With the arrival of my 5600X (and having too many CPU Coolers on hand and before having to RMA another 5600X), I had some time on my hands, so I decided to test out the 5600X with them to get some temperature numbers for reference, before shoving them into a small case.
If anyone is interested, the link to the results are here. Warning though, no fancy charts or graphs to be found, pure text/numbers dump. Link to the photo album is here.
Tl;dr? 5600X at stock values, most coolers will be able to handle it. Not much surprising results, bigger coolers have better dissipation, thus better scores and temps. Personal favourite still is the Thermalright AXP90s.
Disclaimer: I am no professional tester, and the numbers are purely for my own reference and use case, as your mileage will vary with factors like silicon lottery.
CPU Coolers tested:
- Noctua NH-D15S Chromax
- Noctua NH-L9a
- ID-Cooling SE-904-XT Slim
- ID-Cooling IS-60 EVO with 12025mm fan
- ID-Cooling IS-60 EVO with 12015mm fan
- ID-Cooling IS-47K
- Thermalright AXP90-X47 Full Back (Black/Copper)
- Thermalright AXP90-X53 Full Black (Black/Copper)
Yes, I didn't test the Noctua NH-L12S and Blackridge, as it seems like it wouldn't fit with my Corsair Vengeance Pro RAMs, but mainly because I didn't want to spend even more money. As far as possible, I used the same NF-A9x14 on the coolers that used a 92mm fan.
Test Methods:
- Stock 5600X Values: PPT: 76W, TDC: 60A, EDC: 90A
- Motherboard limits: PPT: 110W, TDC: 70A, EDC: 113A (Approximately, fluctuates according to HWinfo64)
- Manual Values: PPT: 88W, TDC: 60A, EDC: 90A
As far as possible, Curve Optimizer was utilized for the Motherboard Limits and the Manual Values at -10 for best two cores, -15 for the rest, +100 MHz boost. As far as possible, stability hasn't been tested intensely, but no crashes from low loads yet.
If you ask me, stock 5600X values are weirdly low, and all the tested coolers are able to keep it respectfully cooled, even the NH-L9a.
With motherboard limits enabled in BIOS with PBO, the PPT/TDC/EDC increases to around 110W (PPT), 70A (TDC), 113A (EDC). And here is where the D15S outshines the smaller SE-904-XT, in all aspects of temperatures, noise and scores. I only dared to test motherboard limits with these two coolers, even though I think the IS-60 EVO and Thermalright AXP-90s would have completed the runs, but at max heat.
Dialing in manual values closer to the previous generation 3700X at 88W, with Curve Optimizer, the 5600X extends its legs and prowess, in both scores and in temperatures as well.
Undoubtedly, the NH-D15S dominates in all aspects (as it should), but here we are to see the results for the lower profile coolers no? Cinebench R23 scores are all roughly equal, most of the time, where max temperatures are the differentiating factor.
The IS-47K and NH-L9a were unable to complete runs of Cinebench R23, unfortunately indicating that 88W might be possibly too much for them to handle. Initially I thought mounting pressure might have been the issue, I repasted and mounted them again, but still failed to complete.
Noise Disclaimer: I don't test noise levels unfortunately. My standing fan next to me is the noise ceiling and is able to drown out most fan noise from the rig. The only fan I can really hear over the standing fan are the three industrialPPC NF-F12s in my main rig when they are going full blast at 2000 RPM.
Although I must say, the Thermalright TL-C9B and TL-9015B are max rpm are more noticeable compared to the Noctua NF-A9x14.
What's next? Moving the parts into an actual case. It's a toss up between the new SGPC K39, or an unnamed K39v2 clone.
