When it was first released Supreme Commander was one of the most demanding RTS (Real-Time Strategy) games ever developed in our opinion. In fact, at the time we had trouble creating a computer that was fast enough to play this game in all its glory. This was a little disappointing and it meant that most could only truly appreciate the game years later, when pricing of the required hardware dropped.

Released back in early 2007, the original Supreme Commander was a hit and is still very popular today. However today we are moving on, as Gas Powered Games releases Supreme Commander 2 for the PC and Xbox 360. This new RTS features full cinematic detail, and in doing so, it requires quite a lot of processing power. With all the visual enhancements enabled, Supreme Commander 2 looks stunning, easily making it one of the best looking RTS games of all time.

On the other hand, as a result you can expect many computers to struggle playing this game with all the eye candy turned on. This can be quite disappointing for some, because as good as the game play and fun factor of Supreme Commander 2 is, having visual settings set at their maximum really does this game justice.

Gas Powered Games has actually been very forth coming with the system specification requirements for Supreme Commander 2. They recommend that users have at least a 3.0GHz dual-core processor such as the Core 2 Duo E6850 or Athlon II X2 245, along with a GeForce 8800 GS or Radeon X1800 graphics card. While none of these specifications are by any means outrageous, they are higher than that of other modern real-time strategy games.

Of course developers are often a fair way off target when it comes to system specifications and even their recommended specifications are never enough to really enjoy the game in all its glory. This is where we come in, testing a range of graphics cards to determine which ones will be worth using and which ones wont.

Given that we feel all games should be played using maximum in-game quality settings, this is generally what we test with. That said, because Supreme Commander 2 is such a demanding RTS game and so many of you are going to want to play it, we have also included a great deal of testing using the medium quality preset as well.

Before we jump into the testing, here is a short summary of the Supreme Commander 2 game. Set 25 years after the events of the original game, Supreme Commander 2 begins with the assassination of the newly elected president of the colonial defense coalition. The CDC members deny involvement, blame each other, and a galactic war ensues!

Like the original, this second installment lets the gamer experience brutal battles on a massive scale. It is possible to create enormous customizable armies and experimental war machines that can change the balance of power at any given moment. Players can take the role of one of the three enigmatic commanders, each representing a unique faction with a rich story that brings a new level of emotional connection to the RTS genre.

By default these presets do not enable any level of anti-aliasing. Rather, the high quality settings employs 8x anisotropic filtering, while the medium quality preset uses 4x anisotropic filtering. However please note that our high quality test results were recorded with 8x anisotropic filtering and 8x anti-aliasing enabled.

The latest official drivers were used for all graphics cards, which saw both Crossfire and SLI working correctly. Something we noticed early on in our testing was that the Radeon HD 5000 series graphics cards received around a 5-10fps performance boost when going from the Catalyst 10.1 to the 10.2 drivers. The high-end cards all gain around 10fps, while the mid-range/low-end cards were boosted by about 5fps.

For measuring frame rates we relied on Fraps, where we recorded three minutes of single player game play. The Intel Core i7 920 processor was used this time and we overclocked it to 3.7GHz in an attempt to remove any CPU bottlenecks that could influence high-end graphics cards scores. Please note that each test was conducted 3 times, with a pre-cache test as well. This gives us very accurate average frame rates.

Finally, we know that many of you will be interested in CPU scaling performance as well, and therefore we down clocked the Intel Core i7 920 processor to see what kind of impact this had on performance. In addition to this we have also tested a range of processors/platforms at the same frequency to determine how well different processors handle this new game.

Finally, we will be looking for an average of 60fps for stutter-free game play.

Hardware - Intel Core i7 965 @ 3.70GHz - x3 2GB G.Skill DDR3 PC3-12800 (CAS 9-9-9-24) - Seagate 500GB 7200-RPM (Serial ATA300) - Radeon HD 5870 (1GB) - Radeon HD 5850 (1GB) - Radeon HD 5830 (1GB) - Radeon HD 5770 (1GB) - Radeon HD 5750 (1GB) - Radeon HD 5670 (512MB) - Radeon HD 5570 (1GB) - Radeon HD 5450 (1GB) - Radeon HD 4890 (1GB) - Radeon HD 4870 (1GB) - Radeon HD 4850 (1GB) - Radeon HD 4830 (512MB) - Radeon HD 4770 (512MB) - Radeon HD 4670 (1GB) - Radeon HD 4650 (1GB) - Radeon HD 3850 (512MB) - Radeon X1950 Pro (256MB) - GeForce GTX 285 (1GB) - GeForce GTX 275 (896MB) - GeForce GTX 260 (896MB) - GeForce GT 240 (512MB) - GeForce 9800 GT (512MB) - GeForce 9600 GT (512MB) - GeForce GT 220 (512MB) - Asus P6T Deluxe Software - Microsoft Windows 7 Ultimate (64-bit) - ATI Catalyst 10.2 - Nvidia Forceware 196.21

In the screen shots above the commander unit features much better texture quality in the first high quality screen shot. The ground texture is also significantly improved and the shadows are much sharper. The use of AA also makes a huge difference as the top image features none of the jagged edging that you will see in the medium quality image.

The next two shots compare water visual quality and while the medium quality water does not look bad, the high quality water is far more detailed. However the biggest difference can be seen around the units and when they are moving. While we cannot show you how much better the units look moving though, you can see the water and the added detail around the boat in the high quality image.

The Radeon HD 5870 is the fastest single-GPU graphics card available and as such it was the fastest graphics card at 2560x1600 with 8xAA/8xAF enabled in Supreme Commander. This was an impressive 38% performance increase over the GeForce GTX 285, the previous holder of the world’s fastest single-GPU graphics card title.

The Radeon HD 5870 was just 23% faster than the Radeon HD 5850, and with an average of 73fps, it would seem that the Radeon HD 5850 is a much better value option for playing Supreme Commander. The new Radeon HD 5830 was 17% slower than the Radeon HD 5850, while it was just 11% slower than the GeForce GTX 285, and finally it matched the performance of the GeForce GTX 275 with 60fps on average.

Further down we have the GeForce GTX 260 with an average of 51fps, making it just 5% faster than the Radeon HD 4890. The Radeon HD 4890 was just 1.5% faster than the Radeon HD 4890, which rendered an average of 47.8fps. The Radeon HD 4870 started to dip down into the low 40’s and this is where gamers will start to notice lag spikes caused by sudden drops in frame rates.

The Radeon HD 5750 averaged just 39fps, while the Radeon HD 4850 was even slower with 34fps. Still, given we are using maximum in-game quality settings here with 8xAA/8xAF enabled, these results are still quite impressive. That said, anything below the Radeon HD 4870 on this graph was not able to deliver what we call perfectly playable performance.

At 1920x1200 we start to run into a CPU bottleneck it seems, with the high-end graphics cards limited to 98–99fps. Here we have further evidence that high-end graphics cards, such as the Radeon HD 5870 and GeForce GTX 285, are a waste for Supreme Commander 2, particularly at this resolution where the GeForce GTX 275 and Radeon HD 5830 graphics cards are just as fast.

Moving further down we have the GeForce GTX 260, Radeon HD 5770 and Radeon HD 4890, which all rendered more than 70fps on average. This is more than enough performance for perfectly playable performance. The Radeon HD 5750, which costs just $140 US, delivered 60fps on average which is quite impressive.

The Radeon HD 4850, 4770 and GeForce 9800 GT all rendered 50fps or more, which is acceptable in our opinion but about as low as you would want to go. Still, we are using maximum in-game quality settings with 8xAA/8xAF, so there are plenty of settings that could be tweaked that would have very little impact on the visual quality.

Now at 1680x1050 the 98–99fps bottleneck is even more apparent, and this is either a limitation put in place by the game or it is a limitation of the processor. In either case, it meant that anything as fast or faster than the GeForce GTX 275 was capable of reaching this limit, which included 4 other tested graphics cards.

This time the GeForce 9800 GT averaged 61fps, meaning that it and the other dozen graphics cards above it on the graph were all able to deliver playable performance. For owners of Radeon HD 4830, 5670, 5570 and GeForce 9600 GT/GT 240 graphics cards, the maximum quality settings with 8xAA/8xAF is a little much and this is not at all surprising. Still, some fine tuning of these settings should see some of these cards provide playable performance.

Now we are testing with reduced visual quality settings by enabling the “medium” quality preset, which also disabled Anti-aliasing and reduces the Anisotropic filtering level to 4x. The high-end graphics cards have been removed as they could handle 2560x1600 with maximum in-game visuals, so there was little point in testing them. However we have kept graphics cards such as the Radeon HD 5830 and GeForce GTX 260 as a reference. Furthermore, a number of older graphics cards have been added to the mix.

As you can see, at 2560x1600 the Radeon HD 5830 managed 73fps while the GeForce GTX 260 averaged 71fps. The Radeon HD 5770 dropped down to 65fps, while the Radeon HD 4870 managed just 57fps. Most of these mid-range/low-end graphics cards struggled with the 2560x1600 resolution, even with the medium quality settings in place. That said, we are not all that surprised as this is an extreme resolution.

Now at 1920x1200 we again find that 98–99fps limit which has been placed on the Radeon HD 5830 and GeForce GTX 260 graphics cards while the Radeon HD 5770 was stuck at 97fps. The Radeon HD 4870 reached 93fps this time, while the 4850 was not far behind with 78fps. This time the Radeon HD 4830 was able to comfortably deliver playable performance, while the Radeon HD 5670 and GeForce 9600 GT/GT 240 struggled, and of course this also includes all the other graphics cards below them.

The 1680x1050 resolution where we expect many of you will be playing Supreme Commander 2, is very graphics card friendly. The Radeon HD 5750 and the other 5 graphics cards above it all reached that 98–99fps limit. The Radeon HD 4770 also came close with 93fps, while the GeForce 9800 GT was slightly slower again with 89fps.

The Radeon HD 5670 and GeForce GT 240 managed 68fps, while the GeForce 9600 GT dropped down to 66fps. Amazingly the old Radeon HD 3850 averaged 60fps here, which is most impressive. Again the Radeon HD 4670 and Radeon HD 5570 struggled to deliver playable performance at this resolution, which was very disappointing.

Something we found a little unexpected was the Core i7 920 scaling results, which showed no difference in performance when going from 3.70GHz down to 2.22GHz. This suggests to us that there is in fact a 100fps limit in place that has been imposed by Supreme Commander 2. This means even when using a powerful graphics card such as the Radeon HD 5870, a Core i7 processor down-clocked to just 2.22GHz will be up to the task!

Here we are briefly comparing a few different CPU architectures. It would seem that for the most part the processor does not play a huge roll in Supreme Commander 2, as these results are nothing like those found when testing StarCraft II. Even a lowly Sempron 140 was just 11% slower than the mighty Core i7 920 (please note both CPU’s are operating at the same frequency) and 2560x1600.

With a single core in use the average CPU utilization recorded when playing Supreme Commander 2 was 93.6% and as you can see, that single core was working flat out.

With two cores enabled the average work load halves to 50.2%, though as you can see the work load is not distributed evenly. The first core is still working the hardest at about 60-70% load, while the second core works at around 30–40%.

Clearly, like StarCraft II and most games for that matter, Supreme Commander 2 is only dual-core optimized, and we see this when looking at the quad-core results. Although average CPU utilization has dropped to just 26.3%, we see that the third and fourth cores are doing very little.

When it came to benchmarking Supreme Commander 2 to see how demanding this new RTS would be, we were pleasantly surprised. The original version was very demanding and I recall testing it just as we have Supreme Commander 2. At the time we threw the best processor and graphics cards at Supreme Commander and even they had a hard time delivering acceptable performance with maximum in-game quality settings enabled.

Here are some comments regarding CPU performance from our conclusion 3 years ago when first testing Supreme Commander…

“The more processing power you throw at this system the better, though I do not agree that multi-core processors ensure excellent performance. I ran a few tests using Core 2 Quad and Duo processors and noticed that the second, third or fourth cores did very little when compared to the first. In the most intense battles the first core would operate at around 90 ~ 100% where as additional cores worked between 20 ~ 50%.”

Also come comments of graphics card performance…

“The graphics card is obviously another key component to improving performance, though unfortunately the latest and greatest solutions again do not guarantee stellar performance. Having said that, anything less than a Radeon X1950 Pro is going to make life difficult when playing on the larger maps with more than 2 players. In fact, using the low quality settings at 1280x1204 saw the X1950 Pro render just 24fps on average. The ultra powerful GeForce 8800 GTX was not a great deal better, producing an average frame rate of just 31fps. The maximum in-game quality settings saw the 8800 GTX fall to an average frame rate of just 22fps.”

Thankfully the situation with Supreme Commander 2 appears very different, at least in respect to the graphics card performance. The fact that we saw graphics cards such as the new Radeon HD 5830 rendering over 90fps at 1920x1200 with maximum in-game quality settings in place was impressive. Even much cheaper graphics cards, such as the Radeon HD 5750, delivered playable performance at this resolution, with an average of 60fps.

Those playing at resolutions such as 1680x1050 or lower can enjoy Supreme Commander 2 in all its glory with older graphics cards such as the GeForce 9800 GT or Radeon HD 4770, which is impressive. That said, Supreme Commander fans that purchase the Radeon HD 5570 will be screaming, as this graphics card put plainly, sucked, even at 1680x1050 with an average of 36fps.

Another interesting discovery was the CPU performance, which appears to have little impact on the game. When comparing platforms we found very little difference in performance between the Core i7 920 and Phenom II X4 965 processors. Furthermore, there was almost no difference between the Phenom II X4 965 and Athlon II X4 635 processors.

CPU utilization was difficult to accurately measure in this game as the processor only starts to become a factor once a large amount of units exist. For example, at the start of any game we found that the CPU utilization would often sit on 0% with our Core i7 920 processor. That said, zooming out would increase utilization to about 10%. However once a large amount of units began to amass we were able to max out a single core.

Finally, for those that cannot play using the high quality settings, Supreme Commander 2 does not look too bad when using the medium quality settings. Furthermore, we found that the medium quality setting would enable slower graphics cards, such as the Radeon HD 3850 and GeForce 9600 GT, to deliver playable performance at 1680x1050 which is quite impressive.