Cooling Performance

For cooling testing, I did change things up from our previous test suite. Our old Ryzen test setup tested across multiple CPUs but the new one focuses on the latest Intel CPU the i9-12900K which is known for generating a lot of heat. I did still do testing with a stock fan profile as well as at 100% fan speeds, but cutting down on the CPU count helped simplify the testing overall. I should also point out that most of the tests are done with the CPU power limits set to the suggested 250 watts with the exception of the unlimited power test. This is the same wattage our CPU review was tested at as well but remember that most of the higher-end motherboards default to a much higher wattage that will auto overclock your CPU even more.

My first test was using AIDA64’s FPU workload which is an extremely demanding workload that normally pushes thermal and power limits. I ran the workload for a half hour until thermals leveled off and recorded the temperature using the CPU dies sensor. The Nitro+ S240-A leveled out at 91c with the stock fan profile and at 89c when I did the same test with the fans turned all the way up. This was better than I expected compared to the larger MAG C280 with its larger fans with the 100% fan speed result being a degree less and was 2c less than the H100i with the stock profile. The stock profile for the H100i was controlled by their own software whereas the others were all run by the Asus standard fan profile not helping the h100i’s performance there.

With that same AIDA64 FPU workload I did also go in and turn the PL1 and PL2 wattages up to allow the CPU to go as high as it could being limited just by thermals. For this, I just documented the wattages on each cooler. Here the smaller size of the Nitro+ S240-A was more apparent, it was able to only edge out another 8 watts over the stock limit.

Next up I switched AIDA64’s workload over to the CPU workload which I consider to be more realistic to gaming and standard PC use where the FPU workload is more demanding like rendering can be. You can see just how big of a difference the workload makes just by the range of temperatures that the coolers are running in here. The Nitro+ S240-A came in at 65c with the stock fan profile which was better than what Corsairs software was letting the H100i do but two degrees warmer than the larger C280. The Nitro+ S240-A did better with the fans cranked up however coming in at 57c which was 5c lower than the others. I should point out that the fans on the Nitro+ S240-A are running at 2400 RPM when at full speed where the C280 fans run at 1800 and the Corsair MagLev fans top out at 2000 RPM. That extra airflow does help give the Nitro+ S240-A a little more headroom. Especially here when things aren’t maxed out like the FPU workload did.

For the last test, I changed the workload over to Blender Benchmark with the BMW render. This render doesn’t run for as long as I am able to run the AIDA64 workloads which I think helped benefit the Nitro+ S240-A here. As soon as the workload hits the fans spin up and the cooler absorbed the thermals from the benchmark a little better than the other two coolers with the stock profile coming in at 86c and the same on the 100% fan speed test. I even went back and retested the other coolers just to make sure the difference wasn’t caused by any other issue. I would bet if you were running a longer render that it would come in closer to the FPU results on the stock fan profile but that extra fan speed on the Nitro+ S240-A helps a lot on the 100% tests.