Lucid Hydra 100 - How it works continued
(Page 3 of 4 )
The Lucid Hydra 100 works in between the chipset and the graphics cards. When the code is sent from the CPU, it goes to the Hydra 100 chip. The chip's job is to analyze the code, and then pass it along to each GPU to do its work. The great part of this is that it can divide up the work equally. This means that each GPU can be utilized the best way.
The great thing about this chip is that it is almost perfectly linearly scalable, or at least a lot better than what SLI and Crossfire can do. Sure, it might work great for 4 GPUs, but what about 10? I’m sure it won’t be able to handle 10 GPUs. It will be interesting to see where the Hydra 100 starts to drop the ball. I’m sure that there will reach a point where the Hydra 100 can’t keep up with the amount of work it needs to push out. As games become more and more graphic-intense, will it be able to handle the greater demand on the GPU’s?
The biggest potential issue I see in the Hydra 100 chips is that it relies on knowing what it is dealing with when the instructions are sent to the chip. It works with OpenGL and DirectX. What if another game engine comes along? This chip won’t know what to do with the instructions. DirectX 11 is on the way soon, will this be compatible?
There are a lot of questions still in the air about the Hydra 100 chip that promises linear scaling GPUs. The only way I see that this chip has a chance is if they release new versions along with updated logic cores.
What looks to be an exciting aspect is that none of the load balancing is automatically loaded onto any one GPU. One of the problems of SLI and Crossfire is that they can’t balance the work between two GPUs of different power. It has been a rocky road for SLI or Crossfire to work with different cards. The Hydra 100 does the load balancing for the cards, so with some work, it should be able to efficiently divide the load between two cards with different GPUs. Hydra 100 doesn’t care if they are the same cores or not; it should be able to easily handle different GPUs -- and not just two different GPUs, but perhaps even different GPUs in each and every slot.
Next: Availability >>
More Video Cards Articles
More By jkabaseball
