Top Graphics Hardware from PC Pitstop 3D Video Tests

PC Pitstop Bump map test

In late August 2008, PC Pitstop added a battery of 3D video tests to its free on line OverDrive scan. (Click to see the PC Pitstop 3D video test roll out announcement). The Overdrive 3D video tests consists of a suite of tests covering 6 areas of performance: Transform, Lighting, Multi Texturing, Blending, Vertex Shader and Bump Mapping. OverDrive consolidates the results into a single 3D performance score.

We thought it would be interesting to see what graphics cards were surfacing as the top 3D video performers based on the test results captured during the first month since its roll out.

On the desktop side, the NVIDIA GeForce GTX 280 scored highest with an approximately 15% advantage over the second place ATI Radeon HD 4800 Series graphics controller. NVIDIA and ATI alternated for the top 4 average scores, at which point it became all NVIDIA to finish out the remaining 6 spots in the top 10 results.

Portables came in with an NVIDIA sweep for the top ten average scores. The NVIDIA GeForce 8800M GTX edged out second place NVIDIA GeForce 9800M GTS by less than a 2% average test score margin.


PC Pitstop 3D Video Test Results – Top 10 Desktop System Graphics Hardware

xpspadoptall.gif


PC Pitstop 3D Video Test Results – Top 10 Portable System Graphics Hardware

xpspadoptall.gif


(Visited 38 times, 1 visits today)

38 thoughts on “Top Graphics Hardware from PC Pitstop 3D Video Tests

  1. This test may reflect the actual readings from the test but should not be used to determine what a good video card is.

    Something you should really keep in mind is that video cards may be ATI or Nvidia but the manufacturer makes or breaks the card as a whole. Just because the GPU mounted on the card is an ATI or Nvidia means little. If the card was manufactured poorly by what ever company then the card will run like crap.

    If you want an accurate depiction of what each card is actually about you would need to go find test results where an ATI card manufactured by ATI solely was tested against an Nvidia card manufactured by Nvidia solely with both cards being tested on an identicle set up under identicle conditions. Otherwise you cannot tell which chip is actually better with any accuracy.

  2. Hi all
    This makes for good reading. Its no good putting a £300 card on a £50 motherboard and a £30 power supply your going to get poor results what ever the card! I have a 1x280gtx runing with vista os and my 3d06 score is 17038 http://service.futuremark.com/compare?3dm06=7838165
    and guess what im still no good at online gaming lol.At the end of the day whats does it matter who is the faster its down to the player in the end!

  3. The Overdrive 3D video tests consists of a suite of tests covering 6 areas of performance: Transform, Lighting, Multi Texturing, Blending, Vertex Shader and Bump Mapping …

    Looks like a pretty thorough Test to me , I think that it is better than 3dm06 , personally , because it is more graphics based ,…

  4. There certainly is no lack of heated comments from both camps in this battle over who has the best, fastest, etc. video card, or which manufacturer is best, NVidia or ATI. These types of arguments have been going on from prior to the PC, Windows or Apple, MS/DOS or IBM/DOS, before modems, et al.

    I base my opinions on actual use, benchmarks, experiences in dealing with the various companies, educated discussions with my peers, etc. Based on 30+ years of professional experience in microprocessing, microelectronics and communications, I’ve seen the industry from its beginnings (before the first 300baud Hayes modem); I’ve also had the enjoyable experience of watching it develop into what it is today (all of the companies starting up, falling on their faces, surviving, fading away, etc. Sorry, I digress…

    Overall, NVidia and NVidia-based graphic cards are a better buy than ATI; you need to pay a lot more $ for high-end hardware, but that’s to be expected. More support, more driver updates and a larger owner/user-base, not to mention better customer service (PNY is a good source and they’ve been around for years). No matter what you choose, if you’re wanting high-end performance, you’ll need to first base a system around a solid motherboard and CPU combination that are matched for each other’s use together and that can run circles around your add-ons. In addition, you must also buy a power supply that provides more than is needed by at least 50%; in most cases that’ll mean you start with 1000W and go up. (Here’s an address for those who aren’t sure about their systems’ power needs; refer to the Recommended Power Supply Wattage Calculator at http://support.asus.com/PowerSupplyCalculator/PSCalculator.aspx?SLanguage=en-us and it should let you know better than just guessing. An example of power needs, take PCIex16 SLI cards. Two of them and you’ll need 1000W to begin with! Rember, all kinds of odd things can/will occur when your power supply is only marginal and benchmarking of any type will result in totally inaccurate outcomes.

    Anyway, that’s all I got for now, gotta run!

    cu

  5. Your new OverDrive tests are good — except — they crash at the 3D tests if you are running an older computer that doesn’t have the latest graphics card. OEM integrated graphics is one thing, but the test crashes with a OEM AGP card w/ 16MB of RAM. There should be a way to skip the 3D portion and still run the rest of the test.

  6. ” txyankee Says:
    October 10th, 2008 at 9:21 am

    My nvidia 7800 GTX out performs them all according to the pc online test results. but only using the older drivers not the newer ones…”

    That would be an interesting glitch. I’m curious to see your results for comparison. I have a 7900GTX TDH Extreme that beats every other 7900 or 7800 that I’ve seen to date in any benchmark, and it sits about where it should on the Pitstop tests.

  7. I noticed somebody said something about thier amd beating an intel quadcore. A multiple core processor must have a load on the processor to cause the other cores to kick in. In other words the quad core must have multiple programs running. That was you Flying_Caveman I believe. Many people have wrong conception of multiple core processors. Also the AMD probably had a higher CPU so if you just running one program on the intel it would win.

  8. Well, I guess my ATI HD3200 integrated graphics at 1152×864/70/32bit is a slow Sally at only 485fps, but I’m not a gamer so what do I care?

  9. @Sps87
    Thanks for the heads up on the power supply, but my ATI Radeon HD3200 is a chip set integrated on the MOB and not a seperate Graphics Card. I have the standard 300W Gateway supply. It could be that its not grounded well and is picking up some noise? But so many Gamers and TV adicts are having the same problem that I doubt thats the case.

    Thanks for the Reply

    Jim
    in Dayton OH

  10. Pawpaws Rig, There are a bunch of excellent vid. cards out there. The 7600gt is an excellent card and although I have an 8800gtx I use my 7600gt everyday. What’s important is that it does what you need and not what it scores.

    Thanks for the input.

  11. I am in my 50’s and built a new system for the 2nd time, the 1st time with an ati card and was not impressed mainly because of causing my sytem to be unstable. I then went to an nvidia 7600gt due to a lack of money, but no crash or lock ups. I rebuilt with an asus m2n32-sli deluxe wifi with corsair pc6400 ram, raptor velocity 300 drive and an nvidia 9600 gt 0c card from BFG-Nvidia. It does what ever I want to do with it with no flaws. So what am I saying, I think there is to much bull about who is the best and need to get back to what works for you and not woory about who is the best. But My pick is, and always will be nvidia due to , to many problems with ati. After all, who cares who has the fastest card as long as the individual and the graphics card they have does what they want and are happy with it. Oh by the way, I am an AMD Processor Fan for sure, no wavering from this, I have an amd x2 6000 black edition. I’m happy, just wish all others were. Yes’ I also use a couple of pc pitstops software and am happy there to.

  12. there is little to no speed (fps)difference when using the agp bus or the pci-e bus card of equal gpu. sure the bus size has increased, but unless u are using a 64bit os with a 64bit game or application there is no advantage to the larger bus size. i have a p4 2.5 w/x1650pro agp xphome, and a amd5400 w/2600pro pci-e, the difference is not as much as u would think being the pci-e is running @ x16 bus speed bi-directional,it in theory should be huge.

  13. Would like to see the list divided.
    One for PCI-cards and one for AGP-cards.

    Think there is a lot of AGP-cards still running.

    My ATI Radeon HD3850 gets 4324 points and that is quite decent on an an old P4, 3.2mHz!

  14. I don’t trust any of this anymore. my old AMD 2500 with a 9800pro beats my not quite as old, 3.2Ghz P4 with a 7800gs. The frame rates I see in games are not even close the P4 is way faster. (I guess it depends on the game?) Also, I see my newer system an AMD X2 6000+ with a 8800gs world ranks higher than some of the intel quads with better graphics cards, yet their scores on cpu, memory, and graphics are higher.

    So, remember its only a benchmark.

  15. I am an owner of a gaming notebook (ASUS still under warranty) with Nvidia GeForce 8600M GT card and I am also one of the complainers who while gaming notices temps running as high as 92 to 98 C! There are so many complaints on forums on these issues that I would not recommend buying any Nvidia product until there is a fix for this one. Drivers fail, black screens, freezes and automatic shut down (probably due to heat) plague many of us Nvidia graphics card users. Nvidia has finally recognized the tip of the iceberg is a lot larger underwater but they have not been forthcoming on the problem.

    I would like to see a test of these graphics cards after 6 months of use. 🙂

  16. I’ve read all the remarks, all the humms’ and haws. My point is if you can’t feed the beast your benchmarking will drop significantly. To get the most of either, you have to have the muscle behind it, to date a 450W PS can’t be used in succesful benchmarking, 600 Watt minimum, seems to be the closest I’ve come to benchmarking a true comparison. Running a 1600 watt PS my results take the nVidia solution way neyond what I get on the ATI solution. To each his own.. 😉

  17. Crap information from yet another user. I’m sorry for NVIDIA, but GDDR5 is twice as fast as GDDR3 (which is all a GTX GPU is designed to handle). Time to jump off the bandwagon, people. ATI finally kicked NVIDIA. Check real PC people by trying MaximumPc.com
    They, unlike you tell the facts. Intel is next to go down.

  18. Hey Guys.

    Thanks for all the input.

    We’re working on the dual gpu situation. As soon as I know something more I’ll post it up. In the meantime please keep testing as this contributes to our database and will only serve to make our results more accurate.

    Thanks a lot.

  19. Chea right. Checke again. You’ve got the 4870×2 scored lower than the GTX 260. Um. Like mentioned many times above…the 4870×2 wipes the floor with any card out there.

  20. If this test is only based on using one card, and not SLI, or Crossfire, I think they seam to be pretty accurate. When I’m in crossfire with my 2-4870’s my score is in the 9200 area, but out of crossfire, and one card disabled, my score goes down to 7000

  21. Whether or not the 4870 X2 is the top performer depends on what benchmark is being run. Older benchmarks such as 3DMark 01, 05, Aquamark, etc do not take full advantage of multiple GPU’s and actually show results similar to the OverDrive 3D performance test. Typically, if a program has not been coded to take advantage of an additional GPU, performance gains from such a card are minimal or non-existent.

    What these OverDrive results have revealed is that on the dual GPU cards, the second GPU is not being challenged and therefore does not contribute to the scoring as expected.

    We are looking at the cause and will see what needs to be done to rectify the situation.

  22. @jhowell6

    honestly what power supply are you running for a 280gtx

    the crappy 250-350 power supplies in gateways or any other prebuilt except for the gaming companies are just too small for a powerful card like that.

  23. My nvidia 7800 GTX out performs them all according to the pc online test results. but only using the older drivers not the newer ones….

  24. I have to laugh at these guys who argue with such “vigor” about Nvidia vs ATI. I have had some experience with both brands of cards and can say that the card alone is not the solution or the problem. I’ve had cards that didn’t live up to all the hype and I’ve seen a number of review sites that identify problems with the same card but manufactured by different firms. Anyway, if your system can’t get the max out of your card, then it doesn’t matter which card you have, it’s going to suck. Its like arguing over which football team is the best based only on the review of the quarterback. It don’t matter how well he throws, if there are no linemen he’s going to get sacked and with no receivers he cant complete squat. Even these results should be viewed in light of the entire machine. Maybe the Nvidia cards tested were just in better systems, who knows.

  25. This makes interesting reading to anyone considering which gfx card to go for.
    My own experience tells me to stay well away from ATI.
    I bought a very expensive ATI X1900XT and was astonished at the running temperatures which seemed always to run in the high 70’s and 80’s – yes in degrees celcius!
    3 Months in and the card overheated no matter what i tried.
    The ATI was also louder than my hoover, mainly due to the venting system blowing air out.
    I got a replacement pretty swift, but decided to pay 20 quid more and opt for the water cooled model instead which was highly recommended at the time.
    It was now good to have a quiet system once more.
    However, after about 6 weeks in i started to get broken pixels and wierd lines in my display.
    I checked the running temperatures and they weren’t over the top.
    72 degrees playing BF2 and around 64 degrees on everything else.
    Then my machine started to crash regularly and after a short while this card died.
    Enough is enough i thought so i bought a Geforce 8800GTX 768Meg and since i have had this card i have had no trouble at all.
    I will never ever go back to ATI!
    Nvidia for the win!

  26. I agree, the 4870×2 dominates anything Nvidia currently has, and this test should reflect that. Also 4850 or 4870 in crossfire will handily beat GTX280 in SLI.

    The rest of the results are pretty accurate. 🙂

  27. There is something seriously wrong with those test results. The 4870X2 is the fastest card available. The GTX280 has the fastest single GPU, but doesn’t beat the X2. And to say the X2 only marginally beats an 8800 Ultra, and is stomped out by a 260? Not a chance. So yeah, gg pitstop. Your test probly wasn’t supported by the drivers properly in most of those cases and you’ve wound up with a broken benchmark. Fail.

  28. The thing is, this IS “real life” testing. So it doesn’t matter about your personal “real world tests” or what you “feel”, Stephen… This IS how people run their systems in “real life”; so until they start running them in your configuration and testing them with your “specific” tests, AMD still loses as of September, ’08. Your “opinion” of what is best doesn’t count. You can cheerlead all you want, but you’re liable to get cut-down, lol! Who’d have thunk that the 8800M GTX would beat the 9800M GTS? If you really want to win the video wars that badly, I guess you’re gonna have to make your own benchmarks and get everyone to accept them on “feel”…

  29. Pretty neat poll and survey but there is something left to be disired when the 4870X2 falls behind the 4870 and GTX 280. In real world tests and games, the 4870 1GB handily beats GTX 260 Core 216 and even challenges the GTX 280. The 4870X2 pretty much wipes the floor with the GTX 280 in all but a few rare cases. Then again, the graphics test is a simple one, and I’m sure since it’s a simple test it’s not coded in parallel streams that ATI cards since the R600 series favour. The cards being sold tell a different story with ATI in the lead and nVidia scrambling, which is good. The next round of cards should be fiercly competitive next year. That means the consumer wins, becasue when one of them gets firmly in the lead, we end up paying more for rehashed video cards with limited features.

  30. You can tell the “3D Video Test” is highly unaccurate. The 4870 X2 beats anything on the market hands down at the moment.

  31. Has anyone looked into or been aware of the grafics problem of ATI and possibly NVIDA cards? I have a new Gateway triple core AMD with 2 Gig of RAM that can’t run Window’s Media Center for more than a couple of minuets without crashing.

    The Error I get from the Event Viewer is ATIKMADG. If you Google “ATIKMADG Vista” it will take you to DriverHeaven.Net. Their, Gamers and TV veiwers alike have been complaining about crashing system since March of 2007 until present day. Over 40 pages of comments?

    Do you have any insite into a fix for this problem. It seem bad enough that it will have to hurt sales of these machines before long. A big part of todays advertising is TV on your PC today.

    Thanks for any thoughts on the matter

    PS: Their is some comments about NVIDA having simular problems, but I haven’t researched it because I have an ATI Graphics Adapter HD 3200 integrate on my MOB.

    Jim Howell
    Dayton, OH

Leave a Reply

Your email address will not be published. Required fields are marked *