Local data share (LDS)

LDS Speed-of-Light

Warning

The theoretical maximum throughput for some metrics in this section are currently computed with the maximum achievable clock frequency, as reported by rocminfo, for an accelerator. This may not be realistic for all workloads.

The LDS speed-of-light chart shows a number of key metrics for the LDS as a comparison with the peak achievable values of those metrics.

CDNA

Metric	Description	Unit
Utilization	Indicates what percent of the kernel’s duration the LDS was actively executing instructions (including, but not limited to, load, store, atomic and HIP’s __shfl operations). Calculated as the ratio of the total number of cycles LDS was active over the total CU cycles.	Percent
Access Rate	Indicates the percentage of SIMDs in the VALU [1] actively issuing LDS instructions, averaged over the lifetime of the kernel. Calculated as the ratio of the total number of cycles spent by the scheduler issuing LDS instructions over the total CU cycles.	Percent
Theoretical Bandwidth Utilization	Indicates the maximum amount of bytes that could have been loaded from, stored to, or atomically updated in the LDS divided as percentage of theoretical peak. Does not take into account the execution mask of the wavefront when the instruction was executed. See the LDS bandwidth example for more detail.	Percent
Bank Conflict Rate	Indicates the percentage of active LDS cycles that were spent servicing bank conflicts. Calculated as the ratio of LDS cycles spent servicing bank conflicts over the number of LDS cycles that would have been required to move the same amount of data in an uncontended access. [2]	Percent

CDNA 2

Metric	Description	Unit
Utilization	Indicates what percent of the kernel’s duration the LDS was actively executing instructions (including, but not limited to, load, store, atomic and HIP’s __shfl operations). Calculated as the ratio of the total number of cycles LDS was active over the total CU cycles.	Percent
Access Rate	Indicates the percentage of SIMDs in the VALU [1] actively issuing LDS instructions, averaged over the lifetime of the kernel. Calculated as the ratio of the total number of cycles spent by the scheduler issuing LDS instructions over the total CU cycles.	Percent
Theoretical Bandwidth Utilization	Indicates the maximum amount of bytes that could have been loaded from, stored to, or atomically updated in the LDS divided as percentage of theoretical peak. Does not take into account the execution mask of the wavefront when the instruction was executed. See the LDS bandwidth example for more detail.	Percent
Bank Conflict Rate	Indicates the percentage of active LDS cycles that were spent servicing bank conflicts. Calculated as the ratio of LDS cycles spent servicing bank conflicts over the number of LDS cycles that would have been required to move the same amount of data in an uncontended access. [2]	Percent

CDNA 3

Metric	Description	Unit
Utilization	Indicates what percent of the kernel’s duration the LDS was actively executing instructions (including, but not limited to, load, store, atomic and HIP’s __shfl operations). Calculated as the ratio of the total number of cycles LDS was active over the total CU cycles.	Percent
Access Rate	Indicates the percentage of SIMDs in the VALU [1] actively issuing LDS instructions, averaged over the lifetime of the kernel. Calculated as the ratio of the total number of cycles spent by the scheduler issuing LDS instructions over the total CU cycles.	Percent
Theoretical Bandwidth Utilization	Indicates the maximum amount of bytes that could have been loaded from, stored to, or atomically updated in the LDS divided as percentage of theoretical peak. Does not take into account the execution mask of the wavefront when the instruction was executed. See the LDS bandwidth example for more detail.	Percent
Bank Conflict Rate	Indicates the percentage of active LDS cycles that were spent servicing bank conflicts. Calculated as the ratio of LDS cycles spent servicing bank conflicts over the number of LDS cycles that would have been required to move the same amount of data in an uncontended access. [2]	Percent

CDNA 4

Metric	Description	Unit
Utilization	Indicates what percent of the kernel’s duration the LDS was actively executing instructions (including, but not limited to, load, store, atomic and HIP’s __shfl operations). Calculated as the ratio of the total number of cycles LDS was active over the total CU cycles.	Percent
Access Rate	Indicates the percentage of SIMDs in the VALU [1] actively issuing LDS instructions, averaged over the lifetime of the kernel. Calculated as the ratio of the total number of cycles spent by the scheduler issuing LDS instructions over the total CU cycles.	Percent
Theoretical Bandwidth Utilization	Indicates the maximum amount of bytes that could have been loaded from, stored to, or atomically updated in the LDS divided as percentage of theoretical peak. Does not take into account the execution mask of the wavefront when the instruction was executed. See the LDS bandwidth example for more detail.	Percent
Bank Conflict Rate	Indicates the percentage of active LDS cycles that were spent servicing bank conflicts. Calculated as the ratio of LDS cycles spent servicing bank conflicts over the number of LDS cycles that would have been required to move the same amount of data in an uncontended access. [2]	Percent

Footnotes

[1] (1,2,3,4)

Here we assume the typical case where the workload evenly distributes LDS operations over all SIMDs in a CU (that is, waves on different SIMDs are executing similar code). For highly unbalanced workloads, where e.g., one SIMD pair in the CU does not issue LDS instructions at all, this metric is better interpreted as the percentage of SIMDs issuing LDS instructions on SIMD pairs that are actively using the LDS, averaged over the lifetime of the kernel.

[2] (1,2,3,4)

The maximum value of the bank conflict rate is less than 100% (specifically: 96.875%), as the first cycle in the LDS scheduler is never considered contended.

Statistics

The LDS statistics panel gives a more detailed view of the hardware:

CDNA

Metric	Description	Unit
Theoretical Bandwidth	Indicates the maximum amount of bytes that could have been loaded from, stored to, or atomically updated in the LDS divided by total duration. Does not take into account the execution mask of the wavefront when the instruction was executed. See the LDS bandwidth example for more detail.	Gbps
LDS Instructions	The total number of LDS instructions (including, but not limited to, read/write/atomics and HIP’s __shfl instructions) executed per normalization unit.	Instructions per Normalization Unit
LDS Latency	The average number of round-trip cycles (i.e., from issue to data-return acknowledgment) required for an LDS instruction to complete.	Cycles
Bank Conflicts/Access	The ratio of the number of cycles spent in the LDS scheduler due to bank conflicts (as determined by the conflict resolution hardware) to the base number of cycles that would be spent in the LDS scheduler in a completely uncontended case. This is the unnormalized form of the Bank Conflict Rate.	Conflicts per Access
Index Accesses	The total number of cycles spent in the LDS scheduler over all operations per normalization unit.	Cycles per Normalization Unit
Atomic Return Cycles	The total number of cycles spent on LDS atomics with return per normalization unit.	Cycles per Normalization Unit
Bank Conflict	The total number of cycles spent in the LDS scheduler due to bank conflicts (as determined by the conflict resolution hardware) per normalization unit.	Cycles per Normalization Unit
Addr Conflict	The total number of cycles spent in the LDS scheduler due to address conflicts (as determined by the conflict resolution hardware) per normalization unit.	Cycles per Normalization Unit
Unaligned Stall	The total number of cycles spent in the LDS scheduler due to stalls from non-dword aligned addresses per normalization unit.	Cycles per Normalization Unit
Mem Violations	The total number of out-of-bounds accesses made to the LDS, per normalization unit. This is unused and expected to be zero in most configurations for modern CDNAu2122 accelerators.	Accesses per Normalization Unit

CDNA 2

Metric	Description	Unit
Theoretical Bandwidth	Indicates the maximum amount of bytes that could have been loaded from, stored to, or atomically updated in the LDS divided by total duration. Does not take into account the execution mask of the wavefront when the instruction was executed. See the LDS bandwidth example for more detail.	Gbps
LDS Instructions	The total number of LDS instructions (including, but not limited to, read/write/atomics and HIP’s __shfl instructions) executed per normalization unit.	Instructions per Normalization Unit
LDS Latency	The average number of round-trip cycles (i.e., from issue to data-return acknowledgment) required for an LDS instruction to complete.	Cycles
Bank Conflicts/Access	The ratio of the number of cycles spent in the LDS scheduler due to bank conflicts (as determined by the conflict resolution hardware) to the base number of cycles that would be spent in the LDS scheduler in a completely uncontended case. This is the unnormalized form of the Bank Conflict Rate.	Conflicts per Access
Index Accesses	The total number of cycles spent in the LDS scheduler over all operations per normalization unit.	Cycles per Normalization Unit
Atomic Return Cycles	The total number of cycles spent on LDS atomics with return per normalization unit.	Cycles per Normalization Unit
Bank Conflict	The total number of cycles spent in the LDS scheduler due to bank conflicts (as determined by the conflict resolution hardware) per normalization unit.	Cycles per Normalization Unit
Addr Conflict	The total number of cycles spent in the LDS scheduler due to address conflicts (as determined by the conflict resolution hardware) per normalization unit.	Cycles per Normalization Unit
Unaligned Stall	The total number of cycles spent in the LDS scheduler due to stalls from non-dword aligned addresses per normalization unit.	Cycles per Normalization Unit
Mem Violations	The total number of out-of-bounds accesses made to the LDS, per normalization unit. This is unused and expected to be zero in most configurations for modern CDNAu2122 accelerators.	Accesses per Normalization Unit

CDNA 3

Metric	Description	Unit
Theoretical Bandwidth	Indicates the maximum amount of bytes that could have been loaded from, stored to, or atomically updated in the LDS divided by total duration. Does not take into account the execution mask of the wavefront when the instruction was executed. See the LDS bandwidth example for more detail.	Gbps
LDS Instructions	The total number of LDS instructions (including, but not limited to, read/write/atomics and HIP’s __shfl instructions) executed per normalization unit.	Instructions per Normalization Unit
LDS Latency	The average number of round-trip cycles (i.e., from issue to data-return acknowledgment) required for an LDS instruction to complete.	Cycles
Bank Conflicts/Access	The ratio of the number of cycles spent in the LDS scheduler due to bank conflicts (as determined by the conflict resolution hardware) to the base number of cycles that would be spent in the LDS scheduler in a completely uncontended case. This is the unnormalized form of the Bank Conflict Rate.	Conflicts per Access
Index Accesses	The total number of cycles spent in the LDS scheduler over all operations per normalization unit.	Cycles per Normalization Unit
Atomic Return Cycles	The total number of cycles spent on LDS atomics with return per normalization unit.	Cycles per Normalization Unit
Bank Conflict	The total number of cycles spent in the LDS scheduler due to bank conflicts (as determined by the conflict resolution hardware) per normalization unit.	Cycles per Normalization Unit
Addr Conflict	The total number of cycles spent in the LDS scheduler due to address conflicts (as determined by the conflict resolution hardware) per normalization unit.	Cycles per Normalization Unit
Unaligned Stall	The total number of cycles spent in the LDS scheduler due to stalls from non-dword aligned addresses per normalization unit.	Cycles per Normalization Unit
Mem Violations	The total number of out-of-bounds accesses made to the LDS, per normalization unit. This is unused and expected to be zero in most configurations for modern CDNAu2122 accelerators.	Accesses per Normalization Unit

CDNA 4

Metric	Description	Unit
Theoretical Bandwidth	Indicates the maximum amount of bytes that could have been loaded from, stored to, or atomically updated in the LDS divided by total duration. Does not take into account the execution mask of the wavefront when the instruction was executed. See the LDS bandwidth example for more detail.	Gbps
LDS LOAD Bandwidth	The effective bandwidth of LDS load operations, accounting for the work-items executed (execution mask). Calculated as the total bytes loaded from LDS divided by the kernel duration.	Gbps
LDS STORE Bandwidth	The effective bandwidth of LDS store operations, accounting for the work-items executed (execution mask). Calculated as the total bytes stored to LDS divided by the kernel duration.	Gbps
LDS ATOMIC Bandwidth	The effective bandwidth of LDS atomic operations, accounting for the work-items executed (execution mask). Calculated as the total bytes accessed by LDS atomic operations divided by the kernel duration.	Gbps
LDS LOAD	The total number of LDS load instructions issued per normalization unit.	Instructions per Normalization Unit
LDS STORE	The total number of LDS store instructions issued per normalization unit.	Instructions per Normalization Unit
LDS ATOMIC	The total number of LDS atomic instructions issued per normalization unit.	Instructions per Normalization Unit
LDS Instructions	The total number of LDS instructions (including, but not limited to, read/write/atomics and HIP’s __shfl instructions) executed per normalization unit.	Instructions per Normalization Unit
LDS Latency	The average number of round-trip cycles (i.e., from issue to data-return acknowledgment) required for an LDS instruction to complete.	Cycles
Bank Conflicts/Access	The ratio of the number of cycles spent in the LDS scheduler due to bank conflicts (as determined by the conflict resolution hardware) to the base number of cycles that would be spent in the LDS scheduler in a completely uncontended case. This is the unnormalized form of the Bank Conflict Rate.	Conflicts per Access
Index Accesses	The total number of cycles spent in the LDS scheduler over all operations per normalization unit.	Cycles per Normalization Unit
Atomic Return Cycles	The total number of cycles spent on LDS atomics with return per normalization unit.	Cycles per Normalization Unit
Bank Conflict	The total number of cycles spent in the LDS scheduler due to bank conflicts (as determined by the conflict resolution hardware) per normalization unit.	Cycles per Normalization Unit
Addr Conflict	The total number of cycles spent in the LDS scheduler due to address conflicts (as determined by the conflict resolution hardware) per normalization unit.	Cycles per Normalization Unit
Unaligned Stall	The total number of cycles spent in the LDS scheduler due to stalls from non-dword aligned addresses per normalization unit.	Cycles per Normalization Unit
Mem Violations	The total number of out-of-bounds accesses made to the LDS, per normalization unit. This is unused and expected to be zero in most configurations for modern CDNAu2122 accelerators.	Accesses per Normalization Unit
LDS Command FIFO Full Rate	The number of cycles where the LDS command FIFO was full, per normalization unit. High values indicate backpressure in LDS instruction dispatch, which may stall wavefronts waiting to issue LDS operations.	Cycles per Normalization Unit
LDS Data FIFO Full Rate	The number of cycles where the LDS data FIFO was full, per normalization unit. High values indicate backpressure in LDS data return path, which may stall wavefronts waiting for LDS read/atomic results.	Cycles per Normalization Unit