with the fans maxed out it wasn't too different from many other laptops, I've tested still a bit loud they're balanced murder was a little quieter than gaming murdered, as we Saw before it gets hotter in gaming mode, I didn't test temperatures for both default fans and manually maxing out the fans as in gaming mode, while playing games or under stress tests.

I found the fans to be maxed out the same regardless, which would give the same results as for overclocking, the 8750 eight CPU can't be overclocked, but I was able to increase the GPU core clock if the max q10 70 little on top of what gaming mode. Sir, these are the average speeds while running the heaven benchmark in the default balanced mode with gaming bird, which essentially overclocks the graphics by a hundred megahertz on the core and three hundred megahertz on the memory. And then with my manual overclock. We get just a little extra, but no much gaming modes already doing a pretty good job, I've retested pub G with both balanced and gaming modes and then with the gaming mode under vaulted and we're, just sings small differences in performance between them. I'Ll also note that I haven't tried manual GPU under bolting here as gaming mode in the Razer synapse. Software has already done this quite well. The differences in performance shown aren't hard and fast rules. There are different factors which will vary results, primarily the temperature you're running in application of thermal paste, and even the specific hardware which comes down to the silicon lottery. You may not be able to under bolt your 8750 h the same as me. It depends on the chip and the specific power requirements, so don't just blindly copy my settings and do some testing to find out where your stable point is for best results.

I'M, not sure how practical it would be to replace the thermal paste in this model, as it uses a vapor chamber cooler which takes up quite a bit of space. Presumably this could be removed and paste is used between it and the CPU and GPU. But this isn't really something I can test in a review unit for a number of reasons. Generally, I can't really tamper with the review unit. The laptop needs to work in the same way for the next person to receive it. If I go ahead and remove the stock double paste and replace my own, I can't put the old paste back, so the next reviewer would experience something different from what you'd actually see with the product and unknowingly report incorrect information due to what I've done under bolting On the other hand, isn't physically intrusive and as we've seen, it did improve performance and temperatures in this particular unit. With no downside once you've got a stable under bolt it's a great way to get back some performance if you're running too hot or facing power limit. Throttling, let me know how much of a performance boost you've found by under vaulting your hardware and what you thought of the improvements here and don't forget to subscribe for the full review of the razorblade 2018 gaming laptop as well as future tech.