| Per Loop | Per Instance | Per Iteration | Float AI | |
| L1, GB | ||||
| L2, GB | ||||
| L3, GB | ||||
| L4, GB | ||||
| DRAM, GB | ||||
| Self bandwidth, GB/s | Utilization, % | Hardware Peak, GB/s | ||
| L1 | ||||
| L2 | ||||
| L3 | ||||
| L4 | ||||
| DRAM | ||||
| No data. Suggestion: Collect Memory Traffic data that includes callstacks. | ||||
| Per Loop | Per Instance | Per Iteration | Float AI | |
| L1, GB | ||||
| L2, GB | ||||
| L3, GB | ||||
| L4, GB | ||||
| DRAM, GB | ||||
| Total bandwidth, GB/s | Utilization, % | Hardware Peak, GB/s | ||
| L1 | ||||
| L2 | ||||
| L3 | ||||
| L4 | ||||
| DRAM | ||||