: News
: Reviews
: Editorials

: Archives
: About
: Contact
: Advertising
: Privacy
: Links
: Forums

 

 

 

Tesla Computing

 

Putting those FLOPS to PRODUCTIVE Use

 

by Josh Walrath

 

            The promise of GPGPU has been covered for the past several years, and with the introduction of programmable graphics processors, that dream was finally able to come to fruition.  The previous generation of DX9 cards dipped their toes into the general computing world, but the architectures were still very oriented towards graphics and fixed function processing.  ATI had a bit of a better time of it due to their thread management, programmable memory controller, and better branching performance with the X1800 and X1900 series, but the architecture was still rather limited in what it could do.  NVIDIA was a bit farther behind with their 7900 series of cards, and even though some basic GPGPU work was done with them, they simply were not flexible enough to handle complex programs and calculations.

            Fast forward to November of last year and we see NVIDIA release a radical new architecture which excels not only in its flexibility, but also overall performance.  The 8800 series of GPUs are the first to support Microsoft’s DirectX 10 and Shader Model 4.0, but these chips are also designed from the outset to be an outstanding performer in general purpose single precision floating point operations.  Seemingly the price for next generation performance in games, DX10 compatibility, and GPGPU support is a huge one.  The G80 chip the 8800 series is based on is comprised of 680+ million transistors packed into a die that is supposedly around 460 mm square.

            Another price that has to be paid is power and heat.  Currently the top end 8800 GTX/Ultra boards consume around 150 to 175 watts at full bore.  They also require a pretty hefty system to whisk the heat away.  NVIDIA uses the older 90 nm process from TSMC with their G80 chips, and they use the newer 80 nm optimized process for their lower end G84 and G86 chips.  But when considering the alternatives for achieving these types of performance numbers, the G80 chips look incredibly cheap from a price/performance/power standpoint.  It would take quite a few dual core/quad core X86 CPUs to match the theoretical output of the G80 in single precision floating point operations. 

Tesla

            NVIDIA is branding their GPGPU products as the Tesla family.  For the first generation of these products we will see three distinct offerings all based upon the G80 processor on a GTX/Ultra PCB, but without the video outputs.  There are three levels of products that NVIDIA will offer.  Initially NVIDIA will be the only company actually selling these products, but perhaps in the future, depending on how successful the Tesla initiative is, we may see other partners taking over responsibility for production of the actual Tesla cards, personal supercomputers, and rackmount servers.

            The most basic product is just a single Tesla board.  This is virtually identical to the above mentioned 8800 GTX/Ultra, but without the output ports as well as double the memory (1.536 GB).  This card is designed to be plugged into a workstation and be used as a co-processing board using GPGPU software.  The initial cost of Tesla C870 is around $1,499.  These boards are specifically designed for GPGPU use, so there potentially could be some significant differences between the $1,499 board and a $529 8800 GTX other than double the memory.  The board itself has a max power rating of 170 watts and can output (in specific situations) 500 gigaFLOPS per GPU. 

            The second tier is that of the “personal supercomputer” which is a box containing 2 x C870 boards.  This product is branded the Tesla D870.  Inside the box it contains 2 x C870 boards, an internal power supply, and the PCI-E switch necessary to allow the box to communicate with the workstation it is attached to.  In the workstation there is installed a PCI-E switch card that is inserted into the PCI-E 16X slot.  An external PCI-Express cable goes from this card to the supercomputing box where it attaches to the internal PCI-E switch that NVIDIA developed (or at least utilizes).  This switch is then connected to the individual cards.  These are all PCI-Express Gen 2 products, so their throughput is pretty impressive (16 GB/sec aggregate bandwidth).  The box consumes around 500 watts at maximum, which is more than double the individual power ratings of the cards contained.  This again is due to other internal components such as the switch and cooling solutions.  This little monster will set a user back $7,500.  For average people, this is a lot of money.  For companies who work in the gas and energy field, this is chump change and well worth the cost in terms of space, power consumption, and overall computational performance in one small package.

            The third tier is that of the GPU Computing Server.  This 1U rackmount product packs in 4 x C870 boards and is branded the Tesla S870.  This unit typically consumes around 550 watts with a max power rating of 800 watts.  In certain circumstances, this setup could produce around 2 gigaFLOPS.  The amount of single precision computing power this unit has is tremendous.  This rackmount connects to another server the same way as the personal supercomputer does above utilizing PCI-Express Gen 2 cables and switches.  These racks are designed to be integrated into larger stacks, and be connected with other 2P/4P/8P servers which would handle the rest of the computational requirements.  These racks are not standalone, and do need to be connected to another server which will handle all the OS, networking, and storage duties.

            NVIDIA eventually plans to have rackmounts which support up to 8 GPGPUs, but that does not appear to be an option at this time.  If one of the master servers has 2 x 16X PCI-E slots, then it can theoretically connect to two of the S870s.

            In terms of hardware NVIDIA appears to have all the bases covered for a comprehensive lineup of GPGPU products.

 

Next:  The Software and Future

 

If you have found this article interesting or a great help, please donate to this site.

 

Copyright 1999-2007 PenStar Systems, LLC.