Date: October 7, 2008 - Author: Rob Williams
It's no secret that the HD 4870 is one of the best overall GPUs on the market right now, but with so much selection from vendors, it's hard to choose the "best" one. Palit has a definite winner with their Sonic Dual Edition though. It's pre-overclocked, runs 20°C cooler than the reference design and carries no cost premium.
Just a month ago, I posted our review of Palit's Radeon HD 4870, and since then, not too much has changed in the GPU scheme of things. AMD released their Radeon HD 4670 and HD 4650 budget cards, but for gamers after great performance, the choices have remained the same since the beginning of summer. There's nothing wrong with that though, as the releases we saw in June from both ATI and NVIDIA are still amazing and I'll say it once again, there's no better time to be thinking about a GPU purchase than right now.
Another interesting event over the summer was to see the tables turn in the market. Even though NVIDIA released their impressive GTX 260 and GTX 280 cards, they were quickly overshadowed by ATI's launch of the HD 4000-series, which delivered performance no one was expecting. Fast forward to the HD 4870 X2, and it's really difficult to look away from any of AMD's offerings. As it stands right now, their current line-up is priced-right and offers insane performance.
When I get asked about the "best" mid-range GPU of the moment, I have no other choice but to say the HD 4870. It's affordable, at just under $300, yet delivers enough performance to power any game on the market at 2560x1600 with good detail settings - even Crysis Warhead. Should you decide down the road that you want even more performance, adding a second identical card to your system is all you need. By the time you feel the need to upgrade, prices should be far lower than they are right now.
Now more than ever, we're seeing graphic card vendors working hard to draw your attention to their offerings, instead of a competitors. Rather than just slap a GPU along with a sprinkle of memory chips on a PCB, they are looking to offer extra features that set their product apart. In the case of the Radeon HD 4870 Sonic from Palit, the extra features include three-phase power, a dual-fan cooling solution, higher clocks and enough connections to keep your monitor stuffed.
As it is right now, the HD 4870 is ATI's second-highest card in their line-up, succeeded only by the HD 4780 X2, which in case you are unaware, is simply two HD 4870's put together. Unlike NVIDIA's GX2 solution, the ATI's X2's stick to one PCB, thankfully. The next best option to the HD 4870 from ATI is of course the HD 4850. That card is similar in most regards except where clocks are concerned. The HD 4870 also features GDDR5, while the HD 4850 sticks to GDDR3.
Model |
Core MHz |
Mem MHz |
Memory |
Bus Width |
Stream Proc. |
| HD 4870 X2 | 750 |
900 |
2x1024MB |
256-bit |
800 |
| HD 4850 X2 | 625 |
993 |
2x1024MB |
256-bit |
800 |
| HD 4870 | 750 |
900 |
512MB |
256-bit |
800 |
| HD 4850 | 625 |
993 |
512MB |
256-bit |
800 |
| HD 4670 | 750 |
900 |
512MB |
128-bit |
320 |
| HD 4650 | 600 |
500 |
512MB |
128-bit |
320 |
| HD 3870 X2 | 825 |
900 |
2x512MB |
256-bit |
320 |
| HD 3850 X2 | 666 |
828 |
2x512MB |
256-bit |
320 |
| HD 3870 | 775 |
900 |
512MB |
256-bit |
320 |
| HD 3850 | 666 |
828 |
512MB |
256-bit |
320 |
| HD 3650 | 725 |
800 |
256,512MB |
128-bit |
120 |
| HD 3450 | 600 |
500 |
256MB |
64-bit |
40 |
Palit's Sonic "Dual Edition" card is named as such because it includes two fans with steeply-curved blades, completely unlike the leaf-blower fan that the reference design includes. The semi-translucent fans are rather quiet, although I don't have a sound meter here to test with. It's much quieter than the reference design fans, however, especially at full load.
Palit also makes improvements with the card itself, using a multi-layer PCB that improves upon the reference, while also utilizing a three-phase power solution - whereas the stock design uses a two-phase design. I am not immediately impressed by the physical appearance of the cooler, and the red is a little gaudy , but at least it matches the both the brand and PCB color.
This card undoubtedly includes most of the connectivity that most people will ever need. Included on the card itself are two DVI-D ports along with a DisplayPort - nice to see given it's a technology that's still catching on. Palit includes VGA and HDMI support by including the appropriate dongles.
Another special feature of the card is its Dual-BIOS design, which opens up the ability to overclock your card on a dime (with the computer off, preferably). With the switch in its original position, the card will act as a normal-clocked HD 4870 (with slightly overclocked memory), but flick it to the Turbo position and the Core clock gets bumped by 25MHz over stock and the memory by 100MHz. A healthy boost, but not jaw-dropping.
To help understand how the special cooler operates, we can look at the side-view photo below.
The card utilizes an array of aluminum fins like most others on the market, but with the help of the dual fans, more of the card is covered and heat gets exhausted out towards the back quickly. The cooler doesn't touch the memory chips, but as we'll find out later, it doesn't matter - the card keeps cool enough even with this design.
To use the card, your PSU should include two PCI-E 6-pin cables, or four if you plan to pick up two for Crossfire mode. With the current prices of the HD 4870 though, the small premium over dual cards would be put to good use on the X2, since it's a single card design and only requires two power connections (8+6 Pin).
With our look at Palit's latest offering taken care of, we highly recommend you peruse our testing methodology on the following page. We recently overhauled our entire test system and also how we conduct testing, so even if you've read one of our GPUs recently, we still recommend looking through.
At Techgage, we strive to make sure our results are as accurate as possible. Our testing is rigorous and time-consuming, but we feel the effort is worth it. In an attempt to leave no question unanswered, this page contains not only our testbed specifications, but also a fully-detailed look at how we conduct our testing.
If there is a bit of information that we've omitted, or you wish to offer thoughts or suggest changes, please feel free to shoot us an e-mail or post in our forums.
The below table lists our testing machine's hardware, which remains unchanged throughout all GPU testing, minus the graphics card. Each card used for comparison is also listed here, along with the driver version used. Each one of the URLs in this table can be clicked to view the respective review of that product, or if a review doesn't exist, it will bring you to the product on the manufacturer's website.
Component |
Model |
| Processor | Intel Core 2 Extreme QX9770 - Quad-Core, 3.6GHz (Overclocked), 1.35v |
| Motherboard | ASUS Rampage Extreme - X48-based, 0501 BIOS (08/28/08) |
| Memory | Corsair XMS3 DHX 2x2GB - DDR3-1333 7-7-7-15-1T, 1.91v |
| ATI Graphics |
Palit Radeon HD 4870 X2 2GB (Catalyst 8.9) Palit HD 4870 512MB Sonic Dual Edition (Catalyst 8.9) Palit HD 4870 512MB (Catalyst 8.9) ASUS EAH4850 512MB (Catalyst 8.9) |
| NVIDIA Graphics |
Palit GTX 280 1GB (GeForce 178.13) Palit 9800 GX2 1GB (GeForce 178.13) ASUS EN9800GTX 512MB (GeForce 178.13) Gigabyte 9600 GT 512MB (GeForce 178.13) |
| Audio | On-Board Audio |
| Storage | |
| Power Supply | |
| Chassis | |
| Display | |
| Cooling | |
| Et cetera |
When preparing our testbeds for any type of performance testing, we follow these guidelines:
To aide with the goal of keeping accurate and repeatable results, we alter certain services in Windows Vista from starting up at boot. This is due to the fact that these services have the tendency to start up in the background without notice, potentially causing slightly inaccurate results. Disabling "Windows Search" turns off the OS' indexing which can at times utilize the hard drive and memory more than we'd like.
For graphic card reviews that pit us with a mid-range card or higher, we test at three popular resolutions that span the mid-range to high-end ground, consisting of monitor sizes of 20" (1680x1050), 24" (1920x1200) and 30" (2560x1600).
In an attempt to offer "real-world" results, we do not utilize timedemos in our graphic card reviews, with the exception of Futuremark's automated 3DMark Vantage. Each game in our test suite is benchmarked manually, with the minimum and average frames-per-second (FPS) captured with the help of FRAPS 2.9.5.
To deliver the best overall results, each title we use is exhaustively explored in order to find the best possible level in terms of intensiveness and replayability. Once a level is chosen, we play through repeatedly to find the best possible route and then in our official benchmarking, we stick to that route as close as possible. Since we are not robots and the game can throw in minor twists with each run, no run can be identical to the pixel.
Each game and setting combination is tested twice, and if there is a discrepancy between the initial results, the testing is repeated until we see results we are confident with.
The six games we currently use for our GPU reviews are listed below, with direct screenshots of the game's setting screens and explanations of why we chose what we did.
1680x1050 |
1920x1200 |
2560x1600 |
Crysis and its Warhead successor are two of the most demanding games on the PC today, and as a result, Anti-Aliasing is not our focus in testing. The noticeable differences come when the advanced options are increased, and to keep things simple, Crytek offers four profiles to choose from: Entry, Mainstream, Gamer and Enthusiast.
We run all three of our resolutions at the "Gamer" setting as it's playable enough on all current mid-range (and higher) GPUs to complete a full run-through. The game unquestionably runs better on "Mainstream", but "Gamer" helps push even the highest-end GPUs to their breaking-point.
1680x1050 |
1920x1200 |
2560x1600 |
The Call of Duty series of war-shooters are without question some of the most gorgeous on the PC (and consoles), but what's great is the fact that the games are also highly optimized, so no one has to max out their machine's specs in order to play it. Since that's the case, the in-game options are maxed out in all regards, except the Anisotropic Filtering, which is set to the center of the slider bar.
1680x1050 |
1920x1200 |
2560x1600 |
It might have been four-years-ago that we were able to play the first installment of the Half-Life 2 series, but it's held up well with its new releases and engine upgrades. This is one title that thrives on both a fast CPU and GPU, and though it's demanding at times, most any recent computer should be able to play the game with close to maxed-out detail settings, aside from the Anti-Aliasing.
In the case of very-recent mid-range cards, the game will run fine all the way up to 2560x1600 with maxed-out detail, minus Anti-Aliasing. All of our tested resolutions use identical settings, with 4xAA and 8xAF.
1680x1050 |
1920x1200 |
2560x1600 |
For as long as the Unreal Tournament series has been around, people have been benchmarking it. So it's a little strange that UT III offers some of the most simplistic in-game graphics settings ever, with the "Texture Detail" and "World Detail" being of most importance. These two options scale between 1 and 5, and we of course use 5 as it can be used on almost any current mid-range GPU as long as the resolution is kept in check.
To help make both the game look better and our test more demanding, we force an Anti-Aliasing setting with the help of either ATI's or NVIDIA's Control Center. Both allow us to force a 4xAA setting, which is where we keep it.
1680x1050 |
1920x1200 |
2560x1600 |
The original S.T.A.L.K.E.R. was one of the most demanding games we've ever tested with, and its Clear Sky sequel is no different. Unlike most other games, S.T.A.L.K.E.R. not only offers quality setting profiles (Medium, High, etc) but also render settings. New to Clear Sky though is the "Enhanced Full Dynamic Lighting DX10" render setting, which enables all of the DX10 goodies that fans were waiting for.
We test the game at this setting for two reasons. First, it's incredibly demanding on even the highest-end GPU, and second, FRAPS will not properly show the FPS counter or record an average FPS with any other render setting - I'm unsure why.
1680x1050 |
1920x1200 |
The Need for Speed series has been a personal favorite ever since I first laid my hands on the third iteration, and I cannot see myself getting bored with any of the games soon. Sadly, Electronic Arts does not allow a 2560x1600 resolution with ProStreet, so we are limited to testing at both 1680x1050 and 1920x1200 only. In-game settings are completely maxed out, with 4xAA and V-Sync Off.
The next release from the NFS series will be Undercover, which is due out in late October. We will more than likely replace ProStreet soon afterwards, and we can only hope that EA learned their lesson from ProStreet and opens up 2560x1600 support.
As PC enthusiasts, we tend to be drawn to games that offer spectacular graphics... titles that help reaffirm your belief that shelling out lots of cash for that high-end monitor and PC was well worth it. But it's rare when a game comes along that is so visually-demanding, it's unable to run fully maxed out on even the highest-end systems on the market. In the case of the original Crysis, it's easy to see that's what Crytek was going for.
Funny enough, even though Crysis was released close to a year ago, the game today still has difficulty running at 2560x1600 with full detail settings - and that's even with overlooking the use of anti-aliasing! Luckily, Warhead is better optimized and will run smoother on almost any GPU, despite looking just as gorgeous as its predecessor, as you can see in the screenshot below.
The game includes four basic profiles to help you adjust the settings based on how good your system is. These include Entry, Mainstream, Gamer and Enthusiast - the latter of which is for the biggest of systems out there, unless you have a sweet graphics card and are only running 1680x1050. We run our tests at the Gamer setting as it's very demanding on any current GPU and is a proper baseline of the level of detail that hardcore gamers would demand from the game.



Palit's overclocked HD 4870 has no trouble beating out the stock-clocked version, although the differences are minor. For Crysis to be enjoyable, the game should deliver at least 30 FPS, which we came very close to hitting at 1920x1200 with this card.
Graphics Card |
Best Playable |
Avg. FPS |
Palit HD 4870 X2 2GB |
2560x1600, Gamer, 0xAA |
31.382 FPS |
Palit 9800 GX2 1GB |
2560x1600, Mainstream, 0xAA |
50.550 FPS |
Palit GTX 280 1GB |
2560x1600, Mainstream, 0xAA |
46.038 FPS |
Palit HD 4870 512MB Sonic |
2560x1600, Mainstream, 0xAA |
34.471 FPS |
Palit HD 4870 512MB |
2560x1600, Mainstream, 0xAA |
32.973 FPS |
ASUS 9800 GTX 512MB |
2560x1600, Mainstream, 0xAA |
30.840 FPS |
ASUS HD 4850 512MB |
2560x1600, Mainstream, 0xAA |
26.530 FPS |
Gigabyte 9600 GT 512MB |
1920x1200, Mainstream, 0xAA |
31.979 FPS |
Although the Gamer profile delivers some sharp looking graphics, shifting to Mainstream doubled the output of our HD 4870 Sonic and allowed us to run the game at 2560x1600 with a comfortable 34 FPS on average The only GPU to date that's capable of running that resolution along with the Gamer setting is ATI's HD 4870 X2. Using two HD 4870s in Crossfire mode would offer similar results.
Crysis Warhead might have the ability to bring any system to its knees even with what we consider to be reasonable settings, but Call of Duty 4 manages to look great regardless of your hardware, as long as it's reasonably current. It's also one of the few games on the market that will actually benefit from having a multi-core processor, although Quad-Cores offer no performance gain over a Dual-Core of the same frequency.
For our testing, we use a level called The Bog. The reason is simple... it looks great, plays well and happens to be incredibly demanding on the system. It takes place at night, but there is more gunfire, explosions, smoke, specular lighting and flying corpses than you can shake an assault rifle at.
Because the game runs well on all current mid-range GPUs at reasonable graphic settings, we max out what's available to us, which includes enabling 4xAA and 8xAF, along with choosing the highest available options for everything else.



Once again, Palit's Sonic card outpaced the stock-clocked model with ease, although unlike Crysis, the differences here actually reach 5 FPS at certain resolutions. For the price, both of the HD 4870's deliver fantastic frame rates.
Graphics Card |
Best Playable |
Avg. FPS |
Palit HD 4870 X2 2GB |
2560x1600, Max Detail, 8xAA |
113.024 FPS |
Palit GTX 280 1GB |
2560x1600, Max Detail, 8xAA |
85.440 FPS |
Palit 9800 GX2 1GB |
2560x1600, Max Detail, 4xAA |
76.192 FPS |
Palit HD 4870 512MB Sonic |
2560x1600, Max Detail, 4xAA |
67.928 FPS |
Palit HD 4870 512MB |
2560x1600, Max Detail, 4xAA |
64.825 FPS |
ASUS 9800 GTX 512MB |
2560x1600, Max Detail, 0xAA |
70.363 FPS |
ASUS HD 4850 512MB |
2560x1600, Max Detail, 0xAA |
69.745 FPS |
Gigabyte 9600 GT 512MB |
2560x1600, Max Detail, 0xAA |
48.180 FPS |
The setting we found to be most playable is the same one used in our 2560x1600 test. Even with the in-game settings maxed out, the card delivers close to 70 FPS, which is more than enough to offer the gameplay experience we're all looking for.
The original Half-Life 2 might have first seen the light of day close to four years ago, but it's still arguably one of the greatest-looking games ever seen on the PC. Follow-up versions, including Episode One and Episode Two, do well to put the Source Engine upgrades to full use. While playing, it's hard to believe that the game is based on a four+ year old engine, but it still looks great and runs well on almost any GPU purchased over the past few years.
Like Call of Duty 4, Half-Life 2: Episode Two runs well on modest hardware, but a recent mid-range graphics card is recommended if you wish to play at higher than 1680x1050 or would like to top out the available options, including anti-aliasing and very high texture settings.
This game benefits from both the CPU and GPU, and the skies the limit. In order to fully top out the available settings and run the highest resolution possible, you need a very fast GPU or GPUs along with a fast processor. Though the in-game options go much higher, we run our tests with 4xAA and 8xAF to allow the game to remain playable on the smaller mid-range cards.



The card continues to perform well, and the pre-overclock helped deliver an extra 4 FPS at our top resolution. Nothing major, but it's an increase nonetheless.
Graphics Card |
Best Playable |
Avg. FPS |
Palit HD 4870 X2 2GB |
2560x1600, Max Detail, 8xAA, 16xAF |
81.418 FPS |
Palit GTX 280 1GB |
2560x1600, Max Detail, 8xAA, 16xAF |
61.437 FPS |
Palit HD 4870 512MB Sonic |
2560x1600, Max Detail, 8xAA, 16xAF |
60.413 FPS |
Palit HD 4870 512MB |
2560x1600, Max Detail, 8xAA, 16xAF |
56.572 FPS |
Palit 9800 GX2 1GB |
2560x1600, Max Detail, 4xAA, 8xAF |
89.596 FPS |
ASUS 9800 GTX 512MB |
2560x1600, Max Detail, 4xAA, 8xAF |
51.272 FPS |
ASUS HD 4850 512MB |
2560x1600, Max Detail, 4xAA, 8xAF |
48.142 FPS |
Gigabyte 9600 GT 512MB |
2560x1600, Max Detail, 0xAA, 8xAF |
52.297 FPS |
Our general rule is that for this particular game to be completely enjoyable, an average of at least 60 FPS is required. When we increased the anti-aliasing to 8xAA, we managed to keep performance above that mark, making it our best playable setting. For an affordable GPU like this one, it's great to see performance like this at such high detail settings.
As odd as it may seem, every single game we currently use for our graphic card benchmarking is a sequel or an entry in a series of games, including this one. The original Unreal Tournament launched in late 1999, and since then, it has become a stature with GPU benchmarking. Similar to Call of Duty, the UT series of games is one that manages to deliver spectacular graphics, but doesn't require a bleeding-edge machine to see them.
UTIII offers a variety of modes and levels, and has some of the most interesting and lush environments ever seen in a video game. If I could choose where I wanted to die, it would most likely be in the Gateway level, which you can see in the screenshot below. This level is one of the most interesting in the game as it's essentially three levels in one, linked together with portals - and it's hard to beat the feeling of scoring a portal frag.
The game might be one of the best-looking currently on the PC, but it doesn't offer robust in-game settings like some others in our suite. Because of this, we are forced to enable anti-aliasing in the control panel of the current graphics card. Both ATI's and NVIDIA's drivers allow us to choose 4xAA, so that's what we stick with throughout all of our testing.



The tradition continues. Palit's pre-overclocked HD 4870 consistently outperforms the stock-clocked model, which isn't at all a surprise. To play in the big leagues though (2560x1600, 4xAA), a beefier GPU is almost a requirement, especially if you play online (it would be odd if you didn't).
Graphics Card |
Best Playable |
Avg. FPS |
Palit GTX 280 1GB |
2560x1600, Max Detail, 4xAA |
72.148 FPS |
Palit HD 4870 X2 2GB |
2560x1600, Max Detail, 4xAA |
55.479 FPS |
Palit 9800 GX2 1GB |
2560x1600, Max Detail, 0xAA |
78.909 FPS |
Palit HD 4870 512MB Sonic |
2560x1600, Max Detail, 0xAA |
60.901 FPS |
Palit HD 4870 512MB |
2560x1600, Max Detail, 0xAA |
57.617 FPS |
ASUS 9800 GTX 512MB |
2560x1600, Max Detail, 0xAA |
48.874 FPS |
Gigabyte 9600 GT 512MB |
2560x1600, Max Detail, 0xAA |
43.781 FPS |
ASUS HD 4850 512MB |
2560x1600, Max Detail, 0xAA |
42.228 FPS |
Like most of the GPUs in our line-up, enabling the 4xAA is a bad move with this card. Only NVIDIA's GTX 280 and ATI's HD 4870 X2 manage to play well with anti-aliasing at our highest resolution, but luckily enough, the game still looks fantastic without it.
When it comes to first-person shooters, post-apocalyptic adventures are a dime a dozen. But when S.T.A.L.K.E.R. was first released in the spring of 2007, it dared to be different. How? By basing the game off of a real-world tragedy, the Chernobyl nuclear disaster, which occurred way back in 1986 near the city of Prypiat in the Ukraine. Despite the disaster happening so long ago, people are still unable to live in the surrounding area, and will be unable to for at least another 150 years.
In addition to the games real-world ties, S.T.A.L.K.E.R. happened to be one of the grittiest, realistic (aside from the problematic AI) and expansive games we've seen on the PC in a while. Having the ability to roam as you like is a huge benefit and really helped make the game feel real. Clear Sky further delivers on what made the original so great, but at the same time, adds support for DX10.
It might be difficult to judge from the screenshot, but Clear Sky (like the original) is one of the most demanding games on the PC today, especially if you wish to play using DX10. To help push all of our GPUs to their breaking-point, we stick to that mode while using the "High" quality setting.



Clearly, the DX10 mode is one that requires an incredible GPU, regardless of what resolution you are running. Even at the modest 1680x1050, the card could only muster 26 FPS, while the X2 version upped it to 49 FPS, which was still not silky-smooth. I look forward to the day when our GPUs can run the game in DX10 with ease. It might take a little while.
Graphics Card |
Best Playable |
Palit HD 4870 X2 2GB |
2560x1600, Enhanced Full Dynamic Lighting, Medium |
Palit GTX 280 1GB |
2560x1600, Enhanced Full Dynamic Lighting, Medium |
Palit 9800 GX2 1GB |
2560x1600, Full Dynamic Lighting, High |
Palit HD 4870 512MB Sonic |
2560x1600, Full Dynamic Lighting, High |
Palit HD 4870 512MB |
2560x1600, Full Dynamic Lighting, High |
ASUS HD 4850 512MB |
2560x1600, Full Dynamic Lighting, Medium |
ASUS 9800 GTX 512MB |
1920x1200, Full Dynamic Lighting, High |
Gigabyte 9600 GT 512MB |
1920x1200, Full Dynamic Lighting, Medium |
Because FRAPS doesn't record the FPS with DX10 mode disabled (I'm unsure why), we don't have performance information for our "best playable" settings. That's fine, as the FPS doesn't tell half the story anyway. For fluid gameplay, we had to drop the DX10 mode down to Full Dynamic Lighting, but were able to keep the detail settings at High.
Where the racing genre is concerned, there are few games like Need for Speed. The first title launched in 1994, and since then, the series has done well to stick to its roots by offering an exciting racing experience that doesn't hinge on being a simulator, like Gran Turismo or Forza. Instead, it delivers close to an arcade-like experience, which seems to be preferred by most people. EA has also kept incredibly regular with the series, having released sixteen different versions in a fourteen year span. That's impressive.
What wasn't impressive was ProStreet, however, as it took the franchise and turned it upside down. Sometimes reinventing a series is a good thing, but with concern to this game, EA should have left things as they were. The developers realized they goofed though, and the upcoming Undercover game (slated for a Nov. 17 release) looks to bring the series back on track. On release, we'll replace ProStreet with Undercover in our testing.
ProStreet offers a wide-range of graphics options, allowing you to intricately tweak the game to work on your machine, regardless of what hardware you have. However, even when using maxed out detail settings, the game is still playable enough to complete a reliable benchmarking run, so we take that route. We also enable anisotropic filtering and 4x anti-aliasing.


This game in particular seems to favor NVIDIA-based GPUs, but the HD 4870 still manages to have a good showing, being out-performed only by the GTX 280 and 9800 GX2. For some unknown reason, I did not have good luck with re-benchmarking the game using an HD 4870 X2, as you can see in the graph. This might not be entirely representative of real-world gameplay, as when that card was first reviewed, I saw far better performance results. Whether this is the fault of the card or the game, I'm unsure.
Graphics Card |
Best Playable |
Avg. FPS |
Palit 9800 GX2 1GB |
1920x1200 Max Detail, 4xAA |
111.112 FPS |
Palit GTX 280 1GB |
1920x1200 Max Detail, 4xAA |
93.939 FPS |
Palit HD 4870 512MB Sonic |
1920x1200 Max Detail, 4xAA |
84.196 FPS |
Palit HD 4870 512MB |
1920x1200 Max Detail, 4xAA |
81.253 FPS |
ASUS 9800 GTX 512MB |
1920x1200 Max Detail, 4xAA |
66.830 FPS |
ASUS HD 4850 512MB |
1920x1200 Max Detail, 4xAA |
64.861 FPS |
Gigabyte 9600 GT 512MB |
1920x1200 Max Detail, 4xAA |
52.189 FPS |
Since ProStreet tops out at 1920x1200, which happens to be far less intensive than 2560x1600, all GPUs in our lineup could run the game fine with our topped-out settings. Our HD 4870's performed almost as well as the GTX 280, which is good to see given the obvious pricing differences.
Although we generally shun automated gaming benchmarks, we do like to run at least one to see how our GPUs scale when used in a 'timedemo'-type scenario. Futuremark's 3DMark Vantage is without question the best such test on the market, and it's a joy to use, and watch. The folks at Futuremark are experts in what they do, and they really know how to push that hardware of yours to its limit.
The company first started out as MadOnion and released a GPU-benchmarking tool called XLR8R, which was soon replaced with 3DMark 99. Since that time, we've seen seven different versions of the software, including two major updates (3DMark 99 Max, 3DMark 2001 SE). With each new release, the graphics get better, the capabilities get better and the sudden hit of ambition to get down and dirty with overclocking comes at you fast.
Similar to a real game, 3DMark Vantage offers many configuration options, although many (including us) prefer to stick to the profiles which include Performance, High and Extreme. Depending on which one you choose, the graphic options are tweaked accordingly, as well as the resolution. As you'd expect, the better the profile, the more intensive the test.
Performance is the stock mode that most use when benchmarking, but it only uses a resolution of 1280x1024, which isn't representative of today's gamers. Extreme is more appropriate, as it runs at 1920x1200 and does well to push any single or multi-GPU configuration currently on the market - and will do so for some time to come.



No surprises here - the card scales just as we'd expect. Palit's Sonic does leap a nice distance past the stock-clocked version, but it's certainly nothing major. What is major is the X2 model which completely dominates everything else, including NVIDIA's highest-end GTX 280.
Before tackling our overclocking results, let's first clear up what we consider to be a real overclock and how we go about achieving it. If you've read our processor reviews, you might already be aware that I personally don't care for an unstable overclock. It might look good on paper, but if it's not stable, then it won't be used. Very few people purchase a new GPU for the sole purpose of finding the maximum overclock, which is why we focus on finding what's stable and usable.
To help find the max stable overclock on an ATI card, we stick to the Catalyst Control Center included with the official driver. Sadly, ATI's limits are real conservative, so it's rare when at least the Core clock can't be totally maxed out, as long as temperatures are kept within check.
Once we find what we feel could be a stable overclock, the card is put through the stress of dealing with 3DMark Vantage's "Extreme" test, looped three times. Although previous versions of 3DMark offered the ability to loop the test infinitely, Vantage for some reason doesn't. It's too bad, as it would be the ideal GPU-stress test.
If no artifacts or performance issues arise, we continue to test the card in multiple games from our test suite, at their maximum available resolutions and settings that the card is capable of handling. If no issues arise during our real-world gameplay, we can consider the overclock to be stable and then proceed with testing.
The default clocks for the Radeon HD 4870 are 750MHz Core and 900MHz Memory. Palit's Sonic card, with the help of the toggle, boosts those figures up to 775MHz Core and 1000MHz Memory, which is minor, but an increase nonetheless. ATI's Catalyst Control Center, sadly, doesn't allow much more breathing room than that, and I'm sure it's to do with the fact that the cards normally have very high temps.
Thanks to these low limits, this is one of those times when the sliders could be maxed out and the card still pass the driver's stability test. So, our final stable overclock is 790MHz Core and 1100MHz Memory. How much benefit does such a small overclock deliver? Let's let the graph explain:

The differences are minimal, so overclocking the card at all is almost moot. There is little sense in pushing the GPU further out-of-spec when the differences are so minor, you'd never know the difference in real-world gameplay. If ATI's control center allowed a higher Core MHz, I'm confident this card could handle it.
Regardless of whether or not you plan to overclock, having reasonable system temperatures is always welcomed. Not only will your machine be more reliable with cooler temps, it will likewise not add any unneeded heat to the room you are in (unless it happens to be wintertime and you keep the windows open, then it might be a good thing).
To test a GPU for idle and load temps, we do a couple things. First, with the test system turned off for at least a period of ten minutes, we measure the room temperature using a Type-K thermometer sensitive of up to 0.1°F. The result from this is placed beside the GPUs name in the graph below. Since we don't test in a temperature-controlled environment, the room temp can vary by a few degrees, which is why we include the information here.
Once the room temp is captured, the test system is booted up and left idle for ten minutes, at which point GPU-Z is loaded up to grab the current GPU Core temperature. Then, a full run of 3DMark Vantage is run to help warm the card up, followed by another run of the same benchmark using the Extreme mode (1920x1200). Once the test is completed, we refer to the GPU-Z log file to find the maximum temperature hit. Please note that this is not an average. Even if the highest point was only hit once, it's what we keep as a result.

Luckily enough, the room temperature was exactly the same between testing the stock-clocked HD 4870 and Palit's Sonic "Dual Edition", and I have to admit... I didn't expect such stark differences. Palit's dual-fan cooler on their overclocked card helped lower the load temperature by 21°C over the stock-clocked card using a reference cooler. Likewise, the idle temperature dropped by 13°C.
These results are very impressive, and I have to wonder why ATI themselves ever stick to the same general reference design from release to release. The sad thing is, while Palit gave us a cooler that dramatically drops the core temperature, we are still unable to go far out of spec with our overclocking, which is too bad, because it's highly likely that the card could handle frequencies beyond what ATI allows with their control center.
In the age where anyone can appreciate good power efficiency, it's almost upsetting to see how much wattage any graphics card manages to pull from our walls. Even the lowest-end models don't seem too impressive when compared to the power efficiency of a CPU, but that's how it is, at least right now. It's interesting to see how different GPUs compare in this regard, as some might perform better than others, but use less power, like we normally see with a shift to a smaller process node.
To help see what kind of wattage a given GPU eats on average, we use a Kill-A-Watt that's plugged into a power bar that's in turn plugged into one of the wall sockets, with the test system plugged directly into that. The monitor and other components are plugged into the other socket and is not connected to the Kill-A-Watt. For our system specifications, please refer to our methodology page.
Like our temperature testing, the computer is boot up and left idle for ten minutes, at which point the current wattage reading is recorded. To test for full load wattage, 3DMark Vantage is again loaded up and run at the "Extreme" setting. The space flight test is used here, with the result being recorded during a specific sequence during that run where it seems to stress the GPU the most.

Palit's Sonic card might have higher clocks than the reference model, but it manages to use less power overall at full load - 7W less. At idle, the power efficiency is near-identical to the reference model, only hogging one extra watt. I assume the savings seen at the full load settings is the result of a cooler-operating card. The dual-fan cooler does a great job keeping temperatures at a minimum and as a result, the card's not overheated and doesn't need to stress the fan. It's a win/win all around.
When AMD first launched their ATI Radeon HD 4000 series, it became quickly obvious who the new leader in GPUs were. Compared head-to-head with NVIDIA's recent top offerings, ATI's cards delivered just what gamers were looking for, and at reasonable prices - so reasonable, that it forced NVIDIA to make stark price drops just to compete.
Well, that was great then, and things are even better now. While the HD 4870 retailed for $299 at launch, it's easy to find models for much less at various e-tailers, especially if you want to take advantage of mail-in-rebates. In the case of this particular card, NewEgg is currently selling it for $269.99, or $249.99 after MIR, which is a deal in every sense of the word - there's virtually no cost premium.
Let's face it. The overclock available with a flick of the switch on this Sonic card is small, and the differences in real-world tests are minimal. What makes this card so great is the general performance we've come to appreciate, along with the features and cooler. It's all made better by the fact that despite the additions, the card is still priced less than most of the competition.
The card features a robust and effective cooler, and a DisplayPort connection built right in. Past that, there is an adapter included for both HDMI and VGA support, so the card pretty-well has it all: performance, features and a great price. It's impossible to go wrong with this card, and once again, Palit has earned itself one of our Editor's Choice awards.
Have a comment you wish to make on this article? Recommendations? Criticism? Feel free to head over to our related thread and put your words to our virtual paper! There is no requirement to register in order to respond to these threads, but it sure doesn't hurt!
Palit Radeon HD 4870 Sonic Dual Edition
Copyright © 2005-2009 Techgage Networks Inc. - All Rights Reserved.