Lucid HYDRA Multi-GPU Technology Launches - End of SLI and CrossFire?

Manufacturer: General

Lucid makes quite a showing


Even more impressive though, is the demonstrated ability that the HYDRA technology has to combine the performance of an AMD and NVIDIA card in the same system.  This is truly the killer feature that will make you want, no NEED, a HYDRA-based motherboard very soon.

But first, a quick refresher from our preview of the HYDRA technology last year:

What is the HYDRA Engine?

At its most basic level the HYDRA Engine is an attempt to build a
completely GPU-independent graphics scaling technology - imagine having
NVIDIA graphics cards from the GeForce 6600 to the GTX 280 working
together with little to no software overhead with nearly linear
performance scaling.  HYDRA uses both software and hardware designed by
Lucid to improve gaming performance seamlessly to the application and
graphics cards themselves and uses dedicated hardware logic to balance
graphics information between the CPU and GPUs.

Why does Lucid feel the traditional methods that NVIDIA and AMD/ATI
have been implementing are not up to the challenge?  The two primary
multi-GPU rendering modes that both companies use are split frame
rendering and alternate frame rendering.  Lucid challenges that both
have significant pitfalls that their HYDRA Engine technology can
correct.  For split frame rendering the down side is the need for all
GPUs to replicate ALL the texture and geometry data and thus memory
bandwidth and geometry shader limitations of a single GPU remain.  For
alternate frame rendering the drawback is latency introduced by
alternating frames between X GPUs and latency required for inter-frame
dependency resolution.

How it Works

HYDRA is a dedicated silicon with sole purpose of scaling GPUs. 
Though there is no graphics processing logic on the HYDRA chip, what
the chip can do is redistribute graphics workloads across multiple GPUs
in real-time.  The HYDRA technology also includes a unique software
driver that rests between the DirectX architecture and the GPU vendor

The distribution engine as it is called is responsible for
reading the information passed from the game or application to DirectX
before it gets to the NVIDIA or AMD drivers.  There the engine breaks
up the various blocks of information into "tasks" - a task is a
specific job that HYDRA defines that can be passed to any of the 2-4
GPUs in the system.  A task might be something like a specific lighting
effect, a post processing run, a specific model being drawn, etc.  The
company founders on hand at the meeting were a little vague about the
algorithms that decide how, and what parts, of the DirectX data are
going to be defined as "tasks" - it is obvious that this is part of the
magic that gives HYDRA its power; it is with these task definitions
that the hardware logic can efficiently distribute the work load across
many GPUs.

Once the tasks have been created, they are then sent over the PCI
Express bus to the HYDRA chip where they are VERY quickly processed and
split between 2 to 4 GPUs.  The HYDRA Engine passes off these tasks to
the GPU, awaits a result and return of finished data or pixels, and is
then responsible for passing that information on to one of the GPUs for
final output to a monitor.  At the outset, this doesn't sound that much
different than what NVIDIA and AMD already do with AFR and SFR
rendering modes, but after seeing the HYDRA technology at work it is
obviously something very different.

By essentially intercepting the DirectX calls from the game to the
graphics cards, the HYDRA Engine is able to intelligently break up the
rendering workload rather than just "brute-forcing" alternate frames or
split frames as both GPU vendors are doing today in SLI and CrossFire. 
And according to Lucid all of this is done with virtually no CPU
overhead and no latency compared to standard single GPU rendering. 

To accompany this ability to intelligently divide up the graphics
workload, Lucid is offering up scaling between GPUs of any KIND within
a brand (only ATI with ATI, NVIDIA with NVIDIA) and the ability to load
balance GPUs based on performance and other criteria.  The load
balancing is based on a couple of key data points: pre-existing
knowledge from the Lucid team about the GPU in question and the
"response time" of the GPU when being sent data from the HYDRA Engine
chip.  The HYDRA driver will actually recognize the GPUs in a system
and will estimate how much processing power each holds but will then
fine tune that estimate based on real-time performance of the GPU in
action.  If a GPU is sent a "task" to perform and the return time on it
is slower than expected, the HYDRA engine will back off slightly and
send more "tasks" to the less-loaded GPUs.  All of this is updated on
the fly, in real time as the game is running.

With the ability to divide up the graphics processing into tasks
and monitor GPU load, the HYDRA engine offers up some very interesting
types of GPU scaling.  Yes, it can and will sometimes run standard
split frame rendering - this is very efficient for per-pixel processing
heavy workloads.  However, the HYDRA can also implement some much more
interesting divisions of work.

Click to Enlarge

Click to Enlarge

What you are seeing above is two monitors, each displaying the
workload of a single GPU in a dual 9800 GT configuration.  This of
course wouldn't normally be how a game is presented to users - it is
simply a way to demonstrate their technology to the press.  Before the
images are merged again you'll see two very different screens - one has
some rendered areas of a level of UT3 with some areas completely in
black while the other monitor has the inverse - opposite rendered areas
and black areas. 

Click to Enlarge - merged image on the left, half of the rendered image on the right; notice no floor on the right, etc.

This is the power of the task-driven graphics workload division. 
You can clearly see that some of the "items" in the world are being
rendered by one GPU while the background and trees by another.  There
is no rigid requirement of certain size or shapes of divisions and thus
many of problems found in "box rendering" are avoided.  If Lucid is to
be believed, this is the division of tasks that are about "even" in
required rendering power.  Lucid pointed out there are many other split
rendering methods that it uses in the background for games but that
demonstrating them to the end users in any way is pretty difficult to
do - techniques like blending pixels in a frame buffer for example. 

Even more importantly though is that this rendering method is NOT
predefined by any driver profile as with NVIDIA's and AMD's SLI and
CrossFire technology.  Instead, because of HYDRA Engine's
pre-processing work, the rendering method can and will change
throughout the game and sometimes even inside of individual frames. 
The HYDRA chip itself does some of this algorithmic work with help from
the driver and task setup process. 

What is new for 2009?

So, at a glance, what is new and being announced this year at IDF?  In short - the actual product you can buy.  Lucid and MSI are teaming to announce the "Big Bang" motherboard that sports the HYDRA 200 chip and three fully functional PCIe slots for multi-GPU gaming.  We are told this board will be available sometime in the next 30 days though pricing is still in the air. 

MSI's upcoming "Big Bang" motherboard

We have a great video of about 20 minutes in length that looks over this new motherboard, shows a couple of Lucid's demos on working systems and talks with the co-founder Offir Remez about Lucid's technology and plans for the future.

On the "Big Bang" motherboard the Lucid HYDRA 200 and P55 chipset sit side by side.

Obviously there is one thing missing from this video and article:
performance results. Lucid did let us see the HYDRA system in action on
the combined AMD/NVIDIA system but the company is not releasing
performance numbers at this time.  Based on what I saw last year as compared to this year, I am confident that it will achieve competitive gaming performance. 
(Sorry we can't say more, they are really tying our hands here.)  We
will have to reserve our final judgment when we test HYDRA 200 in our
labs, but if the idea of "universal GPU scaling" interests you, you'll
want to keep checking back for more information on this motherboard
very soon.

This is the most important system running both an AMD Radeon HD 4890 and GeForce GTX 260 and scaling!

There are still some lingering questions I have for the HYDRA technology: are there stuttering issues?  Will day one scaling for new gaming titles be high enough?  We will be getting our answers very soon.  For now though, this is yet another reason to be excited about PC gaming!