AMD FirePro W9100 Announced: Doing Work in Hawaii.

Subject: General Tech, Graphics Cards | March 26, 2014 - 05:43 PM |
Tagged: amd, firepro, W9100

The AMD FirePro W9100 has been announced, bringing the Hawaii architecture to non-gaming markets. First seen in the Radeon R9 series of graphics cards, it has the capacity for 5 TeraFLOPs of single-precision (32-bit) performance and 2 TeraFLOPs of double-precision (64-bit). The card also has 16GB of GDDR5 memory to support it. From the raw numbers, this is slightly more capacity than either the Titan Black or Quadro K6000 in all categories. It will also support six 4K monitors (or three at 60Hz), per card. AMD supports up to four W9100 cards in a single system.

View Full Size

Professional users can be looking for several things in their graphics cards: compute performance (either directly or through licensed software such as Photoshop, Premiere, Blender, Maya, and so forth), several high-resolution monitors (or digital signage units), and/or a lot of graphics performance. The W9100 is basically the top of the stack which covers all three of these requirements.

View Full Size

AMD also announced a system branding initiative called, "AMD FirePro Ultra Workstation". They currently have five launch partners, Supermicro, Boxx, Tarox, Silverdraft, and Versatile Distribution Services, which will have workstations available under this program. The list of components for a "Recommend" certification is: two eight-core 2.6 GHz CPUs, 32GB of RAM, four PCIe 3.0 x16 slots, a 1500W Platinum PSU, and a case with nine expansion slots (to allow four W9100 GPUs along with one SSD or SDI interface card).

View Full Size

Also, while the company has heavily discussed OpenCL in their slide deck, they have not mentioned specific versions. As such, I will assume that the FirePro W9100 supports OpenCL 1.2, like the R9-series, and not OpenCL 2.0 which was ratified back in November. This is still a higher conformance level than NVIDIA, which is at OpenCL 1.1.

Currently no word about pricing or availability.

Source: AMD

March 26, 2014 | 06:32 PM - Posted by Daniel Masterson (not verified)

Hopefully the bitcoiners will go for this and free up the 290X stock.

March 26, 2014 | 06:42 PM - Posted by Scott Michaud

Stock is not really that bad, lately. For instance, you can get the ASUS R9 290X DirectCU II for $599 at Amazon. That's only 30$ over MSRP.

March 26, 2014 | 06:54 PM - Posted by JosiahBradley

You can even get the PowerColor 290X at MSRP:

March 26, 2014 | 06:44 PM - Posted by AMDbumlover (not verified)

doesn't AMD support OCL 2.0 on GCN parts? or is that just kaveri [in that one specific setup?] ?

March 27, 2014 | 01:24 AM - Posted by Scott Michaud

They actually do not support OpenCL 2.0 at all, apparently, at least not as a whole, temporarily. They have drivers which are classified as OpenCL 1.2 but include extensions for some OpenCL 2.0 features. It sounds like Kaveri can be upgraded to OpenCL 2.0 with a driver update at some point, but I have not heard anything about that for their discrete graphics. That said, it is possible that they are developing an OpenCL 2.0 driver for their add-in boards. If they surprise me with one, then awesome.

March 27, 2014 | 12:43 PM - Posted by Anonymous (not verified)

For the price they are charging for pro graphics cards, over standard gaming cards, they should have the latest driver support for all the graphics APIs. The only excuse for not having the support is if it really need hardware support for the API to work properly, and then they should have it at least implemented in software if possable. Pro graphics cards like this should transition to more of a GPU/CPU accelerated type of system with a CPU/s on Die with the GPU for a low latency total graphics system on a PCI card. A large on Die RAM whould also help, as would a pro graphics tuned dedicated OS running on GDDR5(12 GB minium) or better memory, and hosted on the PCI based device. Having a CPU share the same Die or package with the GPU and RAM will enable a descrete device to be able to ray trace and not have to rely solely on the motherboard CPU and its high latency narrow width data bus. Nvidia is developing an entire package on a specilized module, that has metal traces for the chips etched out on silica, with an ultrawide data/ultra fast bus directly connected to the memory/other components, that are placed/stacked on the module and connected to the silica etched metal traces along with GPUs, and CPUs, if the customer wants. expect this technology, adopted for similar IBM technology, to begin to appear in the consumer market also.

March 27, 2014 | 03:47 PM - Posted by Anonymous (not verified)

"is developing" lol

AMD is developing things too don't forget. ;)

Back to the present and reality, W9100 is and will be the undisputed leader in workstation class cards for some time...until AMD's next FirePro that is. A 512 bit bus, 16GB frame buffer, memory stacking for 64GB addressible memory, DPFP, unmatched integer performance, all in a 440 mm chip. NV can't touch that, unless of course we are talking fantasies, buzzwords and ever changing roadmaps....

March 27, 2014 | 03:49 PM - Posted by Anonymous (not verified)

*1/2 rate DPFP. 2.6 TFLOPS peak.

March 27, 2014 | 12:33 AM - Posted by eddie (not verified)

Do those cards really have backplates.. And I would have thought they would have at least let sapphire throw a custom coolers one these things because from what we have seen so far, AMDs coolers suck. Its not like it would add that much onto the already considerable price as these are workstation cards and therefore will come with a workstation price tag.

March 27, 2014 | 09:20 AM - Posted by The Smith (not verified)

The workstation in the picture doesn't conform to even the minimum specs; it has only one CPU!

March 27, 2014 | 12:38 PM - Posted by Anonymous (not verified)

Scott Michaud , definitely one of my favorite PCPer editors.
Fine read as always Scott.

Post new comment

The content of this field is kept private and will not be shown publicly.
  • Lines and paragraphs break automatically.
  • Allowed HTML tags: <a> <em> <strong> <cite> <code> <ul> <ol> <li> <dl> <dt> <dd> <blockquote><p><br>
  • Web page addresses and e-mail addresses turn into links automatically.

More information about formatting options

By submitting this form, you accept the Mollom privacy policy.