![[H]ardOCP](/images/blank.gif){kind=small}
- Date:
- Sunday , June 27, 2004
- Author:
- Brent Justice
- Editor:
- Kyle Bennett
- Google +1
NVIDIA SLI
It’s back! NVIDIA is re-introducing multiple video card 3D acceleration, and it may just change your mind on your next video card purchase.
NVIDIA SLI Multi-GPU:
Welcome to NVIDIA SLI Multi-GPU technology.
Over ten years ago back in 1993 the only way to get graphics parallelism was in extremely expensive supercomputers, such as the Onyx. Even with all that you still only got the level of graphics you see above; that is SGI’s Performer running on Onyx.
1998 is when 3DFX introduced their Voodoo2 video cards capable of doubling them in order to gain more performance. Their SLI technology is shown above. A ribbon cable was required to connect the two Voodoo2 video cards together, and an external pass-through cable was required as well as a regular 2D graphics card. The games that the Voodoo2 were accelerating fast back there were games such as Quake2 which is also shown above on Voodoo2 SLI.
Now in 2004 NVIDIA has introduced their new SLI technology, but note that it does not stand for Scan Line Interleaving. This new technology is called "Scalable Link Interface" or SLI for short. Of course all of us old timers still get excited when we hear "SLI" and NVIDIA knows this. The new SLI is possible because of the new PCI-Express bus that was introduced last week. NVIDIA is telling us that when used with the Unreal3 engine, they are seeing increases in frame rate of up to 87%. To qualify this statement, I think it is needed to note here that an engineer stood behind this statement, not just NVIDIA's marketing department.
Thanks to PCI-Express we will be seeing motherboards with multiple X16 PCI-Express slots. This means multiple PCI-Express x16 graphics cards can now be installed! From a technical standpoint, when using this new form of SLI, one board will be running on a X16 while the other will be using a X8 PCI-E bus, but both cards will be using a X16 slot configuration.
In the first picture above you can see a die shot of the NV40 and you will see there is in fact circuitry in the die for this new SLI technology. In the second picture above you can see what two of the video cards will look like in SLI configuration on the PCI-E x16 bus. Note that each card is being provided power with the new power connector for PCI-E video cards. This means if you do not have these special connectors on your power supply it will take no less than four Molex connectors to power both cards.
This is how the new Scalable Link Interface technology works between both GPUs. As you can see these cards are NOT rendering even or odd lines, they are in fact rendering whole sections of the screen at a time. In this picture it looks like one card is rendering three fourths of the screen while the other card is doing one fourth of the screen. NVIDIA has a dynamic load balancing logic that will assign each card even workloads rather than simply splitting the screen into equal size portions. The load balanced image is then put together for you on your display.
NVIDIA is taking SLI technology on the road. If you are at any of these places during these times make sure to check it out in person. If appears they will also have this setup at this years Quakecon this year in Grapevine, Texas. Make sure to stop on by, and you just may see us there as well.