Parallel Computing

Due to the slowing of Moore’s Law and the widespread use of multi-core microprocessors, parallel computing is becoming an increasingly important technique used to achieve better performance. In essence,, parallel computing divides large problems into smaller ones and solves them concurrently. But memory management and distributed message passing pose a problem in transiting a linear algorithm into a parallel one.

At Ion Machines, we develop solutions that reason about system topologies and data movement. Dynamic analysis of data movement and resource usage permits us to make the overall system more efficient and smarter. We are especially proficient in shared memory architectures and in using parallel libraries to the fullest extent possible.

Learn More

Kalman Filtering
Go Back