RedMagic 10S Pro benchmarks and throttling tests
The RedMagic 10S Pro is a mid-season upgrade of the 10 Pro. That phone arrived late last year as one of the first Snapdragon 8 Elite devices and it was the first RedMagic to use liquid metal to improve cooling. The company has had time to refine its setup since then.
The “Leading Version” of the Snapdragon 8 Elite is now available, which overclocks the CPU and GPU. The two Orion Prime cores now run at 4.47GHz, up from 4.32GHz on the original model, the six performance cores stay put. Also, the Adreno 830 GPU now runs at 1.2GHz, up from 1.1GHz.
RedMagic 10S Pro in Nightfall, Moonlight and Dusk
There are other changes – the RAM (up to 24GB) is now LPDDR5T, which should offer more bandwidth than the LPDDR5X used in the 10 Pro. Storage is up to 1TB of UFS 4.1 Pro.
Performance
We tested the RedMagic 10S Pro with and without its cooling fan to see how much of a difference it made. Let’s look at performance first – note that the fan doesn’t really do much for peak performance, it’s mostly about sustained performance, which is coming up in the next chapter.
With the Prime CPU cores clocked higher, the 10S Pro edges out ahead of the 10 Pro in the Geekbench single-threaded test and is ahead of a pack of Snapdragon 8 Elite and Dimensity 9400 powered phones. However, once the two Prime cores have to share the power budget with the six Performance cores, things even out with the original model. There are still applications that do heavy single-threaded tasks, they should see a small bump. Optimized apps and games won’t see a big benefit.
The GPU overclock has a bigger impact for gaming and helped the 10S Pro open up a lead on competing devices in the Wild Life Extreme benchmark. That will translate in higher fps in games that weren’t hitting the refresh rate cap and more stable gameplay in games that were.
Strangely, the RedMagic doesn’t do so well on the Solar Bay test, 3DMark’s ray tracing benchmark. The 10S Pro posts a higher score than the 10 Pro, but neither can match other Snapdragon 8 Elite and Dimensity 9400 phones.
AnTuTu is a whole-system test, so it tests additional components like RAM and storage bandwidth. The 10S Pro shows a small but noticeable improvement over the 10 Pro.
Stability
With the original 10 Pro, RedMagic added liquid metal – this is an alloy that melts at very low temperature and has very high thermal conductivity. It’s used in gaming PCs and laptops as it performs better than typical thermal compound.
In the RedMagic 10S Pro there is 36mm² of liquid metal in direct contact with the chip (30% more than there was on the 10 Pro). This guides heat to a 12,000mm² vapor chamber that spreads the heat around. But the phone doesn’t rely on just passive cooling – it could, but for gaming you can enable the 23,000rpm fan.
The cooling setup of the RedMagic 10S Pro
This fan sucks in cool air from the outside, pulls it across the phone over the chipset and dumps the now warm air out the other side. In our RedMagic 10 Pro review we noted that this system wasn’t as effective as it has been on previous generations. Let’s see if the refined cooling system on the 10S Pro does any better.
Unfortunately, the CPU still throttles pretty heavily – down to 56% from peak with the fan on, compared to 54% with the fan off (on the 10 Pro it was 57% and 54%, respectively). However, the fan does smooth out some of the bumpiness in the chart.
CPU stability test: fan off • fan on
Moving over to the GPU and the results aren’t any better. The stability reported by 3DMark in the Wild Life Extreme test is within the margin of error, as is the performance – both the best and lowest score from the test.
GPU stability test: fan off • fan on
Does the fan actually do anything? Indeed it does – here’s a look through a thermal camera showing the phone under load. You can see the warm air coming out of the phone heating up the desk below (the camera can’t see the air itself, obviously).
RedMagic 10S Pro in thermal vision
The RedMagic 10S Pro is a very fast phone – performance is not an issue. However, the active cooling system does not bring enough benefit to justify the extra cost and complications (e.g. the basic water resistance) associated with it.