Sign in with
Sign up | Sign in

Raw-power and Fill Rate Tests, Continued

ATI Radeon HD 4850: Smarter by Design?
By , Florian Charpentier

Let’s continue the Pixel Shader 2.0 tests:

Radeon-HD4850 Geforce-9800-GTX+ GTX-260 GTX-280 Grafikkarten

The AMD architecture wins out – the 800 processing units are in evidence and despite their higher frequency, the GT200’s 240 units couldn’t keep up. Now let’s see the results with a newer version, using modern shaders.

Radeon-HD4850 Geforce-9800-GTX+ GTX-260 GTX-280 Grafikkarten

Again it’s not even close: the RV770’s 800 ALUs left the GeForce in the dust. On the other hand, as soon as the test puts the accent more on texture instructions...

Radeon-HD4850 Geforce-9800-GTX+ GTX-260 GTX-280 Grafikkarten

...the RV770 and its 40 texture units can’t compete with the G92’s 64 and the GT200’s 80. There’s no surprise as far as raw fill rate goes – the number of ROPs haven’t changed, still being limited to 16:

Radeon-HD4850 Geforce-9800-GTX+ GTX-260 GTX-280 Grafikkarten

All the GPUs logged scores that were very close to their theoretical values (10800 Mpixels/s for the 9800 GTX, 19264 for the 280 GTX and 24800 for the 3870x2). But we should mention the surprising results our 4850 attained, which even surpassed (if slightly) its theoretical value (10000 Mpixels/s). That was probably the result of the slight overclocking on the Asus model we tested (we’ll come back to that later).

Display all 84 comments.
This thread is closed for comments
Top Comments
  • 11 Hide
    Malovane , June 25, 2008 12:21 PM
    No offense, Fedy Abi-Chahla and Florian Charpentier, and thanks for the hard work, but I think the article should be revised a bit. First off, this should be a review of graphics cards.. not a burned out overclocked Asus motherboard. If you attribute your 4850 test crashing due to your motherboard.. why throw in results of 0 across the board for the 4850? You just corrupted your data and made the final fps averages meaningless, which is the thing people were generally interested in. Secondly, why in the world are you including tests that don't fit the definition of "playable" on any card in your test lineup (Crysis 2560x1600). It just throws off averages, as people aren't going to run this game at 7fps! If there's no card in the lineup that gets close to 30fps in a certain test, just move on! Save it for the quad crossfire or triple sli tests or something. You're giving high weights to resolutions that only a fraction of a percentage point of dedicated gamers can utilize (and those wouldn't bother with a single GPU). Lastly, please get those annoying gigantonormous screenies out of the review. It makes the review look like it was done by kindergarteners.
  • 11 Hide
    Sceptrix , June 25, 2008 3:06 PM
    Oh I see. Because the motherboard can't run GRID, the 4850 gets a score of zero in graphics tests. That makes perfect sense. After all, anyone using this GPU MUST have the same mobo.
Other Comments
  • 4 Hide
    Neog2 , June 25, 2008 9:13 AM
    Wow $200 in Best Buy for a HD 4850,
    $450 in Best Buy for a GTX 260.
    And the 4850 is pretty close to the 280.

    Ouu the 4870 is going to give Nvidia a run for there money
    for the first time in a while.
  • 2 Hide
    Sarcastic , June 25, 2008 9:14 AM
    Good stuff now we just need some 4870 benchies!
  • 2 Hide
    Prodromaki , June 25, 2008 9:55 AM
    Oced Asus and 4850 instead of 4870 + too many games based on engines favoring nVidia...

    P.S. +1000 -> 2222
  • -1 Hide
    randomizer , June 25, 2008 10:10 AM
    Florian you put a 1920×1077 image on the Crysis page! :lol: 
  • 2 Hide
    Anonymous , June 25, 2008 10:24 AM
    For Mass Effect the Engine limits the Maximum framerate to 62FPS. You can change this in the BIOENGINE.INI file (in the Documents\BioWare\Mass Effect\Config\ folder on Vista) by changing the value:

    MaxSmoothedFrameRate=62 in the Engine.GameEngine section
  • 6 Hide
    puterpoweruser , June 25, 2008 12:18 PM
    I can't believe it took nVidia coming out with a new card again to have tom's make a review finally of the 4850.

    "it was unavailable due to the sloppy handling of this launch"
    Seriously? AMD can't control if their retail partners screwed the pooch on the release date, because they were so anxious to get people this great product. They made sure the product was readily available well before the launch date.

    They should be praised for not having a paper launch, not told that it was a sloppy launch, very poor form saying that.

    Hell i went to best buy and bought 2 4850's on sunday, when the cards weren't even supposed to be available yet, the guy told me "they have been in stock for over a month in the back, they aren't supposed to be available yet but i can get two for you." Were the AMD police supposed to come and smack best buy on it's hand and keep me from giving them profits?

    Sorry if i'm ranting, just put the blame where it belongs.
  • 11 Hide
    Malovane , June 25, 2008 12:21 PM
    No offense, Fedy Abi-Chahla and Florian Charpentier, and thanks for the hard work, but I think the article should be revised a bit. First off, this should be a review of graphics cards.. not a burned out overclocked Asus motherboard. If you attribute your 4850 test crashing due to your motherboard.. why throw in results of 0 across the board for the 4850? You just corrupted your data and made the final fps averages meaningless, which is the thing people were generally interested in. Secondly, why in the world are you including tests that don't fit the definition of "playable" on any card in your test lineup (Crysis 2560x1600). It just throws off averages, as people aren't going to run this game at 7fps! If there's no card in the lineup that gets close to 30fps in a certain test, just move on! Save it for the quad crossfire or triple sli tests or something. You're giving high weights to resolutions that only a fraction of a percentage point of dedicated gamers can utilize (and those wouldn't bother with a single GPU). Lastly, please get those annoying gigantonormous screenies out of the review. It makes the review look like it was done by kindergarteners.
  • 7 Hide
    puterpoweruser , June 25, 2008 12:27 PM
    I didn't finish reading the whole article yet but was the driver hotfix and the current 8.6 driver applied to the 4850?? It improved performance and stability greatly as i saw, it make the actual clock speed the card is set it run nicely and gives it great overhead to overclock through the CCC
  • 3 Hide
    draxssab , June 25, 2008 12:48 PM
    Who wants the Radeon 4800 full revew? (including the 4870, that do better than the GTX 280 in some games!)

    http://www.hardware.fr/articles/725-8/dossier-amd-radeon-hd-4870-4850.html

    In french, but the graphs talk by themselves. Ho, and if you want a short translation = impressive and incredibly more efficient than Nvidia (if you compare the size of the GPU, yes it's A LOT more efficient)
  • 1 Hide
    spaztic7 , June 25, 2008 12:58 PM
    These reviews are getting better! Although I have seen many benchmarks and tests of the 4850 before this, I still love seeing how the 48x0 line is doing against the green machine! Anandtech.com has a kill 4870 review!
  • 2 Hide
    Haiku214 , June 25, 2008 1:05 PM
    The downside is the 4850 GPU temp! Makes me rethink about buying one...
  • 0 Hide
    GlItCh017 , June 25, 2008 1:26 PM
    Great read, I've read up on the 4870 on Anandtech as well and it's pretty much the same deal with completely dominating the new Nvidia cards. Performing like a GTX 280 at $100 cheaper. I love the amount of games benchmarked results can vary so much from game to game especially with CF and SLI. Good read!
  • 1 Hide
    homerdog , June 25, 2008 2:05 PM
    This is quite the turnaround for ATI and we haven't even seen R700 yet. Good times!
  • 1 Hide
    dragoncyber , June 25, 2008 2:14 PM
    I'm still waiting to see how the 4870X2 scales in crosfire before I make any decisions. My 8800 GT SLI setup is still beating out the GTX280 and scales better than CrossFire currently. Hopefully by the time the 4870X2 comes out ATI will have fixed the drivers and scaled the crossifre better for more games. What we all know is that obviously the 4870x2 will outperform the GTX280, we must then wait and see if Nvidia goes 260X2 or 280X2 to stay on top of the graphics King of the Hill game.

    The next couple of months are going to be interesting.
  • 0 Hide
    techtre2003 , June 25, 2008 2:23 PM
    extremetech.com has numbers up for the 4870.
  • 6 Hide
    mikeinbc , June 25, 2008 2:23 PM
    Absolutely BIAS REVIEW!

    The review uses OLD DRIVERS for the ATI cards & gets worse from there.

    The drivers used for Nvidia's cards> ForceWare 177.34 beta. Release Date: June 15, 2008
    The drivers used for HD4850> Catalyst 8.22 Release Date: July 17, 2006!!!
  • 0 Hide
    Ogdin , June 25, 2008 3:05 PM
    HardwareCanucks has a detailed 4870 review also.
  • 11 Hide
    Sceptrix , June 25, 2008 3:06 PM
    Oh I see. Because the motherboard can't run GRID, the 4850 gets a score of zero in graphics tests. That makes perfect sense. After all, anyone using this GPU MUST have the same mobo.
  • 0 Hide
    Florian Charpentier , June 25, 2008 3:16 PM
    mikeinbc > How can you believe that a 2 year-old driver can support a new generation GPU? The 8.22 beta driver was simply the only driver available that supported the 4850 at the time of review. It is more something like a Catalyst 8.7 beta.

    Malovane > Unfortunately we needed to review Crysis in 25*16 in order to be able to calculate the average framerates of each board with each resolution, including the 25*16 that is playable on most games. And to show that no board can currently run Crysis in 25*16.

    "If you attribute your 4850 test crashing due to your motherboard.. why throw in results of 0 across the board for the 4850?" > No, I attribute the crashing to the Asus Radeon 4850 TOP that we reviewed.
  • 1 Hide
    Florian Charpentier , June 25, 2008 3:21 PM
    By the way, as I explained again in the review, I didn't include the 0 score of the 4850 with GRID in the sum-up benchmarks.
Display more comments