Slide 1 Slide 2 Slide 3 Slide 4 Slide 5 Slide 6 Slide 7 Slide 8 Slide 9 Slide 10 Slide 11 Slide 12 Slide 13 Slide 14 Slide 15 Slide 16 Slide 17 Slide 18 Slide 19 Slide 20 Slide 21 Product List
Image of MoSys Accelerator Bandwidth Engine: High Performance/High Bandwidth - BE3 RMW

An even larger benefit can be achieved with an RMW operation. In a system that performs these operations and does not have it resident in the memory, it essentially requires at least two command cycles: a read of the initial data that needs to be modified and a write to re-deposit the data back into the memory after the host has modified the previously read data. In placing the ALU logic on the memory die there is significant savings to the control of passing data between the host and memory multiple times as well as freeing up the host to perform other functions while the memory is modifying the data. To move functions like the ALU (and Burst) onto the memory die has a valuable upside potential and minimal impact on die resources.

PTM Published on: 2021-08-11