Home
last modified time | relevance | path

Searched refs:EightBitIntGemm (Results 1 – 8 of 8) sorted by relevance

/external/gemmlowp/eight_bit_int_gemm/
Deight_bit_int_gemm.h55 void EightBitIntGemm(bool transpose_a, bool transpose_b, bool transpose_c,
62 void EightBitIntGemm(bool transpose_a, bool transpose_b, bool transpose_c,
Deight_bit_int_gemm.cc304 void EightBitIntGemm(bool transpose_a, bool transpose_b, bool transpose_c, in EightBitIntGemm() function
310 ScopedLock sl(GlobalMutexes::EightBitIntGemm()); in EightBitIntGemm()
341 void EightBitIntGemm(bool transpose_a, bool transpose_b, bool transpose_c, in EightBitIntGemm() function
347 ScopedLock sl(GlobalMutexes::EightBitIntGemm()); in EightBitIntGemm()
408 ScopedLock sl(GlobalMutexes::EightBitIntGemm()); in SetMaxNumThreads()
414 ScopedLock sl(GlobalMutexes::EightBitIntGemm()); in FreePersistentResources()
/external/gemmlowp/
DREADME.txt82 2. EightBitIntGemm standard interface
86 implementation of the standard EightBitIntGemm interface. The header
141 Since it covers also the EightBitIntGemm interface, it needs to be
/external/gemmlowp/profiling/
Dinstrumentation.h87 static Mutex* EightBitIntGemm() { in EightBitIntGemm() function
/external/gemmlowp/doc/
Dless-than-8-bit.md64 ### The BitDepthSetting parameter in the EightBitIntGemm interface
66 Accessing less-than-8-bit computation via the EightBitIntGemm is very simple:
67 EightBitIntGemm takes a BitDepthSetting enum which allows to choose among a
Dlow-precision.md60 EightBitIntGemm paradigm, where quantized matrix multiplication takes the
/external/gemmlowp/test/
Dbenchmark_meta_gemm.cc98 gemmlowp::eight_bit_int_gemm::EightBitIntGemm( in run_gemm()
Dtest.cc251 eight_bit_int_gemm::EightBitIntGemm( in Gemm()
1152 gemmlowp::eight_bit_int_gemm::EightBitIntGemm( in TestWithSmallData()
1174 gemmlowp::eight_bit_int_gemm::EightBitIntGemm( in TestWithRealData()