GQA ragged prefill causal h16 kv2 d128
0 eligible runs
gqa-ragged-attention
Batched Grouped Query Attention prefill with ragged (variable-length) inputs. Causal mask is applied. Captured from Qwen3-32B at TP=4 during total prefill. 16 q-heads, 2 kv-heads per device, head_dim=128.
Current records
No published measurement for the selected workload.
Implementations
Implementation
Runtime
Best latency
Evidence
Availability
Semantics
Inputs and outputs
qbf16 [total_q, num_qo_heads, head_dim]
kbf16 [total_kv, num_kv_heads, head_dim]
vbf16 [total_kv, num_kv_heads, head_dim]
qo_indptrint32 [len_indptr]
kv_indptrint32 [len_indptr]
sm_scalefp32 scalar
outputbf16 [total_q, num_qo_heads, head_dim]
lsefp32 [total_q, num_qo_heads]
Axes and behavior
total_qvariable
head_dimconstant = 128
total_kvvariable
len_indptrvariable
num_kv_headsconstant = 2
num_qo_headsconstant = 16
determinismunspecified
constraintsNo mutation or aliasing
Identity
No source imports for this operation yet.How records are decidedJSON