While it's always exciting to receive review samples from the big boys in the hardware world, it's also nice to take a look at something a little different and this almost always comes from the little guys. Now, Vvikoo are manufacturers of some top end nVidia and ATI based GPU solutions, so using the word "little" is probably a little harsh, but they are definitely the new kid on the block and have a lot to prove. What I have with me today is an 8800GT, but not just any GT... oh no. This one has 1gb of DDR3, a huge aftermarket Zalman cooler, and it's purple.

System Configuration Used:
- Test Setup
- Power Supply: Spire Rocketeer 600w
- Motherboard: Gigabyte P35T-DQ6 rev. 1.0
- Processor: Intel Core 2 Duo E6750 @ 3.2 GHZ
- Hard Drive: Maxtor DiamondMax 20 80GB SATA
- Video: Vvikoo 8800GT Max 1GB, PNY 8800GTS 320MB, MSI 8800GT 512MB
- Memory: Crucial DDR3 12800 2GB @ 1600 MHZ
Software Configuration - Operating System: Windows XP Pro x86
- Video Driver: NVIDIA ForceWare Version 169.21
- 3DMark2006
- Crysis
- Lost Planet
- Unreal Tournament 3
- Core Clock: 600MHz
- Memory Clock: 1900MHz
- Memory Type: GDDR3
- Memory interface: 256Bit
- Memory capacity: 1024MB
- Manufacture Process: 65nm
- Bandwidth (GB/s): 57.6
- Standard Shader Model: 4.0 (Vista Only)
- Standard DirectX Version: 10.0 (Vista Only)
- OpenGLR2.0 support
- 128-Bit HDR Lighting
- HDCP Compliant
- NVIDIA PureVideo HD Technology
- NVIDIA SLI Technology ready
- NVIDIA nViewer Multi-Display Technology
- Interface: PCIe 2.0
- Power Consumption (Watts): 138
















Seriously, for a next-revision GPU that "high-end", that doesn't seem too pretty
There are a couple other 1GB GTs. Asus makes one, and Palit used to. But it's kind of funny, because Palit stopped producing it and said 1GB didn't make enough of a performance difference to justify the cost. You're better off getting a G92 GTS if you're going to spend that much.
That cooler is great though. I hope they put it on more GTs.
That's an important consideration.
That's an important consideration.
Indeed it is, or not, in fact you should complain at the text source and not here, right?
But I don't care because i'm not planing buy a new card.
Appologies for the Crysis results, they are as you say, a load of crap
I am now running our test rig with a new benchmark (the command prompt packaged with the game) to give you the correct results. As you pointed out (as you changed our methods as a result) the card does get 2-3FPS at best at 1600x1200.
You can read more about this here if you want: http://forums.xsreviews.co.uk/showthread.php?t=2999 This thread also explains the reason behind the 1GB of memory not boosting performance. In summary, DX9 does not register all of the available memory and hence doesn't use all of the provided chips. In DX10 the stakes change, but we have yet to test this.
It is never our intention to mislead anyone, but I fear thats what has happened, and I appologise sincerely for this. I don't expect you to trust our site again, but I feel you deserve an explanation.
The results themselves are correct in what they are, but were shown in a fashion to indicate in-game performance, rather than this half benchmark we were using.
It says "Operating System: Windows XP Pro x86" at the top.
It says (Standard Shader Model: 4.0 (Vista Only) Standard DirectX Version: 10.0 (Vista Only)) near the bottom.
Does that mean this test was with Win XP minus those two items or is the OS used displayed wrong?
How much better would it have been if it would have been on Vista with DX 10 and the Standard Shader Model?
Thanks,
Those two specification points are 'Vista only' as DX10 supports SM 4.0 while XP (DX9) doesn't. We are working on DX10 benchmarks, but as you can already tell we have enough trouble with DX9 before trying something else
As a point of reference, what game tests do you percieve as being useful in testing DX10 performance?
Thank you for your time, SorX
Commenting has either been disabled on this article or you are not logged in. Click here to login or register, its free!
Note: Anonymous commenting is disabled in order to keep the quality of responses to a high standard.