Interesting Article on the State of Lucid Hydra

[defaultluser]

I expected problems with the technology of mixing multiple unrelated GPUs. But still, I'm surprised by the issues Hardware Heaven encountered now 8 months after the first boards/benchmarks were released (plenty of time for drivers to mature).

http://www.hardwareheaven.com/articles.php?articleid=147

I guess the question is, is this all just a driver support problem, or are there just too many technical hurdles to make this a workable technology?

Anyway, I thought I'd pass this along since a number of high-end multi-GPU motherboards are getting the Lucid treatment, and a large number of you run with 2 or more GPUs in exotic setups

 

[WBurchnall]

I'm not sure how cherry-picked that one bunchmark could be ...but... yeah those Crysis: Warhead numbers with Hydra mixed video cards look pretty awful...

 

[Sovereign]

Not saying Hydra is "all that," but I do feel obliged to point out Anand got some pretty decent results. It's not great, but it certainly looks better in this Jan '10 set of tests than in the referenced Sept. '10 Crysis test. I think, like SLI when it first came out, Hydra is highly dependent on which game you're using.

 

[Godmachine]

Well it was extremely well hyped by sites for its one of a kind design and intent. The idea behind it is wonderful but driver support for this kind of technology is without a doubt THE most important aspect of it. You can't do what Hydra does and only have mediocre driver support 8 months after it ships.

I think if Nvidia or AMD had enough common sense one would purchase this technology and put it out on a board of some kind. They have the cash flow to assign a driver team to tackle it and really make some advances on the drivers/performance. One of them could put themselfs ahead of the game.

 

[Savoy]

Sadly if this tech had any real merit from way back when we first heard about it then I'm sure nvidia or amd would have bought lucid out a long time ago. I'm not surprised this is pretty much flopping on its face.

 

[Kueller]

Well it was extremely well hyped by sites for its one of a kind design and intent. The idea behind it is wonderful but driver support for this kind of technology is without a doubt THE most important aspect of it. You can't do what Hydra does and only have mediocre driver support 8 months after it ships.

I think if Nvidia or AMD had enough common sense one would purchase this technology and put it out on a board of some kind. They have the cash flow to assign a driver team to tackle it and really make some advances on the drivers/performance. One of them could put themselfs ahead of the game.

But neither AMD nor Nvidia has any motivation at all to develop a product that encourages you to choose a competitor's video card when you're currently locked into doubling up on one or the other. If anyone were to pick it up it would have to be Intel.

 

[RamonGTP]

I think this technology was dead before it ever got started personally.

 

[gmkmay]

Well it was extremely well hyped by sites for its one of a kind design and intent. The idea behind it is wonderful but driver support for this kind of technology is without a doubt THE most important aspect of it. You can't do what Hydra does and only have mediocre driver support 8 months after it ships.

I think if Nvidia or AMD had enough common sense one would purchase this technology and put it out on a board of some kind. They have the cash flow to assign a driver team to tackle it and really make some advances on the drivers/performance. One of them could put themselfs ahead of the game.

I would put money on this actually being Lucid's business model. Maybe not AMD or Nvidia, but I'm sure they're end goal was/is to sell the technology. I'm betting they actually wished they could have sold it before the release, thus the hype. They must have known the performance wasn't going to win over too many people and most of their funds probably went into marketing (like big expensive sales pitches to motherboard manufacturers). If it doesn't sell soon I'd wager bankruptcy somewhere in 2011.

Then again, maybe I'm wrong.