TGEMV_ACC

1// PTO-INSTRUCTION: {"assembly":["TGEMV_ACC <bundle operands>"],"block":["BSTART.TGEMV.ACC AType","B.DATR BType, RMode, Sat (optional; BType defaults to AType)","B.FPATR PreQuantMode, ReluMode, GroupNCode, RowMaxEn, GroupMaxEn, RowMaxInit, MaxAbsEn, TransA, TransB, CScaleEn (exactly one)","B.DIM LB0 M (optional, default 1; TGEMV permits only M=1)","B.DIM LB1 N (optional, default 1)","B.DIM LB2 K (optional, default 1)","B.IOT ordered Local mathematical sources: C CUBE_M16/M32 accumulator matching A with encoded selector distinct from DstTile, A CUBE_M16/M32 primary, B CUBE_N8 primary","B.IOT D matching A's CUBE_M16/M32 layout with a distinct encoded destination index, optional RowMaxOut, optional GroupMaxOut destinations","B.IOT/B.IOR postprocess operands selected by B.FPATR","BSTOP or the next BSTART completion boundary"],"catalog_indices":[99],"catalog_records":[{"arguments":[{"operand":"destination0"},{"operand":"source0"},{"operand":"source1"},{"operand":"source2"}],"command_mnemonic":"BSTART.TGEMV.ACC","contract_status":"reviewed-complete","datr_contract":{"allowed_nonzero_fields":["DataType","RMode","Sat"],"pad_union":"must-zero"},"disposition":"accepted-direct-operation","effect_contract":"TGEMV_ACC","family":"CUBE","fault_contract":"ExecuteTileInstruction","function":18,"legality_handler":"TileOperandsLegal_TGEMV_ACC","name":"TGEMV_ACC","operands":[{"field":"destination0","role":"destination"},{"field":"source0","role":"accumulator"},{"field":"source1","role":"left-vector"},{"field":"source2","role":"right-matrix"}],"restart_contract":"CompleteBundleAtWithAcceptedApplicabilityRules","semantic_handler":"TGEMV_ACC","state_effects":["operand:destination0:destination","operand:source0:accumulator","operand:source1:left-vector","operand:source2:right-matrix"]}],"classification":["matrix-and-matrix-vector","matrix-vector"],"contract":{"block_composition":["BSTART.TGEMV.ACC AType","B.DATR BType, RMode, Sat (optional; BType defaults to AType)","B.FPATR PreQuantMode, ReluMode, GroupNCode, RowMaxEn, GroupMaxEn, RowMaxInit, MaxAbsEn, TransA, TransB, CScaleEn (exactly one)","B.DIM LB0 M (optional, default 1; TGEMV permits only M=1)","B.DIM LB1 N (optional, default 1)","B.DIM LB2 K (optional, default 1)","B.IOT ordered Local mathematical sources: C CUBE_M16/M32 accumulator matching A with encoded selector distinct from DstTile, A CUBE_M16/M32 primary, B CUBE_N8 primary","B.IOT D matching A's CUBE_M16/M32 layout with a distinct encoded destination index, optional RowMaxOut, optional GroupMaxOut destinations","B.IOT/B.IOR postprocess operands selected by B.FPATR","BSTOP or the next BSTART completion boundary"],"canonical_assembly":["TGEMV_ACC <bundle operands>"],"defaults":["Encoded DataType is always AType. Omitted B.DATR preserves AType as BType, selects RNE, and disables saturation.","Omitted LB0, LB1, and LB2 default M, N, and K independently to one; TGEMV fixes M to one.","Exactly one all-zero B.FPATR selects no conversion, activation, or reduction; B.IOR and auxiliary B.IOT operands exist only when a selected postprocess mode requires them.","Local A uses persistent CUBE_M16 or CUBE_M32, Local B uses persistent CUBE_N8, and D is newly allocated in A's M layout; Local C also uses A's M layout. M is fixed to one; N and K are arbitrary positive values independent of per-PE TSize.","TransA=0 and TransB=0 select no logical transpose. TGEMV requires both controls to remain zero.","C and D are both mandatory. In decoded blocks C's six-bit relative selector must differ from zero-extended DstTile before rename; direct Tile calls require destination TileIndex to differ from accumulator TileIndex."],"encoding_class":"selector-encoded-block-operation","examples":["BSTART.TGEMV.ACC AType; B.DATR BType, RMode, Sat (optional; BType defaults to AType); B.FPATR PreQuantMode, ReluMode, GroupNCode, RowMaxEn, GroupMaxEn, RowMaxInit, MaxAbsEn, TransA, TransB, CScaleEn (exactly one); B.DIM LB0 M (optional, default 1; TGEMV permits only M=1); B.DIM LB1 N (optional, default 1); B.DIM LB2 K (optional, default 1); B.IOT ordered Local mathematical sources: C CUBE_M16/M32 accumulator matching A with encoded selector distinct from DstTile, A CUBE_M16/M32 primary, B CUBE_N8 primary; B.IOT D matching A's CUBE_M16/M32 layout with a distinct encoded destination index, optional RowMaxOut, optional GroupMaxOut destinations; B.IOT/B.IOR postprocess operands selected by B.FPATR; BSTOP or the next BSTART completion boundary"],"exceptions":["A reserved DataType or fixed-bit mismatch raises Fault_IllegalInstruction before block state changes.","Missing, duplicate, or non-Matrix B.FPATR use raises Fault_BundleControl before allocation or payload effects.","Illegal types, dimensions, masks, binding streams, descriptors, shapes, capacities, aliases, readiness, or postprocess values raise Fault_TileLegality before source snapshots and effects."],"field_contracts":{},"field_zero_meanings":{},"legality":["The carrier selects exactly CUBE Function 18 and TileOperation_TGEMV_ACC.","Local A uses persistent CUBE_M16 or CUBE_M32, Local B uses persistent CUBE_N8, and D is newly allocated in A's M layout; Local C also uses A's M layout. M is fixed to one; N and K are arbitrary positive values independent of per-PE TSize.","C and D are both mandatory. In decoded blocks C's six-bit relative selector must differ from zero-extended DstTile before rename; direct Tile calls require destination TileIndex to differ from accumulator TileIndex.","TGEMV is Local-only: TransA and TransB are zero and every effective Shared binding rejects before effects.","AType and BType must be supported ordinary Matrix types from one numeric class. C is one explicit Local MxN accumulator source and D is a distinct newly published destination; C's encoded relative selector and D's zero-extended DstTile hand must differ before rename. M is fixed to one and every Shared binding is illegal.","Every common nonzero four-bit PE_MASK is legal; all four PEs complete cooperative Shared readiness while only selected PEs allocate and publish. Mask zero is a strict no-op before descriptor reads, faults, allocation, readiness checks, or lifetime effects.","B.DATR permits only BType, RMode, and Sat. Exactly one B.FPATR is mandatory and closes the conditional postprocess schema."],"memory_effects":["none"],"operands":[{"field":"destination0","role":"destination"},{"field":"source0","role":"accumulator"},{"field":"source1","role":"left-vector"},{"field":"source2","role":"right-matrix"}],"ordering":["Complete schema, field, type, dimension, descriptor, shape, capacity, readiness, alias, and allocation preflight precedes every source snapshot and destination effect.","D and every enabled reduction output publish as one atomic group; rejection publishes none and successful sources persist."],"standalone_opcode":false,"state_effects":["Multiply the matrix by the vector and accumulate into the supplied Tile.","After complete preflight, execute TGEMV_ACC with the operand bindings listed above; destination definedness changes only as specified by that handler.","For Local execution, publish D with A's CUBE_M16 or CUBE_M32 layout and final output dtype; ordinary Bias, MX scales, and enabled reduction auxiliaries keep their operation-owned layouts.","C is snapshotted before multiplication and remains descriptor-and-payload unchanged after success or rejection."]},"depends_on":["PTO-BLOCK-MODEL-SCHEMA-PROFILE-ENCODING"],"engine":"CUBE","id":"PTO-TILE-TGEMV-ACC","mnemonic":"TGEMV_ACC","summary":"Multiply the matrix by the vector and accumulate into the supplied Tile.","surface":"tile"}2// PTO-REVIEW: {"review_method":"formal-definition-read","outcome":"FORMAL-COMPLETE","reviewed_fields":["assembly","encoding","defaults","operation","state","memory","ordering","faults","reserved"]}3// NDF-BEGIN: PTO-TGEMV-ACC-CONTRACT-0014// ndf: kind=contract level=L1 layer=tile status=accepted5// TGEMV_ACC MUST select CUBE Function 18 and TileOperation_TGEMV_ACC.6// Its encoded DataType is AType; omitted B.DATR MUST preserve AType as7// BType, omitted LB0/LB1/LB2 MUST default M/N/K independently to one,8// Local N and K MUST be positive and independent of per-PE TSize; M9// MUST equal one.10// AType and BType must be supported ordinary Matrix types from one numeric11// class. C is one explicit Local MxN accumulator source and D is a newly12// published destination; C and D may use one architectural Tile name. M is13// fixed to one and every Shared binding is illegal.14// TransA and TransB MUST remain zero and every Shared binding MUST reject.15// Every common nonzero Local PE mask is legal; mask zero is a strict no-op.16// C and D MUST both be explicit; direct Tile calls MUST also use different17// destination and accumulator TileIndex values. C MUST persist unchanged.18// Exactly one B.FPATR MUST close postprocess defaults and operands.19// Complete schema, type, shape, capacity, definedness, readiness, alias,20// and allocation preflight MUST precede source snapshots and effects;21// successful outputs MUST publish atomically and sources MUST persist.22// NDF-END: PTO-TGEMV-ACC-CONTRACT-00123// DOC-BEGIN: decode24readonly func InstructionContractOperation_TGEMV_ACC()25    => TileOperation26begin27    return TileOperation_TGEMV_ACC;28end;29// DOC-END: decode30// DOC-BEGIN: operation31readonly func InstructionContractCubeFunction_TGEMV_ACC()32    => integer {0..31}33begin34    return 18;35end;36
37readonly func InstructionContractSharedOperandsAllowed_TGEMV_ACC()38    => boolean39begin40    return FALSE;41end;42
43readonly func InstructionContractOperandsLegal_TGEMV_ACC(44    destination: TileIndex,45    accumulator: TileIndex,46    left_vector: TileIndex,47    right_matrix: TileIndex) => boolean48begin49    return TileOperandsLegal_TGEMV_ACC(50        destination,51        accumulator,52        left_vector,53        right_matrix);54end;55
56readonly func InstructionContractHandler_TGEMV_ACC()57    => TileSemanticHandler58begin59    return TileHandler_TGEMV_ACC;60end;61// DOC-END: operation62