Statements (64)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
instruction set architecture extension
|
| gptkbp:abbreviation |
gptkb:Advanced_Vector_Extensions_512
|
| gptkbp:category |
SIMD instruction set
x86 instruction set extension |
| gptkbp:compatibleWith |
gptkb:ARM_processors
gptkb:Apple_Silicon AMD processors |
| gptkbp:developedBy |
gptkb:Intel
|
| gptkbp:documentation |
gptkb:Intel_64_and_IA-32_Architectures_Software_Developer’s_Manual
|
| gptkbp:feature |
bit manipulation instructions
gather/scatter instructions conflict detection instructions expand/compress instructions 4FMAPS (4 FMA packed single precision) 4VNNIW (4 VNNI packed word) BF16 (bfloat16 support) BITALG (bit algorithms) BW (byte and word) BW (byte and word) extensions CD (conflict detection) DQ (doubleword and quadword) DQ (doubleword and quadword) extensions ER (exponential and reciprocal) F (foundation) GFNI (Galois field instructions) IFMA (integer fused multiply-add) IFMA (integer fused multiply-add) extensions PF (prefetch) VAES (vector AES instructions) VBMI (vector byte manipulation instructions) VL (vector length) VL (vector length) extensions VNNI (vector neural network instructions) VP2INTERSECT (vector pair intersection) VPCLMULQDQ (carry-less multiplication) VPOPCNTDQ (vector population count doubleword and quadword) fused multiply-add (FMA) integer and floating-point operations mask registers (k0-k7) per-lane predication vector length agnostic operations |
| https://www.w3.org/2000/01/rdf-schema#label |
Intel AVX-512
|
| gptkbp:introducedIn |
2016
|
| gptkbp:predecessor |
gptkb:Intel_AVX2
|
| gptkbp:registration |
ZMM0-ZMM31
|
| gptkbp:successor |
none
|
| gptkbp:supportedBy |
gptkb:Intel_Skylake-X
gptkb:Intel_Xeon_Phi_Knights_Landing |
| gptkbp:supports |
SIMD operations
|
| gptkbp:usedBy |
gptkb:Intel_Xeon_Phi_processors
gptkb:Intel_Xeon_Scalable_processors gptkb:Intel_Ice_Lake_processors gptkb:Intel_Tiger_Lake_processors Intel Alder Lake processors (limited support) Intel Core i9 (select models) |
| gptkbp:usedIn |
gptkb:machine_learning
high-performance computing scientific computing data analytics |
| gptkbp:width |
512 bits
|
| gptkbp:bfsParent |
gptkb:Intel_Deep_Learning_Boost
gptkb:Xeon_Gold_6000_series gptkb:Intel_Advanced_Matrix_Extensions |
| gptkbp:bfsLayer |
6
|