Alphacool Eiswolf Baseline Benchmark – GTX 1080 Eiswolf Thermals

Starting with a baseline thermal analysis, this simple test runs FurMark for just 25 minutes to provide us with an understanding of performance prior to endurance testing.

For baseline GPU temperatures, we’re seeing the Alphacool GPX Pro idling at 5.4C delta T prior to any testing. This compares to about 2C delta T on the GN Hybrid build. If you’re wondering how that makes sense, here’s why: There are two things that increase the GPU diode temperature of the GPX Pro. The first is that the GPX Pro is a full plate, so we’re sinking the VRAM, VRM, and even back-side PCB temperatures into the same cooling solution. The second is that the Alphacool unit is connected to the Eisbaer cooler, which means we’ve got the CPU on the same loop. The CPU has to be first in this particular loop, so even though it is idle and not producing any heat, the liquid temperature is going to be a little bit higher.

As for load temperatures, we’re looking at the GPX Pro at about 15.5C and the GN Hybrid at about 15.3C. No real difference here, but we do see that the Alphacool unit is more efficient at keeping the GPU temperatures under control when it’s under load.

Let’s move on to endurance testing.

Alphacool Eiswolf Endurance Testing vs. Hybrid GTX 1080

This first chart shows the GPU, MOSFET #3, and PCB backside temperatures for the devices. The probes remained in the same place for each test.

We’re seeing load temperatures are about the same when tested over a 1-hour FurMark VRM burn-in. The FETs are significantly different, though, with the Alaphacool GPX Pro keeping its MOSFET #3 around 6C cooler than the GN Hybrid solution, which relies entirely on a baseplate and VRM blower fan for VRM cooling.

As for the PCB backside temperatures, our primary theory for the increase in GPX Pro PCB backside temperatures is that the massive backplate and thermal pads, which connect to the front-side, are helping to better spread heat across the entire card. This can cause a thermocouple to look like it’s experiencing a higher temperature, since it’s indirectly touching the backplate and is the central transfer location for the heat.

Alphacool Eiswolf Endurance Overclock Testing

Endurance tests with GPU overclocks manually applied position the GPX Pro and GN Hybrid about 0.3C apart in GPU temperature, with the GPX Pro again advantaged in MOSFET temperatures. The probe on the PCB is again reporting higher temperatures, but that’s because – again – we’ve got a thermocouple sandwiched between a thermal pad and backplate. All the heat is transferring through that area. Given that this thermocouple is positioned on a hotspot, the 42C PCB backside temperature is still way more than acceptable. It’s not like we’re hitting 130C on this PCB.

Alphacool Eiswolf – CPU & GPU Performance with Eisbaer

These next tests are in more of a vacuum – we’re only looking at the GPX Pro for these.

The first is a benchmark with the CPU being loaded with Prime95 LFFTs, maxing out load at 100%, while the GPU is being tortured with FurMark. This results in high temperatures in both the CPU and GPU, and since we’re sharing a loop, that’s an important thing to look at.

The results position the CPU at 52C dT under load, or around 70-73C when accounting for ambient. This is with the GPU at around 19C delta T, or roughly 4C higher than with the CPU under minimal load.

The CPU cooling isn’t great, to be honest, but it is cooling two hot devices at 100% load each – that’s no easy task. Regardless, we’ll talk about that in our future Eisbaer review.

Alphacool Eiswolf – Low RPM Performance

With RPMs dropped to 800 (from ~1100), we’re seeing a performance output that’s only about 1.4C warmer in idles for the GPU, and about ~2C warmer in GPU load temperature. Not a big deal, considering reduction of noise levels overall.

Alphacool Eiswolf – Noise Levels with Eisbaer

We’re seeing a noise output of approximately 42.2dBA with the Eisbaer at 11000RPM, or its maximum fan speed. This is total system noise, but we’ve got a passive PSU and no case fans, so it is actually about the noise you can expect from the Eisbaer standalone.

The noise levels drop to 33.5dBA when under 800RPM, and our temperature impact isn’t that substantial if not loading both the CPU and GPU simultaneously. Makes sense to run at 800RPM, if noise is a concern.

Continue to Page 3 for more information.