SciEngines Products and Services

SciEngines HPC solutions

SciEngines is a young innovation-driven company located in Kiel, Germany. The company was established in 2007 on the vision that we can satisfy the computational needs in science and engineering.

To do so, SciEngines offers a number of reconfigurable computing (RC) products, ranging from PCI express accelerator cards with 4-8 chips to large-scale FPGA clusters with dozens, hundreds or thousands of FPGAs for high performance computing (HPC) applications. Our architectures provide a maximum in computing ressources, and adjustments can be made to emphasize raw processing power, data-throughput or memory.

Our current flagship products in the area of reconfigurable HPC are:

COPACOBANA - 120 Spartan-3 1000 or Virtex-4 SX35 FPGA computer
RIVYERA - 16 to 128 Spartan-3 5000 FPGA cluster

These products greatly accelerate computation in a variety of fields, including biosciences, cryptography, engineering, financial mathematics, image & signal processing, simulations in physics or logistics, etc. while helping to save money due to an energy efficiency that is 100 times better than what competitive technologies offer.

Should you have specific hardware requirements that aren't fulfilled by our standard, commercial off the shelf (COTS) products, please let us know and we will be happy to provide our custom hardware design services. In case that you are interested in some specific algorithms that is not yet ported to our machines, we are also happy to provide algorithm optimization as well as implementation support. Please refer to our products, solutions and services pages, or contact us by This e-mail address is being protected from spambots. You need JavaScript enabled to view it for further information.

LATEST NEWS

Growing with your application: RIVYERA-UPLINK

RIVYERA-UPLINK, is a full backplane bandwidth machine-interconnection feature, which is shipping with all RIVYERA. UPLINK allows a logical and physical expansion of your machine without the need to modify your algorithms.

Ever thought about a 512 FPGA CLUSTER or more? With less than 2.5 kW power consumption and (for specific algorithms) more than 75'000 CPUs performance? No problem, just directly connect 4 RIVYERA through UPLINK and you get one 12 HU machine with full backplane bandwidths from the first to the last FPGA.