Design Services and Partnerships

 

Apogee understands the added value of partnerships and can assist in the following areas:

Digital Signal Processing

Hardware and System Design

Algorithm Conversion

Digital Signal Processing (DSP)

We are DSP experts and have core capabilities in the following areas:

  • DSP Filter Design
  • Multirate Signal Processing
  • Digital Down Conversion, Upconversion
  • Wideband Spectrum Signals Processing
  • MIMO Filters
  • Rate Conversion
  • Real Time DSP
  • Digital Resampling
  • Digital IF Data Processing

Hardware Design, System Design, FPGA Development, Ethernet Processing

  • System Design
  • Hardware Design, Development, and Manufacture
  • Prototype Development
  • High Speed, High Density, Digital and Mixed-Signal Circuit Board and Electrical Design
  • Schematic Entry - Mentor Graphics DxDesigner, Cadence Orcad
  • Ethernet Packet Processing Applications
  • 10 GbE Ethernet, UDP, VITA 49
  • VHDL Coding, Simulation, Test
  • Xilinx / Altera Firmware
  • FPGA Developer Support
  • FPGA IP Cores
  • Sensor Acquisition
  • Li-ion Battery Powered Solutions

Algorithm Conversion

  • GPU to FPGA (HDL)
  • Matlab to FPGA (HDL)
  • Software (C, Python, OpenCL) to FPGA (HDL)

In today's development environment algorithms are typically developed in computer programs that provide tailored tools for quickly developing algorithms, analyzing data and exploring multiple approaches.   These tools allow a solution that demonstrates the algorithm to be quickly developed faster than writing in traditional languages such as C/C++ where extra effect must be exerted to properly manage details like memory allocation and garbage collection.  Also these tools tend to support more mathematical functions than are available in middle and lower level languages.    The downside to using such tools is the performance related to speed of processing is easily degraded and can take significant amount of work to optimize for high throughput and low memory usage.  Also they are usually ran on larger processors and not supported on smaller processors.   This limits the performance for embedded size applications demanding low power and small physical sizes.

Apogee’s algorithm service provides the critical knowledge base to take your prototype, high-level algorithm and develop a high-performance embedded solution.  Apogee will analyze your current algorithm and product goals (e.g. physical shape, portability, power, weight, and price) and develop a processor architecture that runs the algorithms and meets or exceeds those goals.

Possible Use Cases for Algorithm Design

  • GPP General Purpose Processor and GPU algorithm to FPGA with ARM processor for power and size reductions.
  • A MATLAB algorithm shown to work but needs high bandwidth, real-time implementation.
  • Algorithm with high memory bandwidths bottlenecked by GPP memory architecture.
  • Final product requires smaller physical size and battery based design, in which current smaller processors aren’t supported by high-level tools.
  • Required temperature ranges are too great for standard air cooled GPP based systems.  The final implementation will need power reduction and more efficient implementation.

 

Please contact Apogee for more information on Design Services.