Issues with the cooling performance of the Swiftech H320 X2.
Issues with the cooling performance of the Swiftech H320 X2.
Hi,
Please let me know if you missed any details. This is my first question in this format.
I purchased the H320 but wasn’t getting the performance I expected. Could someone check if these numbers are typical?
I also have an FX-8320 with 1.4V OC to 4.375, which is new to adjusting voltages for OC—let me know if I should be able to raise the voltage.
When I run Prime95, it goes over 65°C before shutting down. I’d like better results and want to know if there’s a fix.
I tried unplugging the PWM for the pump, hoping it would run at full power instead of scaling, but performance and sound stayed the same. This makes me think the pump might be faulty.
I also considered the installation could be wrong, so I reinstalled it and added a bit more paste just in case I was too light before, but nothing changed.
Thanks in advance!
There’s no chance the H100i will surpass that cooler—there must be another factor at play. Are your idle temperatures within normal ranges or are they above what you expect? How rapidly do your load temperatures decrease when the load returns to idle? I’d be cautious about assuming the pump is the problem, since DDCs are generally reliable, but it’s possible something else could be affecting performance.
Are the fans spinning at their maximum? The cooler is securely attached. I don't specialize in AMD, but I suggest resetting your overclock to standard levels and restarting Prime. This helps confirm you're working within normal factory parameters and ensures the cooler functions properly, indicating your overclocks aren't causing overheating.
The fans are operating at maximum capacity. It is positioned at the top in Pull position.
This model isn't the premium version, so I questioned whether the fans might not be cooling efficiently given the maximum RPM of 1800.
Running stock showed temperatures reaching up to 56°C using Prime95, which suggests OC made a reasonable adjustment; however, my previous Corsair H100i performed better despite sounding like a jet engine.
I prefer the H320 for future expansion plans, so this seems like a more suitable option overall. I'm still frustrated with the cooling performance so far.
Is there a method to determine if the pump is the problem without replacing it entirely?
There’s no chance the H100i will outperform that cooler—there must be another factor at play. Are your idle temperatures within normal ranges or are they above what you expect? How rapidly do your load temperatures decrease when the load returns to idle? I’d be cautious about assuming the pump is the problem, since DDCs are generally reliable, but it’s possible something else could be affecting performance.
That's what I expected too. I had higher expectations for this.
The unit is operating at the OC, and it was left undisturbed for a while with a temperature of 41 C. Not terrible, but I was hoping for something in the 30s like with the H100i.
The temperature seems to drop faster (or at least according to memory) compared to the H100i, but I still need better cooling.
At the very least, this has helped me realize I wasn't mistaken when I thought the temperatures were unusual.
I might reach out to the company to request a replacement pump or similar.
Unfortunately, I purchased this from a store in another town—about four hours away.
Honestly, the voltage was actually higher for the same clock when I installed the H100i. I reduced it in an attempt to improve temperatures on the swiftech, and surprisingly it remained stable at those voltages, so I maintained them that way. Still, it was a good idea, though not the main cause. Upgrading from a 280mm radiator (H100i) to a 360mm (swiftech) should have provided more space for fans if needed, but I wasn’t expecting fans to be the problem. After a day of operation with no leaks detected, that seems unlikely. The only remaining possibility is the pump itself. I even switched to Arctic Silver instead of the standard model to rule out that as the issue.