In the past, when a seriously fast graphics card has been launched, we have taken the time to pair it with a range of CPU’s in order to determine just how much processing power it requires to deliver breath taking frame rates. This was first done using the GeForce 8800 GTX, which reigned supreme well past the typical life expectancy of a graphics card. It was not until the GeForce 9800 GX2 was released that we bothered to look at CPU scaling in video games again.

This latest generation of AMD and Nvidia graphics cards has brought some real monsters to light, such as the GeForce GTX 280 and now the Radeon HD 4870 X2. Originally we had planned to use the GeForce GTX 280 for this latest CPU scaling article. However given how much faster the Radeon HD 4870 X2 is in games that scale well with Crossfire, we decided to use an ATI based graphics card for the first time.

Now because this is primarily a CPU scaling article, we have selected benchmarks that do work well with Crossfire technology. There would have been little point in trying to work out what kind of processor gamers are going to require in order to power something like the Radeon HD 4870 X2 at full speed if only one GPU is working. Therefore the list of games includes Company of Heroes, Devil May Cry 4, Enemy Territory Quake Wars, Supreme Commander, and Unreal Tournament 3.

Rather than focus on the Intel LGA775 platform as we have in the past, a range of AMD Phenom X4 processors will be introduced, along with a number of older AMD Athlon64 X2 processors. The Athlon64 X2 stood out as the number one gaming processor for many years, and it was only once the Core 2 Duo series was introduced that this all changed. Therefore we suspect that many gamers are still fragging away with older Athlon64 X2 processors, so these results should prove useful.

In games such as Enemy Territory Quake Wars, Supreme Commander, and Unreal Tournament 3 for example, the Radeon HD 4870 X2 is hands down the fastest single graphics card in the world right now. So the idea behind this article is to determine what kind of CPU gamers will require to get the most value out of this high-end graphics card. Previous CPU scaling articles have told us that a 3.0GHz Core 2 Duo processor offers high-end gamers the best performance/price ratio, so it shall be interesting to see how much the Radeon HD 4870 X2 takes advantage of raw processing speed.

* = Overclocked
** = Under-clocked

Hardware - Core 2 Quad Q6600 266 x 6 = 1.60GHz (DDR3-1066)** - Core 2 Quad Q6600 266 x 7 = 1.86GHz (DDR3-1066)** - Core 2 Quad Q9650 333 x 6 = 2.00GHz (DDR3-1333)** - Core 2 Quad Q6600 266 x 8 = 2.13GHz (DDR3-1066)** - Core 2 Quad Q9650 333 x 7 = 2.33GHz (DDR3-1333)** - Core 2 Quad Q6600 266 x 9 = 2.40GHz (DDR3-1066) - Core 2 Quad Q9650 333 x 8 = 2.66GHz (DDR3-1333)** - Core 2 Quad Q9650 333 x 9 = 3.00GHz (DDR3-1333) - Core 2 Quad Q9650 366 x 9 = 3.30GHz (DDR3-1467)* - Core 2 Quad Q9650 400 x 9 = 3.60GHz (DDR3-1600)* - x2 Kingston HyperX 2GB PC3-13000 Module(s) - Seagate 500GB 7200-RPM (Serial ATA300) - Diamond Radeon HD 4870 X2 (2GB) - ASUS P5E3 Premium (Intel X48) Software - Microsoft Windows Vista Ultimate SP1 (64-bit) - Intel System Driver 8.4.0.1016 - ATI Catalyst 8.8 WHQL

Hardware - Core 2 Duo E6320 - 266 x 6 = 1.60GHz (DDR3-1066)** - Core 2 Duo E6320 - 266 x 7 = 1.86GHz (DDR3-1066) - Core 2 Duo E6550 - 333 x 6 = 2.00GHz (DDR3-1333) - Core 2 Duo E6420 - 266 x 8 = 2.13GHz (DDR3-1066) - Core 2 Duo E6550 - 333 x 7 = 2.33GHz (DDR3-1333) - Core 2 Duo E6600 - 266 x 9 = 2.40GHz (DDR3-1066) - Core 2 Duo E6700 - 333 x 8 = 2.66GHz (DDR3-1333) - Core 2 Duo E6850 - 333 x 9 = 3.00GHz (DDR3-1333) - Core 2 Duo E8400 - 366 x 9 = 3.30GHz (DDR3-1467)* - Core 2 Duo E8400 - 400 x 9 = 3.60GHz (DDR3-1600)* - x2 Kingston HyperX 2GB PC3-13000 Module(s) - Seagate 500GB 7200-RPM (Serial ATA300) - Diamond Radeon HD 4870 X2 (2GB) - ASUS P5E3 Premium (Intel X48) Software - Microsoft Windows Vista Ultimate SP1 (64-bit) - Intel System Driver 8.4.0.1016 - ATI Catalyst 8.8 WHQL

Hardware - Phenom X4 9150e - 200 x 8 = 1.60GHz (DDR2-1066)** - Phenom X4 9150e - 200 x 9 = 1.80GHz (DDR2-1066) - Phenom X4 9350e - 200 x 10 = 2.00GHz (DDR2-1066) - Phenom X4 9550 - 200 x 11 = 2.20GHz (DDR2-1066) - Phenom X4 9650 - 200 x 11.5 = 2.30GHz (DDR2-1066) - Phenom X4 9750 - 200 x 12 = 2.40GHz (DDR2-1066) - Phenom X4 9850 - 200 x 12.5 = 2.50GHz (DDR2-1066) - Phenom X4 9950 - 200 x 13 = 2.60GHz (DDR2-1066) - Phenom X4 9950 - 200 x 14 = 2.80GHz (DDR2-1066)* - Phenom X4 9950 - 200 x 15 = 3.00GHz (DDR2-1066)* - x2 Kingston HyperX 2GB PC2-8500 Module(s) - Seagate 500GB 7200-RPM (Serial ATA300) - Diamond Radeon HD 4870 X2 (2GB) - ASUS M3A79-T Deluxe (790FX) Software - Microsoft Windows Vista Ultimate SP1 (64-bit) - ATI Catalyst 8.8 WHQL

Hardware - Athlon 64 X2 4000+ - 200 x 8 = 1.60GHz (DDR2-1066)** - Athlon 64 X2 4000+ - 200 x 10 = 2.00GHz (DDR2-1066) - Athlon 64 X2 4400+ - 200 x 11 = 2.20GHz (DDR2-1066) - Athlon 64 X2 4800+ - 200 x 12 = 2.40GHz (DDR2-1066) - Athlon 64 X2 5200+ - 200 x 13 = 2.60GHz (DDR2-1066) - Athlon 64 X2 5600+ - 200 x 14 = 2.80GHz (DDR2-1066) - Athlon 64 X2 6000+ - 200 x 15 = 3.00GHz (DDR2-1066) - Athlon 64 X2 6400+ - 200 x 16 = 3.20GHz (DDR2-1066) - Athlon 64 X2 6400+ - 233 x 14 = 3.26GHz (DDR2-1211)* - Athlon 64 X2 6400+ - 266 x 13 = 3.46GHz (DDR2-1383)* - x2 Kingston HyperX 2GB PC2-8500 Module(s) - Seagate 500GB 7200-RPM (Serial ATA300) - Diamond Radeon HD 4870 X2 (2GB) - ASUS M3A79-T Deluxe (790FX) Software - Microsoft Windows Vista Ultimate SP1 (64-bit) - ATI Catalyst 8.8 WHQL

The results now recorded with the Radeon HD 4870 X2 looked much more realistic giving the Core 2 Quad processors a slightly advantage over their dual-core versions just as we saw in the other four games tested. We apologize for any inconveniences the original results may have caused and I am sorry it took so long to work out the problem.

When testing Company of Heroes with the Intel Core 2 Quad range we found quite a significant difference in performance depending on the processors clock frequency. As you can see the Q9650 running at its default frequency of 3.0GHz was 9% faster than the Q6600 at 2.4GHz which worked out to be 7fps.

A similar performance trend is found when testing the Core 2 Duo processors, though this time a 12.5% margin favors the E6850 3.0GHz processor over the E6600 running at 2.4GHz! It seems that once 3.0GHz is reached the Core 2 Duo processors hit somewhat of a limit at around 80fps.

Unfortunately the Phenom X4 processors scale quite poorly in Company of Heroes and even at the same clock speed the Core 2 Duo is superior. Despite being overclocked to 3GHz the Phenom X4 9950 was limited to 72fps matching that of the Core 2 Duo E6600.

Although the Phenom X4 processors were much faster than the Athlon64 X2 processors in Company of Heroes, the margins were not nearly as significant as those seen between the Core 2 Duo and Phenom X4 processors. The Athlon64 X2 6400+ was 40% faster than the 4000+, and from the 4800+ onwards we saw a 2fps jump every time the frequency was increased by 200MHz.

When looking at all forty processor configurations combined you can see that it is the Core 2 Duo/Quad processors that dominate this test when paired with the Radeon HD 4870 X2. It seems that the higher clocked Core 2 Duo processors hit a wall at around 80fps where as the quads push through to roughtly 90fps. The Atlhon64 X2 processors are showing their age here though the overclocked 6400+ at 3.46GHz was able to defeat the Phenom X4 9850.

Those planning on playing Devil May Cry with the Radeon HD 4870 X2 need not worry about frame rates. Still those wanting maximum returns need nothing faster than a 2.33GHz Core 2 Quad processor, as the slightly under clocked Q6600 configuration was nearly as fast as the Q9650 clocked at 3.60GHz, which was impressive. Dropping the frequency to 2.13GHz did see a significant drop in performance, which leveled off around the 1.86GHz - 2.00GHz mark.

The Core 2 Quad processors clearly had an advantage over the dual-core parts, as the E6850 was 13fps slower than the Q9650, which was working at the same clock frequency. The Radeon HD 4870 X2 was not feeling its limits till we began overclocking the E8400 processor to 3.30GHz and 3.60GHz, though the performance of the E6850 at 3.0GHz was still quite impressive.

The Phenom X4 9950 was 40% faster than the Phenom X4 9150e, rendering 19fps more at 1920x1200. It seems for optimal performance the Phenom X4 9850 is required, as this processor was able push the Radeon HD 4870 X2 to the max with a clock speed of just 2.50GHz, which is impressive. Pushing the Phenom X4 to 3.0GHz allowed for just 8fps more.

The Athlon64 X2 processors continued to show their age, falling 100fps or more behind the Phenom X4 processors. Despite being overclocked to 3.46GHz, the Athlon64 X2 6400+ was only able to average 169fps, while the under clocked Phenom X4 9150e at just 1.60GHz managed 189fps.

The high-end Intel Core 2 Duo and Core 2 Quad, along with the AMD Phenom X4 processors, are able to push the Radeon HD 4870 X2 to the limit in Devil May Cry 4 without requiring any overclocking. It is interesting to note that while the flagship processors from these series rendered well over 240fps on average, the 3.46GHz overclocked Athlon64 X2 6400+ managed just 169fps. The Core 2 Quad Q9650 was the fastest test CPU using its default settings, followed by the Phenom X4 9950, and then the Q6600.

There was a 12% performance difference when going from the Q6600 to the Q9650 in Enemy Territory Quake Wars. Every time the frequency gains a few hundred MHz the performance is increased by roughly 10fps up until 3.00GHz is reached. After that every few hundred MHz accounts for just a few frames per second until we reached 140fps at 3.60GHz! Therefore it would seem, in order to get the most out of a Radeon HD 4870 X2 in Enemy Territory, a 3.0GHz Core 2 Quad processor will be ideal.

Again the quad-core processors seem to provide reasonable performance gains in Enemy Territory Quake Wars, as the 2.4GHz Q6600 is easily able to outperform the 3.0GHz E8650. Even at 3.60GHz the E8650 struggled to max out the Radeon HD 4870 X2, though at this frequency it did beat the fastest Phenom X4 configuration.

Again the Phenom X4 processors did not scale too poorly in Enemy Territory Quake Wars, despite losing once again to the Core 2 Duo processors. Strong performance gains are seen when going from 1.80GHz to 2.30GHz, while just a few frames were gained each step when going from 2.30GHz to 2.60GHz. Overclocking the Phenom X4 9950 did allow us to squeeze a little more performance out of the Radeon HD 4870 X2, though with 121fps at 3.0GHz the Phenom X4 was only able to just outperform the Q6600 at 2.4GHz!

Once again the Athlon64 X2 processors continue to show their age, as the 6400+ is significantly slower than the under clocked Phenom X4 9150e. Only at 3.46GHz is the Athlon64 X2 6400+ able to edge out the crippled 9150e by a small margin.

The Radeon HD 4870 X2 performance continued to improve as we threw more CPU power at it, with the fastest configuration being the Core 2 Quad Q9650 at 3.60GHz, which delivered 140fps. The Q9650 was also the fastest processor tested at its default frequency, rendering 133fps, followed by the Q6600 with 119fps. The Phenom X4 9950 averaged 115fps, and the 9850 managed 113fps.The Athlon64 X2 6400+ clocked at 3.46GHz averaged just 84fps, matching the performance of the Core 2 Duo E6550 at just 2.0GHz!

Supreme Commander is a game that I have personally spent quite a bit of time playing, and from my experiences I can say that this is a game that really benefits from as much processing power as you can throw at it! Ohh, and a Radeon HD 4870 X2 never hurts performance either. As you can see, overclocking the Q9650 past 3.0GHz allowed us to crack the 80fps barrier, which allowed for very smooth game play regardless of what was taking place. The Q6600 averaged 71fps, while the under clocked 1.60GHz configuration managed just 59fps.

The performance difference between the Core 2 Quad and Core 2 Duo processors is far less significant in Supreme Commander. The E6850 is able to reach 76fps without any overclocking which was impressive, while at 3.60GHz it did manage 81fps. Again it seems that the more CPU power you throw at the Radeon HD 4870 X2 in games such as Supreme Commander the better.

The Phenom X4 9950 was 8fps slower than the Core 2 Duo E8650, while the Core 2 Duo processors still remained faster on a clock for clock basis. That said, the Phenom X4 processors did perform reasonably well when testing with Supreme Commander, though the drop in frame rate does render the Radeon HD 4870 X2 somewhat useless on this platform.

Clock for clock the Athlon64 X2 processors were another 10fps slower than the Phenom X4 processors, meaning that once again they really were showing their age. While we do not expect that anyone would pair a modern day $600 US+ graphics card with an out dated processor, it is interesting to see how things have progressed in the gaming world over the past few years.

Although the difference between the overclocked Athlon64 X2 6400+ and the Core 2 Quad Q9650 processors does not seem all that significant in Supreme Commander, it is. The overclocked Core 2 Quad Q9650 was able to deliver 36% more performance at 1920x1200 using maximum in game quality settings combined with 8xAA/16xAF. Although a number of processors did produce the same result, the performance still improved with the clock speed for the most part.

The under clocked Core 2 Quad Q6600 at 1.60GHz is still able to average 100fps, while at its default clock frequency of 2.40GHz an average of 125fps was achieved. However as the processor frequency increased so too did the average frame rate, and the Q9650 averaged 132fps. Increasing the clock frequency to 3.60GHz allowed for just 139fps, so again it would seem that a 3.0GHz Core 2 Quad is again ample for the most part.

Once again we find that the quad-core processors do provide a significant performance advantage in Unreal Tournament 3. Although in the end they are able to push the Radeon HD 4870 X2 to similar heights at lower clock frequencies, the dual-core processors really suffer. The Core 2 Duo processors show no signs of maxing out the Radeon HD 4870 X2, as the average frame rate constantly increases with the processors clock speed.

Interesting despite being overclocked to 3.0GHz, the Phenom X4 9950 was unable to outperform the Core 2 Quad Q9650 at just 2.33GHz. In fact, at its default clock frequency, the Phenom X4 9950 was only able to match the performance of the Q9650 at just 2.0GHz! Once the Phenom X4 clock frequency dropped below 2.20GHz the Unreal Tournament 3 performance dropped significantly.

The Athlon64 X2 processors struggled to perform in Unreal Tournament 3 as they failed to push the Radeon HD 4870 X2 anywhere near its limits. In fact, the overclocked Athlon64 X2 6400+ processor at 3.46GHz was only just able to outperform the under clocked Phenom X4 9150e.

Here we can see again that the Athlon64 X2 is now really starting to show its age, that said you won’t see any Pentium 4 results in this article. The once powerful Athlon64 X2 4000+ produced just 59fps on average, which is barley within playable limits. Overclocking the 6400+ to 3.46GHz allowed for 79fps, which while playable, makes this overclocked processor 20% slower than the Core 2 Duo E6600, which is an old processor itself.

The quad-core processors were the fastest tested, with the standard Core 2 Quad Q9650 rendering an average of 132fps, followed by the Q6600 with 125fps. The Phenom X4 processors did shy off a bit in Unreal Tournament 3, with the 9950 managing just 113fps making it slower than the dual-core E6850 which averaged 118fps. Slowly but surely the frame rates continued to rise each time the processor became a little bit faster, and it looks as though we never saw the limits of the Radeon HD 4870 X2 in this game.

Generally the performance margin between the Core 2 Duo and Phenom X4 processors was quite minimal, though more often than not the Intel dual-core processors did outperform the AMD quad-core parts. This is neither new nor surprising, as we have seen this time and time again in the past.

Although it was not all that surprising to see just how much the Athlon64 X2 architecture has aged since we last looked at it, we were still somewhat shocked at how slow the 3.46GHz 6400+ processor was. Obviously we do not expect that AMD Athlon64 X2 owners are going to race out and purchase a Radeon HD 4870 X2, but we still felt that including this processor series would be useful. Those still using an Athlon64 X2 should seriously consider upgrading their platform, even if they only plan to purchase a single current generation AMD or Nvidia graphics card.

Out of the five games used for testing, Devil May Cry was the only one that allowed the processors to find the limits of the Radeon HD 4870 X2. It would seem that this game is more GPU dependent than the others, as the Phenom X4 9850/9950 and Core 2 Quad Q6600/9650 processors were able to max out the graphics card. While the Core 2 Duo processors also did very well in this game, the overclocked 3.46GHz Athlon64 X2 6400+ processor was slower than the E6320.

Another game that strongly favored the quad-core processors was Enemy Territory Quake Wars. Here the Core 2 Quad Q9650 managed 133fps, while the Q6600 averaged 119fps. The Phenom X4 processors were next in line, with the 9950 averaging 115fps, and the 9850 reached 113fps. In fact the Phenom X4 9750 and 9650 processors were also able to outperform the E6850, which only narrowly edged out the 9550.

The Supreme Commander results were more like what we were expecting. Here the Core 2 Quad processors were only slightly faster than the dual-core versions, while the Phenom X4 processors were able to match the Core 2 Duo’s for the most part. However, Supreme Commander was another game that continually saw the Radeon HD 4870 X2 become faster, as the processors became more powerful and we never really saw the limits.

Finally, Unreal Tournament 3 painted an interesting picture, as there was a significant difference in performance between the fastest and the slowest processors tested. Once again the Core 2 Quad Q9650 was the king here, delivering 132fps at its default clock frequency of 3.0GHz, which was impressive when compared to the E6850 for example, which managed just 118fps at the same frequency. The best Phenom X4 processor was of course the 9950, which averaged 113fps, edging out the old E6700 by 2fps.

Again the Radeon HD 4870 X2 limits were not seen, as the results continued to improve with the processors clock speed. While we always suspected that the world’s fastest graphics card would require the world’s fastest processor, we know now for certain that it does. Those with lower-end Core 2 Duo/Quad processors can squeeze a great deal of performance out of the Radeon HD 4870 X2 simply by overclocking their processor past 3.0GHz, which is well within the limits of these processors.

Due to a lack of raw clock speed, the Phenom X4 processors did fall short in most tests, and even at 3.0GHz the 9950 struggled with the competition. Given that all tests were conducted at 1920x1200 with 8xAA/16xAF enabled, the results were impressive, and while the Phenom X4 processors did deliver acceptable performance, processors such as the Core 2 Quad Q9650 proved that the Radeon HD 4870 X2 still has much more potential just waiting to be unleashed.

Reviewed By Steven Walton