Building-up a semi-custom liquid cooling loop is a bit of a new trend, spawned from a surge in AIO dominance over the market. The ease of installation for AIOs greatly exceeds what’s possible with an open loop, with the obvious loss of some customization and uniqueness. The cooling loss, although present, isn’t necessarily a big factor for the types of buyers interested in AIO CLCs rather than open-loop alternatives. Ever since we saw PNY’s solution years ago, though, and then more recently EVGA’s quick disconnect solution, the market has begun to burgeon with semi-custom loop “CLCs.”
An example of these semi-custom CLCs would be the EK Waterblocks Predator XLC 280 that we benchmarked in our Kraken X62 review. Today’s review also focuses on one of these semi-custom liquid cooling solutions, featuring benchmarks of the Alphacool Eiswolf GPX Pro on a GTX 1080. Our testing looks into thermal performance under baseline conditions (versus a GN Hybrid DIY option), frequency stability and performance, overclocking, and FPS impact. We’ve got a few noise and CPU tests too, though this will primarily focus on the GPU aspect of the cooling. The Alphacool Eiswolf GPX Pro does not work as an out-of-box product, necessitating our purchase of the Alphacool Eisbaer to hook into the system (CPU cooler + radiator). The Eiswolf GPX Pro is a $130 unit, and the Eisbaer cost us ~$145.
This unit was provided by viewer and reader ‘Eric’ on loan for review. Thanks, Eric!
Alphacool Eiswolf Specs & Details
The Eiswolf GPX-1080 deploys a quick-release solution for integration into a semi-custom liquid cooling loop, and differentiates itself from pre-built Hybrid products – like the EVGA Hybrid – by providing full coverage of the PCB. The Eiswolf uses massive aluminum fins and a full-coverage baseplate that directly contacts VRAM and the VRM, using a built-in pump to circulate liquid through whatever the attached radiator may be. This isn’t the same as a full-coverage water block, though; the Alphacool solution is somewhat modular, in that the pump can be separated from the aluminum baseplate and backplate. This theoretically allows for the unit to be upgraded for future GPU releases, though that would require that the mounting hole spacing remains the same.
Out of the pump block protrudes two tubes, an in and an out valve, and those tubes terminate with locking valves. These can then be connected to compatible products, like Alphacool’s Eisbaer 280mm CPU cooler that we used for this testing. EK WB’s quick-release valves will not work with the Eiswolf, but we do have the EK WB XLC Predator and will soon pit it against the Eiswolf.
Alphacool’s tubes use an 11mm outer diameter and are rubberized, encased in coils to help prevent bends that would kink flow. G1/4” threads allow for relatively universal fitment to other liquid cooling products, and you could even avoid using the locking quick releases if desired – though value plummets, since you may as well go for an open loop setup at that point.
The plate provides full coverage of the PCB, but it is not a full-coverage waterblock.
The entire unit is comprised of a backplate, baseplate with integrated pump, and thermal pads that you have to cut to size. You’ll have to buy your own radiator for this to actually do anything. This does have its own pump, so you could technically run it isolated from the CPU loop, but it’s better value to hook into a compatible CPU block.
Alphacool Eiswolf Installation
Installation was somewhat covered in our previous video, where we blindly walked through the installation process to determine how difficult it would really be. Here’s that content:
Installing the Eiswolf isn’t particularly challenging, but it can be annoying at times. The thermal pads, for instance, need to be cut to size by the user – about half the pads are pre-cut, but it’s up to you to do the rest. A thoughtfully supplied template by Alphacool does make this easier, but it just seems like a silly place for the cooling company to save money when they spent so much on the heatsink. Regardless, not the biggest deal in the world. Just sort of poor design choices, considering the competition has better-sized pads that are stickier (and won’t fall off during mounting).
The quick-release system is a little more obnoxious. It’s finnicky. The threads will sort of lock, but if you apply extra pressure for reassurance (to ensure no leakage), they’ll slip and need to be re-tightened. This is a casualty of the weak plastics in the valve. Alphacool could improve this by adding some user feedback for a fully secured and sealed connection, similar to the EK clear ‘pop’ upon locking.
Note that you should also expect a few drops of fluid to leak out during connection. We’d recommend putting down paper towels as a safety, then running a leak test prior to full operation.
Speaking of leaks, during connection to our Eisbaer 280 unit, we had an issue where the cheap plastic locking mechanism and its spring popped out of the valve, resultant of those aforementioned slipping threads. Fortunately, we hold the tubes up when installing these types of things, so no fluid was lost – just be careful when you’re connecting everything.
Some tests below:
Kombustor's implementation of FurMark
GTA V and DiRT Rally
Overclocking and overvolting
Baseline tests & real-world tests
Thermocouples are positioned on MOSFET #3 and on the PCB backside. We use K-Type thermocouples that are calibrated with ice water and boiling water. The thermocouples used are flat and are self-adhesive (from Omega), as recommended by thermal engineers in the industry -- including Bobby Kinstle of Corsair, whom we previously interviewed.
K-type thermocouples have a known range of approximately 2.2C. We calibrated our thermocouples by providing them an "ice bath," then providing them a boiling water bath. This provided us the information required to understand and adjust results appropriately.
As for other concerns, these were largely discussed in that EVGA test planning content. We'd mostly have to look out for (1) thermal conductivity and the impact of a thermocouple in its area of placement, and (2) electrical conductivity and avoiding inadvertent damage to components by accidentally causing an electrical short.
With Kinstle's help, we were able to locate flat thermocouples with an adhesive that will not prohibit transfer of heat between the MOSFET casing and its present thermal pads.
Our next point of concern was smaller, as it'd be easier to resolve and spot: EMI caused by inductors or the power plane PCB. We were able to avoid electromagnetic interference by routing the thermocouple wiring right, toward the less populated half of the board, and then down. The cables exit the board near the PCI-e slot and avoid crossing inductors. This resulted in no observable/measurable EMI with regard to temperature readings.
We decided to deploy AIDA64 and GPU-Z to measure direct temperatures of the GPU and the CPU (becomes relevant during torture testing, when we dump the CPU radiator's heat straight into the VRM fan). In addition to this, logging of fan speeds, VID, vCore, and other aspects of power management were logged. Because VRMs are not measurable through software, our direct thermocouples will handle that aspect of testing.
The test platform is detailed below:
|GN Test Bench 2015||Name||Courtesy Of||Cost|
|Video Card||EVGA GTX 1080 FTWs||EVGA||~$740|
|CPU||Intel i7-5930K CPU 3.8GHz||iBUYPOWER
|Memory||Corsair Dominator 32GB 3200MHz||Corsair||$210|
|Motherboard||EVGA X99 Classified||GamersNexus||$365|
|Power Supply||NZXT 1200W HALE90 V2||NZXT||$300|
|SSD||HyperX Savage SSD||Kingston Tech.||$130|
|Case||Top Deck Tech Station||GamersNexus||$250|
|CPU Cooler||NZXT Kraken X61 CLC||NZXT||$110|