AI Engine Intrinsics User Guide
(AIE) v(2024.1)
|
Simple MAC intrinsics.
For general information on all vector MAC intrinsics go here.
Vector Multiply | |
v16acc48 | mul (v16int16 a, v16int16 b) |
Lane by lane 16-bit real multiply. | |
v8acc48 | mul (v8int16 a, v8int16 b) |
Lane by lane 16-bit real multiply. | |
v8cacc48 | mul (v8cint16 a, v8cint16 b) |
Lane by lane 16-bit complex multiply. | |
v4cacc48 | mul (v4cint16 a, v4cint16 b) |
Lane by lane 16-bit complex multiply. | |
v8acc80 | mul (v8int32 a, v8int32 b) |
Lane by lane 32-bit real multiply. | |
v4acc80 | mul (v4int32 a, v4int32 b) |
Lane by lane 16-bit real multiply. | |
v4cacc80 | mul (v4cint32 a, v4cint32 b) |
Lane by lane 32-bit complex multiply. | |
v2cacc80 | mul (v2cint32 a, v2cint32 b) |
Lane by lane 32-bit complex multiply. | |
Vector Multiply Accumulate | |
v8cacc48 | mac (v8cacc48 acc, v8cint16 a, v8cint16 b) |
Lane by lane 16-bit complex multiply-accumulate. | |
v4cacc48 | mac (v4cacc48 acc, v4cint16 a, v4cint16 b) |
Lane by lane 16-bit complex multiply-accumulate. | |
v8cacc48 | msc (v8cacc48 acc, v8cint16 a, v8cint16 b) |
Lane by lane 16-bit complex multiply-accumulate. | |
v4cacc48 | msc (v4cacc48 acc, v4cint16 a, v4cint16 b) |
Lane by lane 16-bit complex multiply-accumulate. | |
v16acc48 | mac (v16acc48 acc, v16int16 a, v16int16 b) |
Lane by lane 16-bit real multiply-accumulate. | |
v16acc48 | msc (v16acc48 a, v16int16 x, v16int16 y) |
Lane by lane 16-bit real multiply-accumulate. | |
v8acc48 | mac (v8acc48 a, v8int16 x, v8int16 y) |
Lane by lane 16-bit real multiply-accumulate. | |
v8acc48 | msc (v8acc48 a, v8int16 x, v8int16 y) |
Lane by lane 16-bit real multiply-accumulate. | |
v8acc80 | mac (v8acc80 a, v8int32 x, v8int32 y) |
Lane by lane 32-bit real multiply-accumulate. | |
v8acc80 | msc (v8acc80 a, v8int32 x, v8int32 y) |
Lane by lane 32-bit real multiply-accumulate. | |
v4acc80 | mac (v4acc80 a, v4int32 x, v4int32 y) |
Lane by lane 32-bit real multiply-accumulate. | |
v4acc80 | msc (v4acc80 a, v4int32 x, v4int32 y) |
Lane by lane 32-bit real multiply-accumulate. | |
v4cacc80 | mac (v4cacc80 a, v4cint32 x, v4cint32 y) |
Lane by lane 32-bit complex multiply-accumulate. | |
v4cacc80 | msc (v4cacc80 a, v4cint32 x, v4cint32 y) |
Lane by lane 32-bit complex multiply-accumulate. | |
v2cacc80 | mac (v2cacc80 a, v2cint32 x, v2cint32 y) |
Lane by lane 32-bit complex multiply-accumulate. | |
v2cacc80 | msc (v2cacc80 a, v2cint32 x, v2cint32 y) |
Lane by lane 32-bit complex multiply-accumulate. | |
v8cacc48 mac | ( | v8cacc48 | acc, |
v8cint16 | a, | ||
v8cint16 | b | ||
) |
Lane by lane 16-bit complex multiply-accumulate.
v4cacc48 mac | ( | v4cacc48 | acc, |
v4cint16 | a, | ||
v4cint16 | b | ||
) |
Lane by lane 16-bit complex multiply-accumulate.
v16acc48 mac | ( | v16acc48 | acc, |
v16int16 | a, | ||
v16int16 | b | ||
) |
Lane by lane 16-bit real multiply-accumulate.
v8acc48 mac | ( | v8acc48 | a, |
v8int16 | x, | ||
v8int16 | y | ||
) |
Lane by lane 16-bit real multiply-accumulate.
v8acc80 mac | ( | v8acc80 | a, |
v8int32 | x, | ||
v8int32 | y | ||
) |
Lane by lane 32-bit real multiply-accumulate.
v4acc80 mac | ( | v4acc80 | a, |
v4int32 | x, | ||
v4int32 | y | ||
) |
Lane by lane 32-bit real multiply-accumulate.
v4cacc80 mac | ( | v4cacc80 | a, |
v4cint32 | x, | ||
v4cint32 | y | ||
) |
Lane by lane 32-bit complex multiply-accumulate.
v2cacc80 mac | ( | v2cacc80 | a, |
v2cint32 | x, | ||
v2cint32 | y | ||
) |
Lane by lane 32-bit complex multiply-accumulate.
v8cacc48 msc | ( | v8cacc48 | acc, |
v8cint16 | a, | ||
v8cint16 | b | ||
) |
Lane by lane 16-bit complex multiply-accumulate.
v4cacc48 msc | ( | v4cacc48 | acc, |
v4cint16 | a, | ||
v4cint16 | b | ||
) |
Lane by lane 16-bit complex multiply-accumulate.
v16acc48 msc | ( | v16acc48 | a, |
v16int16 | x, | ||
v16int16 | y | ||
) |
Lane by lane 16-bit real multiply-accumulate.
v8acc48 msc | ( | v8acc48 | a, |
v8int16 | x, | ||
v8int16 | y | ||
) |
Lane by lane 16-bit real multiply-accumulate.
v8acc80 msc | ( | v8acc80 | a, |
v8int32 | x, | ||
v8int32 | y | ||
) |
Lane by lane 32-bit real multiply-accumulate.
v4acc80 msc | ( | v4acc80 | a, |
v4int32 | x, | ||
v4int32 | y | ||
) |
Lane by lane 32-bit real multiply-accumulate.
v4cacc80 msc | ( | v4cacc80 | a, |
v4cint32 | x, | ||
v4cint32 | y | ||
) |
Lane by lane 32-bit complex multiply-accumulate.
v2cacc80 msc | ( | v2cacc80 | a, |
v2cint32 | x, | ||
v2cint32 | y | ||
) |
Lane by lane 32-bit complex multiply-accumulate.
v16acc48 mul | ( | v16int16 | a, |
v16int16 | b | ||
) |
Lane by lane 16-bit real multiply.
v8acc48 mul | ( | v8int16 | a, |
v8int16 | b | ||
) |
Lane by lane 16-bit real multiply.
v8cacc48 mul | ( | v8cint16 | a, |
v8cint16 | b | ||
) |
Lane by lane 16-bit complex multiply.
v4cacc48 mul | ( | v4cint16 | a, |
v4cint16 | b | ||
) |
Lane by lane 16-bit complex multiply.
v8acc80 mul | ( | v8int32 | a, |
v8int32 | b | ||
) |
Lane by lane 32-bit real multiply.
v4acc80 mul | ( | v4int32 | a, |
v4int32 | b | ||
) |
Lane by lane 16-bit real multiply.
v4cacc80 mul | ( | v4cint32 | a, |
v4cint32 | b | ||
) |
Lane by lane 32-bit complex multiply.
v2cacc80 mul | ( | v2cint32 | a, |
v2cint32 | b | ||
) |
Lane by lane 32-bit complex multiply.