The VectorBlox MXP™ Matrix Processor™ is a scalable soft-core processor designed for FPGAs. It implements classic massively parallel vector processor algorithms traditionally used in scientific super-computers. In addition to 1D vectors, the MXP™ processor also operates on 2D and 3D matrices, increasing the application kernels it can enhance. Programming is purely done in C/C++, combining high-performance with easy programmability.
The MXP™ enhances the performance of standard Xilinx MicroBlaze processors by a minimum of 1 to 2 orders of magnitude in many applications. It is optimized to fit into a standard FPGA as a plug-in IP block. Users write standard C programs augmented with a set of optimized MXP™-specific extensions. These extensions are easy to learn and represent typical operations like vector-add or vector-multiply.
- Accelerates time to market
- Only need C/C++ programming skills – No RTL or HW skills required
- Decouples HW and SW design
- HW speeds delivered in software
- Up to 1000x enhancement to standard MicroBlaze
- Supports Xilinx XPS integration tools
- Enhances standard MicroBlaze processors
- Scalable from V1 to V128 lanes
- Predesigned logic description of the vector processor, available as a plug-in IP module
- C compiler with MXP™ extensions and libraries
- Optional functional simulator; run and debug MXP™ C programs on any computer system
- Machine Vision
- Imaging
- Smart Camera
- Industrial Control
- Financial
- Encrypted RTL
- Functional Simulator (optional)
- C Extensions and Libraries for MXP™
- User Documentation
- Sample Code
Available Formats
MXP™ Matrix Processor™ is available in the following configurations:
| Description | Target FPGA | Fabric | Scalability |
|---|---|---|---|
| MXP-M-Fixed | Xilinx – MicroBlaze | FSL + AXI3 or AXI4 | Fixed size, e.g. V8 |
| MXP-M-Configurable | Configurable size, e.g. V1 to V128 |