NVIDIA Announces AMD Integrated Graphics
by Wesley Fink on September 20, 2005 1:08 PM EST- Posted in
- CPUs
The GeForce 6100 Family
There are two Northbridges, the 6100 and the 6150, and 2 Southbridges, the 430 and 410, that can be mixed and matched to cover a wide range of Integrated Video Solutions.
While nVidia did not include this specification in their release, there is one large potential negative with GeForce 6100. 6100 uses only 2 pixel pipelines, the same number of pixel pipelines used in ATI's Radeon Express 200. Since the gaming performance of the ATI was barely acceptable, even for the low end, the performance of the 6100 is not likely to be very exciting.
On the positive side, GeForce 6100 does NOT automatically disable on-board graphics when a PCIe Graphics Card is inserted. That means the integrated graphics plus graphics card can drive up to 4 monitors if the motherboard supports both integrated video outputs. This is also a feature on the ATI Radeon Xpress200 chipset.
The most feature-laden combination is the GeForce 6150 with the nForce430, including unique features like TV Encoder, High Definition (1080ip/1080i) playback, and Gigabit LAN.
More basic configurations are available for use in applications where the top features are not needed, or price is a big concern. NVIDIA has no plans to provide SLI capabilities on the GeForce 6100 boards. The boards are ATX and aimed more at the mainstream market served by system builders, system integrators, and OEMs.
There are two Northbridges, the 6100 and the 6150, and 2 Southbridges, the 430 and 410, that can be mixed and matched to cover a wide range of Integrated Video Solutions.
Specifications: | NVIDIA GeForce 6150 NVIDIA nForce 430 |
NVIDIA GeForce 6100 NVIDIA nForce 430 |
NVIDIA GeForce 6100 NVIDIA nForce 410 |
CPU | Athlon 64 or Sempron | Athlon 64 or Sempron | Athlon 64 or Sempron |
PureVideo (High Definition) | Yes | Yes | Yes |
DirectX® 9.0 Shader Model 3.0 Support | Yes | Yes | Yes |
TV Encoder | Yes | No | No |
TMDS/DVI | Yes | No | No |
Graphics Clock | 475 MHz | 425 MHz | 425 MHz |
PCI-Express | 1x16 2x1 |
1x16 1x1 |
1x16 1x1 |
MPEG-2/WMV9 Playback | HD(1080p/1080i) | SD | SD |
Video Scaling | High Quality(5x4) | Basic (2x2) | Basic (2x2) |
SATA/PATA drives | 4/4 | 4/4 | 2/4 |
SATA speed | 3Gb/s | 3Gb/s | 3Gb/s |
RAID | 0,1,0+1,5 | 0,1,0+1,5 | 0,1 |
NVIDIA MediaShield | Yes | Yes | Yes |
NVIDIA ActiveArmorTM Firewall | Yes | Yes | - |
Ethernet | 10/100/1000 | 10/100/1000 | 10/100 |
USB ports | 8 | 8 | 8 |
NVIDIA nTuneTM Utility | Yes | Yes | Yes |
While nVidia did not include this specification in their release, there is one large potential negative with GeForce 6100. 6100 uses only 2 pixel pipelines, the same number of pixel pipelines used in ATI's Radeon Express 200. Since the gaming performance of the ATI was barely acceptable, even for the low end, the performance of the 6100 is not likely to be very exciting.
On the positive side, GeForce 6100 does NOT automatically disable on-board graphics when a PCIe Graphics Card is inserted. That means the integrated graphics plus graphics card can drive up to 4 monitors if the motherboard supports both integrated video outputs. This is also a feature on the ATI Radeon Xpress200 chipset.
The most feature-laden combination is the GeForce 6150 with the nForce430, including unique features like TV Encoder, High Definition (1080ip/1080i) playback, and Gigabit LAN.
More basic configurations are available for use in applications where the top features are not needed, or price is a big concern. NVIDIA has no plans to provide SLI capabilities on the GeForce 6100 boards. The boards are ATX and aimed more at the mainstream market served by system builders, system integrators, and OEMs.
44 Comments
View All Comments
Johnmcl7 - Wednesday, September 21, 2005 - link
That was exactly what I was thinking, as far as I'm aware the Xpress200 integrated graphics card is only a two pipeline solution.John
toyota - Wednesday, September 21, 2005 - link
are we the only ones that noticed?? i wish they would correct this then. anandtech made the 2 pipeline nvidia sound inferior to the ati. everthing that i have seen says 2 pipelines for the ati too. i would like a correction or explanation.Johnmcl7 - Wednesday, September 21, 2005 - link
Looks like it, even though we seem to be stuck in bold now!It's a fairly large mistake, as it means the nvidia chip is likely to offer similar performance or better, rather significantly inferior.
John
TrogdorJW - Wednesday, September 21, 2005 - link
Wow... we need to add editing to posts. At least for some of the admins/writers....Here, let's try to kill the bold.
Did it work? (I think there were two stray bold tags.)
TrogdorJW - Wednesday, September 21, 2005 - link
Hooray! :Dxsilver - Tuesday, September 20, 2005 - link
To those naysayers thinking that nvidia is going to stumble under the 90nm process, it doesnt look good. Granted this chip is much simpler than the r520Or could it be the reason it is only 2 pixel pipelines due to yield issues with the low nm? interesting....
JarredWalton - Tuesday, September 20, 2005 - link
I think there's about 50 to 100 million transistors in the typical chipset already, so adding ~25 million for a two pixel pipeline GPU isn't insignificant. (Just guessing on numbers - I could be off.) Certainly, we're not talking 300 million transistors like G70 or R520, but chipsets have a lot of other stuff without adding a GPU core. 475 MHz on the GPU sounds at least somewhat promising.tfranzese - Wednesday, September 21, 2005 - link
But, don't forget they split it into a two chip solution now. This isn't simply adding the transistor counts of the GPU to the previous chipset's.JarredWalton - Thursday, September 22, 2005 - link
True. Of course, GPU manufacturers want to protect their discrete graphics markets. You also won't get discrete GPU performance without adding on-board (or on-chip) RAM, as the shared memory doesn't offer nearly enough bandwidth for a high-powered GPU. If ATI or NVIDIA could develop a board with high-powered IGP - like say 8 pipelines or more - but that chipset cost $100 instead of $50, board manufacturers and end users wouldn't buy it. They try to add as much power to the IGP as possible without increasing costs much, which leads to things like shared RAM and 2 pipelines. At least, that's my take.Doormat - Tuesday, September 20, 2005 - link
NEXT. No HDMI out = no hi-def video replay under windows Vista. Good thinkin' nVidia.