It had been reasonable to expect good overclocking from GeForce 7900 GT or GS cards because the GPU’s frequency potential had been far from exhausted, but we are far from optimistic about the GeForce 7950 GT model. 650MHz is obviously the practical limit for the G71 and the percentage of chips capable of working at such a high frequency is rather low. As opposed to that, almost each sample of the G71 is stable at 550MHz, but further overclocking depends entirely on your luck. We were rather lucky as we managed to increase the GPU clock rate to 610MHz and our card was stable at that, but we don’t expect each sample of the GeForce 7950 GT to be as good at overclocking as ours.
The memory chips showed much better overclockability, speeding up from 700 (1400) MHz to 790 (1580) MHz. So, although we didn’t surpass the memory frequency of the original GeForce 7800 GTX 512, we got very near to the level of the GeForce 7900 GTX.
Our overclocked GeForce 7950 GT turned to be much alike to the GeForce 7900 GTX, so we decided not to benchmark it in this mode. Its performance just wouldn’t differ much from that of the GeForce 7900 GTX.
We want to remind you that the stock cooler installed on the GeForce 7950 GT has too low performance to attempt serious overclocking. You should replace it or use additional cooling means to avoid damaging your graphics card at overclocked frequencies.
Testbed and Methods
We testedthe performance of our Nvidia GeForce 7950 GT graphics card on the following platforms:
- AMD Athlon 64 FX-60 CPU (2x2.60GHz, 2x1MB L2 cache)
- Abit AN8 32X mainboard (nForce4 SLI x16) for Nvidia GeForce cards
- Asus A8R32-MVP Deluxe mainboard (ATI CrossFire Xpress 3200) for ATI’s Radeon cards
- OCZ PC-3200 Platinum EL DDR SDRAM (2x1GB, CL2-3-2-5)
- MaXT 512Bor MaXLine III 7B250S0 hard disk drive (Serial ATA-150, 16MB buffer)
- Creative SoundBlaster Audigy 2 sound card
- Enermax Liberty 620W power supply (ELT620AWT)
- Samsung SyncMaster 244T monitor (24”, 1920x1200@75Hz max display mode)
- Microsoft Windows XP Pro SP2, DirectX 9.0c
- ATI Catalyst 6.10
- Nvidia ForceWare 91.47
The drivers were set up in such a way as to provide the highest possible texture filtering quality.
- Catalyst A.I.: Standard
- Mipmap Detail Level: High Quality
- Wait for vertical refresh: Always off
- Adaptive antialiasing: Off
- Temporal antialiasing: Off
- High Quality AF: On
- Other settings: by default
- TeXT 512MBure Filtering: High Quality
- Vertical sync: Off
- Trilinear optimization: Off
- Anisotropic mip filter optimization: Off
- Anisotropic sample optimization: Off
- Gamma correct antialiasing: On
- Transparency antialiasing: Off
- Other settings: by default
We selected the highest possible graphics quality level in each game. We didn’t edit the games’ configuration files. The speed was measured using the game’s integrated tools or, if not available, by means of the Fraps program. We also measured the minimum speed of the cards where possible.
Besides the two standard resolutions of 1280x1024 and 1600x1200 pixels, we also used a resolution of 1920x1200 pixels with an aspect ratio of 16:10 in games supporting widescreen modes. We enabled FSAA and anisotropic filtering from the game’s menu. If this was not possible, we forced them using the appropriate driver settings.
We decided not to run any tests with disabled FSAA, because the technical specification of the graphics cards priced between $259 and $299 allow us to hope for relatively high performance even with FSAA 4x enabled. We ran the tests with disabled FSAA only for those games that do not support FSAA due to the specifics of their engine or use HDR (FP16). The thing is that the GeForce 7 family cannot perform FSAA along with floating-point HDR.
Unfortunately, at the time we were working on this article we didn’t have a Radeon X1950 XT 256MB at our disposal therefore we had to use a slightly more expensive model with 512MB of memory. Here is the full list of graphics cards, besides Nvidia GeForce 7950 GT that participated in our test session:
- ATI Radeon X1900 XT (R580, 625/1450MHz, 48pp, 8vp, 16tmu, 16rop, 256-bit, 512MB)
- ATI Radeon X1950 Pro (RV570, 575/1380MHz, 36pp, 8vp, 12tmu, 12rop, 256-bit, 256MB)
- Nvidia GeForce 7900 GTX (G71, 650/1600MHz, 24pp, 8vp, 24tmu, 16rop, 256-bit, 512MB)
- Nvidia GeForce 7900 GT (G71, 450/1320MHz, 24pp, 8vp, 24tmu, 16rop, 256-bit, 256MB)
- Nvidia GeForce 7900 GS (G71, 450/1320MHz, 20pp, 7vp, 20tmu, 16rop, 256-bit, 256MB)
The following games and applications were used in this test session:
First-person 3D shooters:
- Battlefield 2
- Call of Duty 2
- Far Cry
- Tom Clancy’s Ghost Recon Advanced Warfighter
- Half-Life 2: Episode One
- Quake 4
- Serious Sam 2
Third-person 3D shooters:
- Hitman: Blood Money
- Splinter Cell: Chaos Theory
- Tomb Raider: Legend
- The Elder Scrolls IV: Oblivion
- Titan Quest
- Pacific Fighters
- X3: Reunion
- Age of Empires 3
- Rise of Nations: Rise of Legends
- Futuremark 3DMark05 build 1.2.0
- Futuremark 3DMark06 build 1.0.2