Upgrading to CUDA FDTD

Acceleware Releases New CUDA Libraries Delivering Increase in Performance

With the new CUDA Library you now can:

• Run Faster Simulations! - Over 2X faster when going from Quadro 5600 to Tesla C1060 + Acceleware FDTD for CUDA.
• Run Bigger Simulations! - Full featured simulations for over 100 MCells on a single card.
• Save Money! - Tesla C1060 is 1/3 the price of the equivalent Quadro FX 5800.
• Do More With Less! - Run the same simulation on half as much hardware at the same speed.
• Windows + Tesla! – Full support of Tesla Products under Windows.

Feature Webinar: CUDA for FDTD (Length: 10 minutes)

Acceleware introduces a new CUDA™ based solution for accelerating EM simulations using FDTD. The new software library has delivered application performance gains of up to 50% compared to the current OpenGL version. The graphs below show how significant the speed ups with the new CUDA based library on the NVIDIA 10-series are:

Save Time	Save Money Libraries

Hover over images above for a larger view.

How to Upgrade?

Gaining the above mentioned performance boost for your simulations is not hard! The images and upgrade notes below show what you need to take advantage of the new CUDA based FDTD library from Acceleware.

Accelerator A30 to NVIDIA Tesla C1060

* ... NVIDIA Quadro NVS 295 needed for graphic output

Upgrade Notes:

• C1060 has no graphics output
• Some workstations will only support one Quadro 5600 / Tesla C1060 due to limited power supply
• Workstations running 5600's should also be able to run C1060
  • See our Certified Systems page for more information
  • See this NVIDIA web page for more information on NVIDIA Tesla C1060
• Free library upgrades available for current maintenance packages
• Ask your partner/Acceleware for details on the 9.0.0 CUDA library

Accelerator A20/D20 to NVIDIA Tesla C1060

* ... NVIDIA Quadro NVS 295 needed for graphic output

Upgrade Notes:

• It is strongly recommended to upgrade your workstation to support your Tesla C1060 card
• Click here for Acceleware's Certified Systems or visit NVIDIA's Tesla Support Systems

D30 to NVIDIA Quadro Plex Model D2