The FirePro Products get Cayman

The AMD FirePro professional graphics lineup is getting access to the new Cayman-class GPU.


(Thanks to Steve Grever for providing insight on the product placement and attending the Professional Graphics Editor’s Day in Austin, TX for us!)

On May 11, AMD invited a handful of technology journalists and hardware reviewers to Northern Islands FirePro Tech Day to unveil a pair of new professional graphics cards – the V5900 and V7900. We were under an NDA to discuss the new GPUs at that time, but now that the gag order has been lifted, we can finally give our readers an in-depth look at these mid-range and high-end graphics card offerings sporting custom features like Eyefinity, Geometry Boost and Power Tune technologies.

Sandeep Gupte, AMD’s product management director, introduced the new graphics cards during the one-day event and stated they will “deliver productivity and performance to professionals regardless of where they are working.” This is an interesting statement, but AMD is committed to providing graphics solutions beyond professional workstations to include mobile workstations, tablets and thin clients to increase productivity and performance across various form factors and operating systems.

Last year’s FirePro lineup helped AMD increase their unit share by six points in the professional graphics market. This share increase puts them at about 16 percent overall, which was also supported by sales with Tier 1 OEMs like HP and Dell. This percentage of market share has improved over the single-digit shares AMD experienced in this market back in 2007.

Hit that "Read More" link below for the full story!

The new V5900 is AMD’s mid-range solution that replaces the V5800 and includes 512 stream processors and supports three simultaneous outputs via one DVI and two DisplayPorts. AMD also doubled the frame buffer to 2GBs of GDDR5 memory, which puts this card squarely in the mid-range class of their FirePro product line. Another interesting feature to note is the V5900 doesn’t require any additional power from a six-pin power connector other than the 75 watts that it pulls from the PCI Express 1.1 slot. This single-slot graphics card retails for $599 MSRP.

(Editor’s Note: It should be pointed out here that both the V5900 and the V7900 FirePro cards are based on the Cayman architecture, first unveiled by AMD with the release of the Radeon HD 6900 series of graphics cards.  This move is a pretty interesting shift in design including a transition from a VLIW5 to a VLIW4 stream processor that can reach higher utilization.  This allows AMD to provide similar performance using ~10% less die area and that can translate into smaller, cheaper or even faster GPUs.  For all the details about the Cayman architecture, be sure to check out our initial review of it.)

The latest addition to AMD’s high-end class of graphics cards is the V7900. The V7900 also includes 2GBs of GDDR5 memory like the V5900, but AMD packed the GPU with 1,280 stream processors! This card also supports Framelock/Genlock (with an S400 module) and Stereo 3D. Another key selling point for the V7900 is that it supports four simultaneous outputs through four DisplayPorts, which will be optimal for Autodesk workstations that want quad high-definition monitors at their disposal. The V7900 is also a single-slot graphics card and retails for $999 MSRP.

<Editor’s Addition> What you might find particularly interesting is that with this new GPU architecture comes some intriguing specification changes.

In this table that compares the V7900 to the V7800 it shows that the "new" card has fewer shader processors as well as a lower peak compute rate on single precision data.  However, double precision performance and memory bandwidth INCREASE leading us towards the path of better professional graphics performance rather than any type of relative gaming performance increase.  Of course you have the feature improvements as well, mentioned above and below.

With the V5900 both single and double precision calculation performance drops and memory bandwidth remains static indicating that most (if not all) performance improvement is a result of the doubling of the frame buffer.  </Editor’s Addition>

Eyefinity, Geometry Boost, Power Tune technologies

With their new V5900 and V7900 professional graphics cards hitting the streets today, AMD focused on three “pillars” where they felt their customer base could benefit from the extra horsepower and advancements their new GPUs offer. These pillars include giving their customers the “edge” they are looking for their business by providing them with leading-edge performance and reliable hardware and optimized drivers.

Their next-gen GPUs all support a host of new technologies like Geometry Boost and AMD PowerTune, but they also have a few more mature innovations like Eyefinity and DisplayPort 1.2 support. We’ll cover each one of these technologies in more detail below.



To help give their customers a competitive edge, AMD introduced Eyefinity to the professional graphics marketplace last year by including this technology with the V5800 and V7800. Eyefinity has been a massive success for AMD, mainly because users can enjoy triple-monitor goodness through one graphics card as opposed to NVIDIA’s Surround technology, which requires two graphics cards to run three monitors simultaneously.

Not only will professionals save money by using one graphics card for triple-monitor workstations, but they will also enjoy less power consumption versus dual-card NVIDIA solutions. Eyefinity also has the eye of professional users because of its support for DirectX 11, OpenGL and OpenCL.


Geometry Boost

AMD’s new Geometry Boost technology is another exclusive feature that
Essentially allows the GPU to process two primitives per clock cycle. This is a huge improvement from previous generations that could only process one primitive per clock cycle. In the professional graphics realm, Geometry Boost could have a huge impact on real world modeling with CAD and DCC, which are geometry intensive with lots of triangles that have to be processed. Doubling the primitive processing rate should yield higher performance numbers when used on medium to large models as well.


PowerTune technology

AMD’s PowerTune technology helps optimize performance for TDP-constrained GPUs. This new tech helps the V5900 and V7900 deliver higher performance that is optimized to the thermal limits of the GPU by dynamically adjusting the clock during runtime based on an internally calculated GPU power assessment.

Eric Demers, AMD’s chief technology officer of their graphics division, said it also improves the mechanism to deal with applications that would otherwise exceed the GPU’s TDP.

“By dynamically managing the engine’s clock speeds based on calculations, which determine the proximity of the GPU to its TDP limit, AMD PowerTune allows the GPU to run at higher clock speeds,” said Demers.

PowerTune is very different from existing GPU power-optimization methods because it dynamically adjusts the performance profile in real time to fit within the TDP envelope. Other solutions typically set the highest state GPU clock speeds based on a worst case TDP approach, which can compromise performance in a majority of applications.

« PreviousNext »