Available for pre-order
View Purchasing OptionsProject update 3 of 5
We have a fantastic update for you! FPGAs, when optimized at a granular level for specific algorithms, can deliver performance far beyond traditional CPU-based solutions. With their efficiency and parallelism, FPGAs are game-changers for accelerators, and we’re here to prove it.
We’ve developed a custom SHA-256 FPGA IP running at 100MHz, executing 66 clock cycles per hash, with 37 parallel instances. This results in a staggering 57.6 MH/s for SHA-256-based applications.
In comparison, a highly optimized Raspberry Pi 5 reaches a maximum of 3 MH/s—meaning we deliver 15x better performance for only 2x the price!
But here’s what makes it even more impressive: the entire system operates at just 12W, making it remarkably power efficient. While CPUs struggle with power-hungry hashing operations, our optimized FPGA-based approach delivers breakthrough efficiency without compromising speed.
At the heart of our system is Vaaman, featuring a 6-core ARM processor running Linux. We utilize Rah, our custom interconnect protocol, to establish high-speed communication between the CPU and FPGA.
For benchmarking, we implemented a standard Bitcoin mining algorithm on our platform. On the FPGA side, our highly optimized SHA-256 IP continuously processes hashes in parallel. Once a valid hash block is found, it is seamlessly transferred back to the CPU—delivering unprecedented efficiency in cryptographic computations.
For every Vaaman user, we are providing an easy-to-use framework to benchmark this accelerator and integrate SHA-256 into their own applications.
With simple APIs (available in C/C++ and Python), users can harness FPGA-accelerated cryptographic processing without complex development overhead.
Stay tuned as we continue to push the boundaries of reconfigurable computing, FPGA acceleration, and high-performance cryptographic processing.
This is just the beginning!