MSI Big Bang Fuzion: Pulling The Covers Off Of Lucid’s Hydra Tech

The Many Heads Of Hydra

Again, we’ll go into more depth on the Fuzion’s basic board design and BIOS when it comes time to explore the similar MSI Trinergy next week. The real differentiator here is Lucid’s Hydra 200 ASIC, which takes the place of Nvidia’s nForce 200. The Trinergy and Fuzion are both $350 P55-based boards, so what this really boils down to is a choice. Choose SLI/CrossFireX with support for up to 3-way SLI. Or choose the flexibility to mix/match cards/vendors.

When you go the former route, you play by the rules set forth by ATI and Nvidia for achieving a compatible multi-GPU configuration. The means identical GPU models in an SLI setup and GPUs from the same family when you go CrossFire. But you also get the validation those companies put into their established technologies.

Should you choose the latter path instead, you have five possible combinations opened up to you, three of which are exclusive to Lucid’s solution:

Example Config
N-Mode: Identical Nvidia Cards
GeForce GTX 285 / GeForce GTX 285
N-Mode: Non-Identical Nvidia Cards
GeForce GTX 285 / GeForce GTX 260
A-Mode: Identical ATI Cards
Radeon HD 5870 / Radeon HD 5870
A-Mode: Non-Identical ATI Cards
Radeon HD 5870 / Radeon HD 4870
X-Mode: Multi-Vendor
Radeon HD 5870 / GeForce GTX 285

So, we have a trio of capabilities you’ve never seen before: N-mode using non-identical Nvidia cards, A-mode using non-identical ATI cards, and X-mode leveraging multi-vendor interoperability, which was a “demo” mode previously, but was recently upgraded to a production status with Lucid’s most recent driver drop.

Achieving The Unachievable

The specifics of how Lucid is able to get GPUs from the same vendor (but with different performance profiles) and GPUs from different vendors working cooperatively is tightly related to the company’s load-balancing algorithms.

There’s a reason why ATI and Nvidia want you to put cards with common performance attributes together.

ATI supports three different modes for displaying your favorite game across a pair of graphics cards: alternate frame rendering, whereby each GPU handles odd or even frames, supertiling mode, which divides the screen into a 32x32 pixel checkerboard of sections rendered alternately by each GPU, and scissor mode, where the screen is split, with one part rendered by GPU 1 and the other rendered by GPU2. ATI’s own product page admits that scissor mode is not as efficient as the company’s other techniques, but works best in OpenGL-based titles. Of course, the supertiling and scissor modes are largely technical additions to CrossFire, since ATI’s own list of best practices suggests programming with alternate frame rendering in mind. Thus, most of the apps you’d run in CrossFire are optimized for this mode anyway.

Nvidia supports two performance modes: split-frame rendering and alternate frame rendering. Split-frame works like ATI’s scissor mode, dividing the frame up to split its workload up between GPUs. And, as with ATI’s implementation, this isn’t as efficient as AFR. Alternate frame rendering functions similarly as well, assigning one card to even frames and the other to odd.

The problem with split-frame/scissor mode is that, while they help alleviate the pixel processing workload, each GPU still has to store the entire frame in its memory, so geometry and (arguably more important) memory bandwidth aren’t helped at all. Meanwhile, tiling is negatively affected by inter-frame dependencies, such as render targets being used in the following frame.

As a result, AFR is most often used. It makes sense, then, that you’d want GPUs with identical performance working on one frame after another. And even when you have this, one frame might take milliseconds longer to render than the one before, resulting in an artifact of multi-GPU configurations referred to as micro-stuttering. Though less perceptible at high speeds, it’s inevitable that I do a story covering CrossFire or SLI and get asked if the hardware in question exhibits signs of micro-stuttering. We’ll go into more depth on this shortly…

Lucid's 1.4 implementation, not driver-dependentLucid's 1.4 implementation, not driver-dependent

…the point is that Lucid did its homework in setting forth Hydra’s minimum requirements and determined it needed to: allow non-identical GPUs to work in the same system, facilitate scalability with more than one card installed, and eliminate the need for over-the-top connectors between cards. And since the company so clearly defines its goals, it becomes particularly easy for us to evaluate where it stands in relation to them today.

According to Lucid, its Hydra engine is able to intercept DirectX/OpenGL calls and, rather than use something like AFR to divvy up the workload, break each frame into “tasks.” A task isn’t necessarily a screen ratio or series of uniform tiles, but could instead be a 3D object, for example—it’s up to the engine to determine which rendering technique to use. Those tasks are then distributed to the installed GPUs (currently limited to two) through the vendor’s driver (which isn’t even aware of Lucid’s software running in front of it, since Hydra employs the standard Device Driver Interface in Windows 7). The completed tasks are returned to what Lucid calls its interop layer, where the frame is composited and sent to the GPU with an attached display.

In order to make this process dynamic—and by that I mean capable of supporting dissimilar GPUs—the Hydra employs a feedback mechanism that evaluates the performance of installed graphics hardware in real-time and load balances accordingly. Thus, in theory, Lucid’s technology circumvents the micro-stuttering issues of alternate frame rendering and resolves the inter-frame dependencies that penalize split-frame rendering.

Create a new thread in the US Reviews comments forum about this subject
This thread is closed for comments
Comment from the forums
    Your comment
  • Maziar
    Nice article,its very good for users for upgrading,because for current SLI/CF you need 2 exact cards but with Lucid you can use different cards as well,but it still needs to be more optimized and has a long way ahead of it,it looks very promising though
  • Von_Matrices
    I'm highly doubtful of the Steam hardware survey. I think it is underestimating the number of multi-GPU systems. I for one am running 4850 crossfire and steam has never detected a multi-GPU system when I was asked for the hardware survey. The 90% NVIDIA SLI seems also seems a little too high to me.
  • Bluescreendeath
    The CPU scores for the 3D vantage tests are way off. You need to turn off PhysX when benchmarking the CPU or it will skewer the results...
  • shubham1401
    Nice concept...
    A long way to go though.
  • Bluescreendeath
    So far, the best scaling has been in Crysis. The 5870/GTX285 combo benchmarks looked very promising.
  • cangelini
    BluescreendeathThe CPU scores for the 3D vantage tests are way off. You need to turn off PhysX when benchmarking the CPU or it will skewer the results...

    It's explained in the analysis ;-)
  • kravmaga
    "But when you spend $350 on a motherboard, you’re using graphics cards that cost more than that. If you’re not, you aren’t doing it right"

    Quoted from the last page; I disagree with that statement.
    There are plenty of people in situations where using this board is a better investment performance per dollars. This is all the more relevant as this technology will undoubtedly find its way into cheaper boards and budget oriented setups where it will make all the sense in the world to bench it using mid-end value parts.

    I, for one, would have liked to see what using gtx260s and 5770s would look like in this same setup. As is, this review leaves many questions unanswered.
  • SpadeM
    Well the review does give an answer in the form of: It's better to run a ATI card for rendering and a nvidia card for physics and cuda (if u're into transcoding/accelerating with coreavc etc.) with windows 7 installed.
    Or at least that is the conclusion i'm comfortable with at the moment.
  • HalfHuman
    i also agree with the fact that a person who will buy this board will necessarily go for the highest priced vid boards. maybe some will but not all. there will be more who will try to save the older vid cards.

    i also understand why you paired the 5870 with nvidia's greatest. there is a catch however... lucid guys did not have the chance to play with 5xxx series too much and you may be evaluating something that is not too ripe. i guess the 4xxx series would have been a better chance to see how well the technology works. couple that with games that are not yet certified for lucid, couple that with how much complexity this technology has to overcome... i think this is a magnificent accomplishment o lucid team part.

    i also think that in order for this technology to become viable it will go down in price and will be found in much cheaper boards. for the moment the "experimenting phase" is done on the expensive spectrum. i saw some early comparisons and the scaling was beautiful. i know that the system put together by lucid... but that is fine since that was only a demo to show that it works. judging on how fast this guys are evolving i guess that they will go mainstream this year.
  • cangelini
    Ah, but if it doesn't offer a better investment in performance per dollar, as is the case now, that statement stands up =)
  • Andraxxus
    I hope that the guys at Lucid will have a chance to continue with this
    wonderful technology.Not long ago mixing ATI with Nvidia was unthinkable
    and many people asked if they could CF or SLI mixed boards on forums. So I think that this is something that should have the support of the people
    that buy GPUs so that we can end this proprietary technology farce (see Physx).I'm not saying that the Physx is bad but the restriction are bad.
    Well in the end I just hope that they won't be bought by a rich so called
    "competitor" that will can the product so that it can keep sucking money
    from the buyer just for minor improvements or rebranding.
  • juanc
    I think that this will really pay if the people develop some driver that can "get the most out of each card" by rendering using each cards "best features" like for example, render the 3D Scene with the GeForce and apply the AA with the ATI and the colouring with the ATI. Balancing using what's best on each card.

    Then I'll get 1 middle of the pack ATI and one middle of the pack NV. Run what runs best on each, or combine the best features of each card togheter.
  • Yuka
    Nice review Mr. Chris, sharp as usual.

    I agree with zipzoomflyhigh, but this chip has a lot of pontential. It needs some polishing or help from ATI and nVidia to make it better. If they can make it some how (ATI and nVidia for Hydra), this would boost up their sales for not being "platform bound" and leasing their multi gpu tech to third parties. I can dream a little, right? lol.

    Anyway, very good news and hopefully nVidia nor ATI will bully this tech.

  • socrates047
    Nvidia has 'x', AMD/ATI has 'y', and Intel brings 'z'.
    Hydra produces 'xyz'.
    this is all value to me... i dont know about you guys.
  • thackstonns
    Here is why I like this technology. I can keep my 4870 and upgrade to a multicard system without having to buy 2 more graphics cards. So I could do a 5870 and instead of moving the 4870 to a different computer I can keep it. Here is where I have that problem though. Physics will suck because of nvidia's restrictions, I will have a hell of a system that will run crysis and looks good, but since the rest of the games are console ports I will be wasting money to play crap quality games.
  • noob2222
    Nice read, but I question the actual useable titles with this Hydra. Testing with games that aren't supported doesn't show what the board can do, but only shows what it can't.
    Using 5/6 titles that aren't supported officially makes this board and technology appear to be an epic failure. Would be nice to know what it does when the game is acually supposed to work, or what happens when the drivers allow these games to work in the future.
  • TeraMedia
    The problem I have with the product is that they are essentially replacing the GPU obsolescence schedule with the chipset obsolescence schedule. And their platform choice makes this particularly bad because while AMD makes an effort to keep their sockets backward-compatible, Intel seems to do the opposite. In fact, Intel now seems hell-bent on segmenting the platform space as much as possible while constraining the product lifecycle as much as possible. Want to reuse your C2Q or upgrade to a 6-core (gulftown, is it?) CPU on this mobo? Good luck with that. With socket 1156, Intel has effectively forced you to buy a new mid-range CPU and constrained you to the mid-range market. If past behavior is any predictor of future behavior, I fully expect the next major generation of Intel CPUs (e.g. 3+ yrs out) not to be compatible with 1156. How long do you think Intel will make advancements on 1156-compatible CPUs?

    So, yes, you can mix GPUs from different generations and even from different vendors. But by the time it even makes sense to do that twice, you'll need to upgrade your whole MB to keep a balanced CPU-GPU system. If the X-mode, A-mode and N-mode scaling were more seamless and effective on the latest HW, and the cost were more in-line with other 1156-socket MBs, I could see this MB making some sense. But given that you need to spend an extra $150+ for this Mobo, I'd rather put that $150 towards the second card or an upgraded card with a longer life span before obsolescence.
  • memeroot
    big fan of the concept and $150 isn't to much for something a bit fun....
    however needs to be x58 and what is the over clocking ability of the board?
    also does it have the same audio advantages
  • xer0
    So what happens when Nvidia (which already has with Physx) or ATI decide to to make drivers (or even firmware) that looks for the competitors's (or lower-end, same-manufacturer cards) and says "Sorry we're being douchebags and turning off functionality and performance features."