OC3D 2020 CPU Cooler Roundup
Published: 8th April 2020 | Source: OC3D Internal Testing | Price: |
When we last revamped our heatsink test system, quad cores were the norm. Higher core counts were expensive and remained exclusive to the HEDT community, leaving Intel's i7-6700K as our best candidate for testing consumer-grade CPU heatsinks. A lot has changed since then.
At the time of writing, Intel's mainstream CPU sockets can support up to 8 cores and 16 threads, and on the AMD side core counts as high as 16 are supported. The quad-core era of 2016 is now over, which means that we need to rethink our CPU testing methodology.
While moving to a newer testbed may seem easy, it must be remembered that legacy results matter. Yes, we will build up a new bank of results over time with our new testing equipment, but legacy results matter. People want to know how new coolers compare to heatsinks like Noctua's NH-D15 or Closed-Loop Liquid Coolers (CLCs) like Corsair's H100i, which means that we have a lot of retesting to do.
- New CPU Coolers - Corsair A500 Air Cooler and Corsair H100i/H115i/H150i RGB Pro XT
- New CPU Coolers - Cougar Helor AIO Liquid Cooler Series
- New CPU Coolers - Fractal Celcius+ Prisma Series of AIO Liquid Coolers
- New CPU Coolers - Gigabyte Aorus Series of Liquid Coolers
- Retested Coolers - Air Heatsinks
- Retested Coolers - AIO Liquid Coolers
- Thermal Testing - Stock i9-9900K
- Thermal Testing Overclockers i9-9900K
New Test System
While AMD's AM4 platform sounds ideal for heatsink testing, with its higher core counts and newer X570 motherboard platform, it is worth noting that it had a downside, variability. With features like XFR and highly aggressive boost clocks, the thermal output of AMD's latest Ryzen series processors is highly variable, making exact thermal testing difficult. This factor is why we opted to use Intel's i9-9900K, as it offers increased clock speed stability and less run-to-run thermal variance.
In our testing, we are using an Intel i9-9900K at fixed voltages on ASUS's ROG Maximus XI Hero WiFi motherboard, with its BIOS set to offer 100% current capacity with load line calibration set to level 8. Fan speeds on our chosen case, Corsair's Obsidian 500D, will be fixed to 1250 RPM (54% speed) to help ensure that the only differentiating factor in our thermal tests is the heatsink we are testing.
Our testing environment will be within a thermally controlled room with an air conditioning system which can heat or cool the environment when necessary. Ambient temperatures for our testing area will typically sit at 18 degrees celsius, with most results sitting within 1 degree of this temperature target. Even so, our thermal results are designed to compare delta temperatures, making ambient temperatures less relevant.
100% perfect thermal testing is impossible when using consumer-grade hardware, but we have done what we can to minimise any potential variable.
Test System Hardware Specs
Below are the full hardware specifications of our new heatsink test system. This system will be kept offline to prevent Windows updates or other software changes from interfering with our thermal testing.
Our results are taken using average CPU thermals across all cores under a sustained CPU load. We use Prime95 version 26.6 with Small FFTs enabled to maximise CPU loads and heat production. This stress test is much heavier than most consumer-grade CPU loads.
OC3D Heatsink Test System
CPU: Intel i9-9900K (Stock at 1.5V & 4.8GHz at 1.2V)
Motherboard: ASUS ROG Maximus XI Hero WiFi Z390
GPU: Nvidia GTX 980 (Reference Model)
Case: Corsair Obsidian Series 500D RGB SE Premium
Fans: 3x Black Case Fans at 1250 RPM
Memory: Corsair Vengeance LP (Black) - 2x8GB at 3200MHz
Storage: Samsung 960 EVO M.2 NVMe SSD
PSU: Corsair RX750X