By acquiring the lock on every cell iteration for every program instance, we inherently are introducing a lot of serial execution into the program, creating a bottleneck that severely inhibits performance.