EVGA have replaced my faulty e-GeForce 8800 GTS 320mb Superclocked graphics card with an e-GeForce 8800 GTS 640mb card. When I received this card I thought they had kindly sent a more powerful GFX card. However a friend told me that the old one is more powerful due to the clock speed of the processor. I ran 3d mark 2006 and am getting scores around 1000 lower on the new card. Here are some statistics I found on the EVGA website
-----------------------------------------------------------
320mb Superclocked
Performance
* NVIDIA GeForce 8800 GTS
* 576 MHz GPU
* 96 Pixel Pipelines
* 400 MHz RAMDAC
Memory
* 320 MB, 320 bit DDR3
* 1700 MHz (effective)
* 68 GB/s Memory Bandwidth
-----------------------------------------------------------
640mb
Performance
* NVIDIA GeForce 8800 GTS
* 511 MHz GPU
* 112 Pixel Pipelines
* 400 MHz RAMDAC
Memory
* 640 MB, 320 bit DDR3
* 1600 MHz (effective)
* 64 GB/s Memory Bandwidth
-----------------------------------------------------------
I contacted EVGA regarding this issue. They replied with the following
“Dear James,
Your findings about performance and featrures are incorrect:
You have sent us a defective graphics card, a 8800GTS with 320 MB of RAM, with a GPU clock of 576 MHz and a memory clock of 1700 MHz
We have returned a new graphics card, a 8800GTS with 640 MB of RAM, with a GPU clock of 500MHz and a memory clock of 1600 MHz.
Having twice the amount of memory on the card should well compensate the slichtly lower GPU and memeory clock rates, and should still give you a by far better gaming experience. This can be prooven by benchmarks such as 3D-Mark 06 and the current version of Futuremark.
Please, after installing this new card, remove the old nnvidia display drivers to install the current version of nvidia display drivers for your card and operating system to ensure optimal product performance,”
Are EVGA correct and should I be experiencing an increase in performance or is my friend and the 3D-Mark 06 result the more accurate. They agreed to replace the fault product with one of equal value so I intend to follow this up if this isn’t the case. However maybe the increase in performance is negligible?
-----------------------------------------------------------
320mb Superclocked
Performance
* NVIDIA GeForce 8800 GTS
* 576 MHz GPU
* 96 Pixel Pipelines
* 400 MHz RAMDAC
Memory
* 320 MB, 320 bit DDR3
* 1700 MHz (effective)
* 68 GB/s Memory Bandwidth
-----------------------------------------------------------
640mb
Performance
* NVIDIA GeForce 8800 GTS
* 511 MHz GPU
* 112 Pixel Pipelines
* 400 MHz RAMDAC
Memory
* 640 MB, 320 bit DDR3
* 1600 MHz (effective)
* 64 GB/s Memory Bandwidth
-----------------------------------------------------------
I contacted EVGA regarding this issue. They replied with the following
“Dear James,
Your findings about performance and featrures are incorrect:
You have sent us a defective graphics card, a 8800GTS with 320 MB of RAM, with a GPU clock of 576 MHz and a memory clock of 1700 MHz
We have returned a new graphics card, a 8800GTS with 640 MB of RAM, with a GPU clock of 500MHz and a memory clock of 1600 MHz.
Having twice the amount of memory on the card should well compensate the slichtly lower GPU and memeory clock rates, and should still give you a by far better gaming experience. This can be prooven by benchmarks such as 3D-Mark 06 and the current version of Futuremark.
Please, after installing this new card, remove the old nnvidia display drivers to install the current version of nvidia display drivers for your card and operating system to ensure optimal product performance,”
Are EVGA correct and should I be experiencing an increase in performance or is my friend and the 3D-Mark 06 result the more accurate. They agreed to replace the fault product with one of equal value so I intend to follow this up if this isn’t the case. However maybe the increase in performance is negligible?