hexagon: MUL_MAT and MUL_MAT_ID rework : 32x32 tiled weight repack, kernel-params, cached graphs (#24954)
* hex-mm: new weight layout and fusion updates * hvx-mm: unroll the new tiled vec_dots to optimize hvx register util * hex-mm: optimize dyn.quant format for q8_0 and q8_1 to reduce overhead in vec_dots. * hvx-mm: parallel quantizer per block for large rows * hvx-mm: simplify and futher optimize dyn.quant and vec_dots * hvx-mm: keep intermediate per tile accumulators in fp16 * hmx-mm: optimize weight dequant by aligning the repacked tiles with the DMA * hmx-mm: remove qweight scratch and just use vtcm_weight * hmx-mm: remove all unused and obsolete code * hmx-mm: the new tiled repack format is here to stay -- rename all x4x2 to _tiled * hmx-mm: improve activation processing with dma prefetch * hex-mm: fix hmx/hvx fallback logic and MUL_MAT_ID allocation (unbreaks OLMoE) * hex-mm: align the weight tiles with dma just like we did in hmx-mm * hex-mm: factor out common mm bits into htp/matmul-ops.h * hex-mm: start moving mm kernel selection to the host * hex-mm: move all of the matmul param compute into the host * hmx-mm: restore pipelined mode * hmx-mm: unroll the dequant functions to optimize register usage * hmx-mm: further improve activation process * hex-mm: use vtcm_seq_alloc for all vtcm allocations and define more common functions * hex-mm: improve mm optimizer to acount for number of activation threads * hex-mm: fix matmul-id kernel params selection (unbreaks OLMoE and LFM) * hexagon: remove support for arch < v73 since HMX is now required for most use-cases * hex-mm: cleanup naming for consistency * hex-mm: make sure matmul fusion accounts for vtcm allocation * hex-mm: minor cleanup for kernel_params definition * hex-mm: replace hardcoded limits with proper checks for vtcm requirements * hex-mm: add support for non-tiled mm as a fallback option and factor out hvx kernels into separate header * hex-mm: remove unused functions * hex-mm: add shorthand for MM_SELECT in run-tool script * hvx-mm: factor out hvx/hmx microkernels and unify matmul entry and dispatch * hex-mm: further cleanup matmul fallback path * hex-mm: refactor matmul entry point and dispatch a bit further * hexagon: update cmake build to enable hmx for everything * hex-ops: optimize kernel_param updates and include summary in the logs * hex-mm: add support for GGML_HEXAGON_MM_SELECT * hex-mm: add hex-common header * hex-mm: pass correct number of tasks to workpool * hex-mm: add proper checks for no-work in dyn.quant tasks * hex-mm: convert all quantizers into a macro * hex-mm: fix hvx-flat fallback to pass all MUL_MAT tests * hex-mm: vectorize q8_1 quantizer * hex-mm: improve fused ffn mm stride handling * hex-mm: consistent use of n_threads and pipeline in kernel_params * hexagon: minor formatting * hex-mm: update MUL_MAT_ID kernel_param handling to make sure host/npu are in sync * hvx-mm: go back to accumulating in fp32 in tiled hvx kernels, more accurate and same perf * hvx-mm: unroll the loops and remove masking that is not needed for tiled accums * hmx-mm: optimize activation processing (slit loops, some unrolling, etc) * hmx-mm: minor optimization for output processing * hex-mm: consistent use of uint32_t and size_t in mm kernels * hex-mm: remove legacy restrictions for rows to be multiple of 256 * hexagon: replace sprintf with snprintf * hex-mm: relax hardcoded nrows checks and rely on VTCM size requirements * hexagon: minor alignment fix * hexagon: fix trailing spaces * hex-mm: relax padding from 256 to 128 (leftovers) * hex-mm: remove redundant checks for weight align to 128 we always use 2D dma for the weights and align them properly * hmx-mm: MUL_MAT_ID better work distribution between hvx threads and hmx tracing * hex-mm: specialize per-token mmid activation handling * hex-profile: update python scripts to handle kernel-params section in the logging output * hex-mm: move n_prefetch (aka dma_depth) into kernel params and remove unused fields * hex-trace: use easier to parse format, simply and fix post-proc scripts * hmx-mm: relax 32 row limit for output processing which helps utilization * hmx-mm: use start-chunk idx for tracing info * hmx-mm: parameterize activation dma pipeline * hexagon: add support for simple graph caching to avoid recomputing kernel-params * hex-mm: remove left-over repack functions * hex-mm: tighten n_prefetch asserts * hex-mm: remove duplicate round/align_up helper * hexagon: cleanup common header used in host/npu * hexagon: update early wakeup threshold * hmx-mm: define cost constants and update solver to assume that repacked ne[1] is padded to 32 * hmx-mm: make precompute_matmul a bit more readable (split into smaller functions, etc) * hex-mm: remove n_threads constraint * hex-mm: minor formatting updates * hex-mm: remove obsolete profiling logs * hex-mm: restore hardcode gate to refuse lm-head to avoid repacking that tensor
This commit is contained in:
Max Krasnyansky
@@ -28,18 +28,19 @@ enum htp_data_type {
|
||||
HTP_TYPE_MXFP4 = 39,
|
||||
|
||||
// types used internally for repack, dyn.quant, etc
|
||||
HTP_TYPE_Q4_0x4x2 = 200,
|
||||
HTP_TYPE_Q4_1x4x2,
|
||||
HTP_TYPE_Q8_0x4x2,
|
||||
HTP_TYPE_MXFP4x4x2,
|
||||
HTP_TYPE_Q4_0_TILED = 200,
|
||||
HTP_TYPE_Q4_1_TILED,
|
||||
HTP_TYPE_Q8_0_TILED,
|
||||
HTP_TYPE_MXFP4_TILED,
|
||||
|
||||
HTP_TYPE_INVALID
|
||||
};
|
||||
|
||||
// Constats for internal types
|
||||
#define QK_Q4_0x4x2 256 // 4x Q4_0 blocks packed with next 4x Q4_0 blocks (size in bytes 128)
|
||||
#define QK_Q8_0x4x2 256 // 4x Q8_0 blocks concat with next 4x Q8_0 blocks
|
||||
#define QK_MXFP4x4x2 256 // 4x MXFP4 blocks concat with next 4x MXFP4 blocks
|
||||
#define QK_Q4_0_TILED 256 // 32x32 Q4_0 tiled layout
|
||||
#define QK_Q8_0_TILED 128 // 32x32 Q8_0 tiled layout
|
||||
#define QK_MXFP4_TILED 256 // 32x32 MXFP4 tiled layout
|
||||
|
||||
|
||||
|
||||
// Mask to enable various stages of the Ops.
|
||||
@@ -57,6 +58,8 @@ enum htp_op_code {
|
||||
HTP_OP_DIV = 3,
|
||||
HTP_OP_MUL_MAT,
|
||||
HTP_OP_MUL_MAT_ID,
|
||||
HTP_OP_MUL_MAT_QKV,
|
||||
HTP_OP_MUL_MAT_FFN,
|
||||
HTP_OP_RMS_NORM,
|
||||
HTP_OP_RMS_NORM_MUL,
|
||||
HTP_OP_UNARY_SILU,
|
||||
@@ -99,7 +102,9 @@ enum htp_op_code {
|
||||
|
||||
#define HTP_OP_MAX_DIMS 4 // aka GGML_MAX_DIMS
|
||||
#define HTP_OP_MAX_INPUTS 6 // aka GGML_MAX_SRCS
|
||||
#define HTP_OP_MAX_OUTPUTS 4
|
||||
#define HTP_OP_MAX_PARAMS 16 // aka GGML_MAX_OP_PARAMS
|
||||
#define HTP_OP_MAX_KERN_PARAMS 32
|
||||
|
||||
#define HTP_OP_MAX_BUFS 16
|
||||
#define HTP_OP_MAX_REQS 256
|
||||
@@ -142,8 +147,10 @@ struct htp_op_desc {
|
||||
uint32_t opcode; // GGML/HTP Op
|
||||
uint32_t flags; // Op flags
|
||||
int32_t params[HTP_OP_MAX_PARAMS]; // Params for the op, e.g. epsilon of RMS norm
|
||||
int32_t kernel_params[HTP_OP_MAX_KERN_PARAMS]; // generic blob for host-precomputed parameters
|
||||
uint16_t src[HTP_OP_MAX_INPUTS]; // Input tensors indices
|
||||
uint16_t dst; // Output tensor index
|
||||
uint16_t dst[HTP_OP_MAX_OUTPUTS]; // Output tensor indices
|
||||
uint16_t pad[2]; // padding to align to 64 bits
|
||||
};
|
||||
|
||||
#ifndef HTP_MAX_NTHREADS
|
||||
|
||||
Reference in New Issue
Block a user