Two GPUs are better than one seems to be the driving philosophy behind the unique series called Mars by Asus, and iLLGaming got the chance to test if that was true. Could two GK104 cores as seen in the GTX 760s really provide enough power to beat a Titan at a fraction of the cost as Asus claims?
Mars is the name given by Asus to its dual GPU single card series built with NVIDIA cores just as Ares is for the AMD counterpart. This is a card designed by Asus in-house and is not built to some manufacturer specification released by NVIDIA. The MARS 760 is the newest entry in the series and leaves no surprise as to origin of the twin GPU cores it holds.
Specification of the MARS 760
Being a custom graphics solution, the MARS 760 features two overclocked GK104 cores with 2GB of GDDR5 VRAM dedicated to each core which adds up to the 4GB prominently advertised on the box. The GK104 cores are fabricated using the 28nm process. The card requires two 8 pin power connectors to feed it’s approximately 300W power requirement.
The dimensions of the two slot card are roughly 28cm x 11.1cm x 4cm and it features a custom Asus ROG (Republic of Gamers) cooler with twin 85mm fans. The side of the card is dominated by a MARS logo that pulsates red when the card is in operation. There is provision for a SLI connector and the card is designed to support up to 4 way SLI with another of its ilk. This feat of engineering was accomplished by the joint efforts of Asus and NVIDIA engineers. The rear panel has two DVI-D connectors, one DVI-I connector and one mini DisplayPort connector. The top of the card is also covered by a heat sheath.
The launch price for the MARS 760 is ₹49,500 which should have dropped since the card’s release in November 2013.
Packaging
The card comes neatly packaged in a box with a flip cover and transparent window to show off the beast inside. It comes with the usual Driver CD, Quick installation Guide, a DVI to D-Sub adapter, a 2x6pin to 1x8pin PCI-E power adapter and includes a unique Asus ROG Magnetic Case Badge for flaunting your new purchase.
[divider]
Testing Methodology
We tested the MARS 760 not only for the average Frames per Second(FPS) but also for the 99th Percentile Frame time which tells us about the performance of the GPU within the second. Within the second testing it is useful to understand micro-stutter which can render a game unplayable despite FPS being high. Fraps 3.5.99 allowed us to calculate both.
Since the card is a Top End GPU, we decided to use highest possible settings in our benchmark games. We tried to disable CPU dependent settings or minimise their impact where possible. VSync and frame buffering were disabled for testing. All tests were run at 1920×1080 on a single monitor configuration.
Test System
CPU: AMD Phenom II X4 965 BE C3 @ 3.8GHz
Motherboard: ASRock 970 Extreme 3
RAM: 2x4GB G.Skill Sniper CL9 1600MHz, 2x2GB G.Skill Ripjaws CL9 1600MHz
PSU: Corsair TX650 650W (for MARS 760), Seasonic S12II Bronze 620W (for others)
HDD: 2xSeagate Barracuda 7200.12 1TB, 1xWesten Digital Red 3TB
OS: Windows 8.1 x64
NVIDIA Driver: 334.67 (Beta Driver)
While this system may not look like a purpose built test rig, we decided to use a normal usage PC so as to better reflect real world scores of the card. The games were also tested with a few applications like Antivirus, Browser, VOIP tool and others running in the background to get a realistic usage scenario. All HDDs were thoroughly defragmented prior to usage. Due to lack of equipment we were unable to conduct acoustic and power testing.
3DMark 11 Firestrike
3Dmark is an artificial Benchmarking tool whose Firestrike Test is very thorough on DirectX 11 cards powering High-end PCs. The full test run for Firestrike includes 2 GPU only tests, a CPU dependent Physics Test and a Combined Graphics and Physics Test. The Tool is also useful for stress testing a GPU when run on loop.
Given that we are looking for the Performance of the Card itself, one should look at the Graphics score and the FPS for Graphics tests 1 and 2. The Physics and Combined tests are CPU dependent which is the limiting factor of our test rig.
Batman Arkham Origins
Batman Arkham Origins is a game that’s been supported by NVIDIA and utilises their PhysX technology to handle physics. We used highest possible settings. PhysX was also set to Enhanced. We ran the inbuilt benchmark tool with a FRAPS timed run of 120 seconds.
The closeness of the results for the MARS 760 and the GTX 770 was surprising and might possibly indicate better optimisation for single GPU core solutions. The 99th percentile frametime for the GTX 770 were also lower than the MARS 760 which is indicative of the effect of SLI in the dual GPU card.
The 99th Percentile graph clearly shows the increased variance in frametimes caused by using two GPU cores in SLI instead of one. The drops in FPS have corresponding spikes in the Frametime graph which were caused by the game loading new areas. There wasn’t any noticeable microstuttering and the game ran smoothly.
Battlefield 4
Battlefield 4 uses the Frostbite 3 engine to push the visual processing boundaries of current hardware. Since Mantle is only for AMD cards, we ran the DX11 version with the highest possible settings. The game offers no benchmark tool, so we used areas from the first single player campaign mission for the FRAPS run.
The MARS 760 shows its raw power with a 10FPS lead over the GTX 770. However, it loses out when comparing the 99th percentile frametime which is it’s weakness due to its multi GPU nature.
The game runs at quite smoothly with everything cranked up having a high average FPS, but hard-core players may want to lower settings in the hopes of going even higher for the fast paced multiplayer. The frametime graph however, is all over the place as compared to a single GPU solution. The microstuttering is not so noticeable as to be immersion breaking though.
Bioshock Infinite
The Unreal Engine 3 powered Bioshock Infinite really pushed the boundaries of visual effects achievable with the ageing engine. It comes with a built in benchmark mode which runs for a lower time period than the standard 120s we used in other tests.
While the MARS 760 outperforms the 770 in average FPS, it seems to have a greater dip in minimum FPS. The 99th percentile frametimes are surprisingly nearly identical.
The frametime graph is very rough and the FPS varies significantly as well. The microstuttering did show in a few places during the test run. It’s possible a bit more optimisation could fix these issues.
Company of Heroes 2
Relic’s Company of Heroes 2 is a tough nut to crack for quite a few GPUs, though it’s dependent on CPUs to a great degree as well. We used highest possible Settings with Physics turned off and Low AA for the short 45s benchmark run the game offers.
The game is CPU dependent and also seems to utilise only one GK104 core and 2GB of VRAM instead of using it in SLI. Thus it scores worse than the GTX 770 in our tests since it effectively functions as a single GTX 760.
The game appeared quite choppy and the frametimes support that. FPS was throttled when CPU heavy action took place on screen. The apparent lack of multi GPU recognition seems to be a game issue though but it does bring out one of the negative points of using a multi GPU setup.
Crysis 3
CryEngine 3 was built to push the PC Hardware to its limits with Crysis 3 and deliver Maximum Eye Candy. We used the highest possible settings for the purpose of our test. The test run consisted of a section of the first mission for 120s as the game surprisingly offers no benchmark tool.
While the 770 makes for a barely playable game, the MARS 760 lets one comfortably play the game in all its graphical glory. There are some dips to barely playable territory but majority of the gameplay is smooth. The 99th percentile frametime is quite another story though.
The gameplay was smooth as can be seen from the frametime graph though there are spikes which correspond with sudden dips in FPS. The graph is not as smooth as that for a single GPU solution which can be explained by the SLI nature of the card.
Far Cry 3
Far Cry 3 is a visually impressive game that lacks a benchmark mode. We used highest possible settings and the test run consisted of a 120s sample from the game’s single player campaign where we spent some time goofing off in the open world.
While the minimum FPS is equal, the MARS 760 performs significantly better than the GTX 770 in this game. For once, it also has a lower 99th percentile frametime.
While the frametime graph is quite rough there wasn’t any visible microstuttering and the gameplay was fluid though there are quite a few rises and dips on the FPS graph. The relatively concentrated frametimes are indicative of a well optimised game.
GRID 2
GRID 2 from Codemasters is one of the few games in our test suite that has no allegiance to either NVIDIA or AMD, which makes it a neutral candidate to better judge the GPU’s performance. We used highest possible Settings to run the benchmark tool offered by the game.
As expected from a neutral game, the performance of the GTX 770 is quite close to the MARS 760 though the latter dips lower in minimum FPS. Surprisingly, the MARS 760 has a lower 99th percentile frametime.
The game had a relatively smooth run as regards frametimes except for that one spike at the beginning. This can be seen reflected in the relatively constant FPS graph. No microstuttering was noticeable.
Sleeping Dogs
While the game is ageing, Sleeping Dogs is still a stunner when it comes to running benchmarks. We pushed the settings to highest possible and ran the game’s built in benchmark run.
The average fps is relatively closer in this game which might be caused by the game seemingly throttling FPS to 60. The MARS 760 does outperform the GTX 770 in all the parameters though.
The Frametime Spikes that are seen are caused by the benchmark transitioning between scenes and there was no microstuttering seen in the rendered scenes. Curiously the FPS drops in the closed spaces in interior areas while remaining quite constant in open world areas. The relatively flat FPS graph indicates the game might have throttled the FPS output to 60FPS.
Tomb Raider
Tomb Raider introduced us to a new Lara with fabled TressFX hair that behaves a lot more realistically than the pre-rendered mop we were used to. However the card does not seem to cope well with this technology and we had to drop it in our test run. We used highest possible settings with TressFX on. The Test run was shorter since the benchmark tool offered by the game runs for less than out 120s target time.
The MARS 760 beats the GTX 770 by a huge margin in this game and has a much lower 99th percentile frametime. It seems the multi GPU solution is better able to absorb the toll taken by TressFX than the single GPU one.
The frametime graph is much smoother than in other games which indicates a bit of optimisation for multi GPU solutions in the game. However, the large spike was noticeable as microstuttering while running the benchmark.
Total War: Rome II
Total War: Rome II is another CPU heavy game that offers significant visual goodness. It offers a benchmark tool that focuses more on GPU power though and that is what we used for our test run of 120s. We set the game to Extreme and Unit Size to Small so as to reduce the impact of the CPU. Unlimited Video Memory was off so the game could scale down visual settings if it reached a bottleneck.
Total War: Rome II is another game that fails to recognise the second GK104 core in SLI along with its 2GB VRAM and effectively makes the MARS 760 a normal GTX 760. The game is quite graphically demanding with neither card able to max it out.
The Frametimes show some massive spikes but given that the FPS is in barely playable territory, it matters little. The game could do with optimisation so as to better utilise the power of Multi GPU solutions.
[divider]
Acoustics
While we were unable to conduct detailed acoustics testing on the card due to lack of equipment, we were able to get a rough idea of how quiet the card is even at full load. We found that the GPU made significantly less noise than our HDDs on idle though it was audible over them at full load. We kept the case at 1m distance with the side panel open. One interesting thing to note though is that the fans would spin at maximum RPM every time the system booted up and were at their noisiest then.
Temperatures
The idle temperature for the card lies in the range of 36-38°C as compared to our ambient temperature of 30°C. Under maximum load the fans allow the temperature to rise up to 80°C and then kick in to keep it there. Try as we might, Asus’s cooling solution rarely let the load temperatures exceed 85°C which is a credit to its effectiveness, given that the fans weren’t running at maximum RPM.
Power
Asus rates the MARS 760 at roughly 300W which makes for quite a power hungry card. Note that an individual GTX 760 is rated at 170W so this solution does consume less power than a normal SLI setup of two GTX 760s.
Overclocking
Despite being factory overclocked by 2-3% on base and 13-14% for memory the card still leaves room for overclocking up to 18% for base and 33% for memory which makes it quite a stellar overclocker. The high quality components used in its construction allow for such gains. It has always been difficult to get Multi GPU configurations to overclock properly, which makes the MARS 760 all the more impressive. All our tests however, were conducted in the stock configuration.
Feature Set
Asus bundles its GPU tweak utility for easy overclocking of the card. With NVIDIA’s GeForce Experience, the card can support features like Shadowplay and game streaming but, there is no way to control its pulsating red MARS logo.
[divider]
Conclusion
The uniqueness of the Asus MARS 760 is its most appealing aspect. The Republic of Gamers line-up of cards boasts of some stellar performers and this is one of them. It sure has a lot of Bragging rights attached to it and the packaging shows it with that magnetic pin.
The cooler design, build and component quality are top notch and allow for quite a bit of overclocking. While the card does outperform a Titan at a fraction of the cost, it’s still pricier than two individual GTX 760s. The value of the card is also diminished when you compare AMD offerings which give about as much performance for an even lower price.
The twin 8 pin PCI-E connectors require a significantly beefy PSU. The lack of HDMI ports or converters is also a drawback if one wishes for easy connection to a TV.
The card also comes with its share of problems associated with multi GPU solutions such as higher variance in frametimes and lack of multi GPU support in some games. While its brute power can muscle through games that support it, it can struggle a lot in games that do not. So one is advised to only consider the card if one plays games that are optimised for it.
While the card has quite a few negatives, its performance and quality are beyond doubt. For this we award it the iLLGaming Silver Badge.
We are extremely grateful to Asus for providing us with a test sample for reviewing.
+Allows for SLI in a single PCI-Ex16 slot motherboard
+Good for High-End 1080p gaming with multi GPU optimised games
+Uniqueness of GPU adds to Brag value of PC
-Pricing is greater than the cost of two individual GTX 760s
-Not all games support multi GPU solutions and the card performance suffers in these
-Needs a beefy PSU with two 8pin PCI-E connectors
[…] 2x4GB G.Skill Sniper CL9 1600MHz, 2x2GB G.Skill Ripjaws CL9 1600MHz PSU: Corsair TX650 650W (for MARS 760), Seasonic S12II Bronze 620W (for others) HDD: 2xSeagate Barracuda 7200.12 1TB, 1xWesten Digital […]