Liquid-cooled graphics cards have blown up over the past year. They've existed before, but never to the level of publicity as spurred-on by AMD's Fury X and nVidia's high-end Maxwell board partners. With the MSI Sea Hawk 980 Ti, we explained why CLC systems for GPUs make excellent sense in the correct use case scenario. We mostly called attention to the obvious thermal reduction on the silicon, increased power efficiency by reducing capacitor leakage, and the ability for high-heat, “Big GPUs” to push more substantial overclocks. This is thanks to an avoidance of thermal throttling of the clock, meaning we become more limited by overall chip stability and BIOS vCore locks.
But liquid doesn't always make sense. CLCs drive the BOM up, increasing what the user pays for the solution. Most CLCs, depending on supplier (read about who really makes liquid coolers), are only good for a few years – five, on average – and that's undesirable to users seeking serious endurance. I'd imagine that most of our audience aims to build or upgrade systems at least once within a five-year period, perhaps mitigating the impact of this consideration. The issue is further diminished by just how easy it is to maintain these things: popping in a new cooler will get it up-and-running again, and they're fairly standardized (in the case of the 970 Hybrid or Sea Hawk, anything by Asetek will work). More work than required for an air-cooled card, but even those face decay from prolonged service life (often thermal compound or pads need to be re-applied).
At the surface, the GTX 970 Hybrid doesn't appear to like an application where “liquid makes sense.” That's what testing is for, and we'll look at use case scenarios for overclocking, ultra-low thermal systems, SFF rigs with thermal concerns, and more.
In this review of EVGA's GTX 970 Hybrid, we benchmark stock and overclocked performance in games (FPS), temperatures, and power consumption.
EVGA GTX 970 Hybrid Specs
| EVGA 970 Hybrid | EVGA 970 SSC | MSI GTX 970 Gaming | GTX 970 Stock | |
| Base Clock (GPU) | 1140MHz | 1190MHz | 1140MHz | 1050MHz |
| Boost Clock (GPU) | 1279MHz | 1342MHz | 1279MHz | 1178MHz |
| Memory Clock | 7010MHz | 7010MHz | 7010MHz | 7000MHz |
| Mem Spec | 4GB GDDR5 256-bit |
4GB GDDR5 256-bit |
4GB GDDR5 256-bit |
4GB GDDR5 256-bit |
| Price | $400 | $350 | $350 | $310 |
At its reference clocks, the original GTX 970 oscillated at 1050MHz base, 1178MHz boost. EVGA's lineup alone contains cards with clocks at 1050MHz (stock) for around $310, 1165MHz cards (“SC Gaming”) at $345, 1190MHz cards (“SSC”) at $350, and now the GTX 970 Hybrid (1140MHz) at $400. And that's just one company – between MSI, ASUS, Gigabyte, PNY, and the half-dozen other manufacturers, clock frequencies scale nearly 200MHz from reference.
The Hybrid's core clock is actually 50MHz lower than EVGA's 970 SSC (1190MHz), effectively guaranteeing lower stock FPS throughput than a device that runs $50 cheaper. The main draw of the Hybrid is clearly its cooling solution, a value-add that will reduce internal case temperatures – useful especially in SFF or shoebox builds – and theoretically increases OC headroom with better thermal management. We'll test that more below.
As with previous CLC-cooled GPUs that we've reviewed, the 970 Hybrid relies on a CLC for direct-to-GPU cooling, sticking with a blower fan for VRM and VRAM dissipation. The CLC used here is the same Asetek unit as found in the GTX 980 Ti Hybrid (detailed here), tuned for GPU cooling with its copper extrusion in the coldplate. MSI & Corsair previously partnered to produce the Sea Hawk 980 Ti, which differs from this approach only by way of deploying a Corsair CLC (built for CPUs) rather than the Asetek GPU CLC. The difference, as we saw in testing, was massive.
The 970 Hybrid uses an identical CLC to the 980 Ti Hybrid. The faceplate is different – uglier, we think, though we don't much comment on aesthetics – but the card design is similar overall. The 970 Hybrid uses a reference board design, which means you should eventually be able to buy a kit to convert other reference 970s to a Hybrid system (EVGA or not). The CLC is the same, blower/VRM cooling is the same, and the overall assembly is the same as the other EVGA “Hybrid” series cards.
Test Methodology
We tested using both of our GPU test benches, as we are in the process of merging into a single GPU test bench. Both are detailed in the tables below. Our thanks to supporting hardware vendors for supplying some of the test components.
The latest AMD Catalyst drivers (15.11.1) were used for testing, except in the case of the Fury X and R9 380/390, which were all on loan and have not been updated since ~15.7. NVidia's 359.00 drivers were used for testing the latest games. Game settings were manually controlled for the DUT. All games were run at presets defined in their respective charts. We disable brand-supported technologies in games, like The Witcher 3's HairWorks and HBAO. GRID: Autosport saw custom settings with all lighting enabled. All other game settings are defined in respective game benchmarks, which we publish separately from GPU reviews. Our test courses, in the event manual testing is executed, are also uploaded within that content. This allows others to replicate our results by studying our bench courses.
Each game was tested for 30 seconds in an identical scenario, then repeated three times for parity.
Z97 bench:
| GN Test Bench 2015 | Name | Courtesy Of | Cost |
| Video Card |
This is what we're testing! |
- | - |
| CPU | Intel i7-4790K CPU | CyberPower |
$340 |
| Memory | 32GB 2133MHz HyperX Savage RAM | Kingston Tech. | $300 |
| Motherboard | Gigabyte Z97X Gaming G1 | GamersNexus | $285 |
| Power Supply | NZXT 1200W HALE90 V2 | NZXT | $300 |
| SSD | HyperX Predator PCI-e SSD | Kingston Tech. | TBD |
| Case | Top Deck Tech Station | GamersNexus | $250 |
| CPU Cooler | Be Quiet! Dark Rock 3 | Be Quiet! | ~$60 |
X99 Bench:
| GN Test Bench 2015 | Name | Courtesy Of | Cost |
| Video Card |
This is what we're testing! |
- | - |
| CPU | Intel i7-5930K CPU | iBUYPOWER |
$580 |
| Memory | Kingston 16GB DDR4 Predator | Kingston Tech. | $245 |
| 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 X41 CLC | NZXT | $110 |
Average FPS, 1% low, and 0.1% low times are measured. We do not measure maximum or minimum FPS results as we consider these numbers to be pure outliers. Instead, we take an average of the lowest 1% of results (1% low) to show real-world, noticeable dips; we then take an average of the lowest 0.1% of results for severe spikes. Anti-Aliasing was disabled in all tests except GRID: Autosport, which looks significantly better with its default 4xMSAA, and Black Ops III. HairWorks was disabled where prevalent. Manufacturer-specific technologies were used when present (CHS, PCSS).
- Battlefront test methodology can be found here.
- Black Ops III test methodology is here.
- The Witcher 3 test methodology is here.
- Fallout 4 test methodology is here.
- Assassin's Creed Syndicate methodology is here.
Overclocking was performed incrementally using MSI Afterburner. Parity of overclocks was checked using GPU-Z. Overclocks were applied and tested for five minutes at a time and, if the test passed, would be incremented to the next step. Once a failure was provoked or instability found -- either through flickering / artifacts or through a driver failure -- we stepped-down the OC and ran a 30-minute endurance test using 3DMark's FireStrike Extreme on loop (GFX test 2).
Thermals and power draw were both measured using our secondary test bench, which we reserve for this purpose. The bench uses the below components. Thermals are measured using AIDA64. We execute an in-house automated script to ensure identical start and end times for the test. 3DMark FireStrike Extreme (GFX test 2) is executed on loop for 25 minutes and logged. Parity is checked with GPU-Z.
Thermals, power, and overclocking were all conducted on the Z97 bench above.
