Quantized dot products for CUDA mul mat vec (#2067)

This commit is contained in:
Johannes Gäßler 2023-07-05 14:19:42 +02:00 committed by GitHub
parent 051c70dcd5
commit 924dd22fd3
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
4 changed files with 427 additions and 93 deletions

View File

@ -68,8 +68,9 @@ option(LLAMA_ACCELERATE "llama: enable Accelerate framework
option(LLAMA_BLAS "llama: use BLAS" OFF) option(LLAMA_BLAS "llama: use BLAS" OFF)
set(LLAMA_BLAS_VENDOR "Generic" CACHE STRING "llama: BLAS library vendor") set(LLAMA_BLAS_VENDOR "Generic" CACHE STRING "llama: BLAS library vendor")
option(LLAMA_CUBLAS "llama: use cuBLAS" OFF) option(LLAMA_CUBLAS "llama: use cuBLAS" OFF)
option(LLAMA_CUDA_FORCE_DMMV "llama: use dmmv instead of mmvq CUDA kernels" OFF)
set(LLAMA_CUDA_DMMV_X "32" CACHE STRING "llama: x stride for dmmv CUDA kernels") set(LLAMA_CUDA_DMMV_X "32" CACHE STRING "llama: x stride for dmmv CUDA kernels")
set(LLAMA_CUDA_DMMV_Y "1" CACHE STRING "llama: y block size for dmmv CUDA kernels") set(LLAMA_CUDA_MMV_Y "1" CACHE STRING "llama: y block size for mmv CUDA kernels")
option(LLAMA_CUDA_DMMV_F16 "llama: use 16 bit floats for dmmv CUDA kernels" OFF) option(LLAMA_CUDA_DMMV_F16 "llama: use 16 bit floats for dmmv CUDA kernels" OFF)
set(LLAMA_CUDA_KQUANTS_ITER "2" CACHE STRING "llama: iters./thread per block for Q2_K/Q6_K") set(LLAMA_CUDA_KQUANTS_ITER "2" CACHE STRING "llama: iters./thread per block for Q2_K/Q6_K")
option(LLAMA_CLBLAST "llama: use CLBlast" OFF) option(LLAMA_CLBLAST "llama: use CLBlast" OFF)
@ -246,8 +247,14 @@ if (LLAMA_CUBLAS)
set(GGML_SOURCES_CUDA ggml-cuda.cu ggml-cuda.h) set(GGML_SOURCES_CUDA ggml-cuda.cu ggml-cuda.h)
add_compile_definitions(GGML_USE_CUBLAS) add_compile_definitions(GGML_USE_CUBLAS)
if (LLAMA_CUDA_FORCE_DMMV)
add_compile_definitions(GGML_CUDA_FORCE_DMMV)
endif()
add_compile_definitions(GGML_CUDA_DMMV_X=${LLAMA_CUDA_DMMV_X}) add_compile_definitions(GGML_CUDA_DMMV_X=${LLAMA_CUDA_DMMV_X})
add_compile_definitions(GGML_CUDA_DMMV_Y=${LLAMA_CUDA_DMMV_Y}) add_compile_definitions(GGML_CUDA_MMV_Y=${LLAMA_CUDA_MMV_Y})
if (DEFINED LLAMA_CUDA_DMMV_Y)
add_compile_definitions(GGML_CUDA_MMV_Y=${LLAMA_CUDA_DMMV_Y}) # for backwards compatibility
endif()
if (LLAMA_CUDA_DMMV_F16) if (LLAMA_CUDA_DMMV_F16)
add_compile_definitions(GGML_CUDA_DMMV_F16) add_compile_definitions(GGML_CUDA_DMMV_F16)
endif() endif()
@ -263,7 +270,7 @@ if (LLAMA_CUBLAS)
if (LLAMA_CUDA_DMMV_F16) if (LLAMA_CUDA_DMMV_F16)
set(CMAKE_CUDA_ARCHITECTURES "61") # needed for f16 CUDA intrinsics set(CMAKE_CUDA_ARCHITECTURES "61") # needed for f16 CUDA intrinsics
else() else()
set(CMAKE_CUDA_ARCHITECTURES "52") # lowest CUDA 12 standard set(CMAKE_CUDA_ARCHITECTURES "52;61") # lowest CUDA 12 standard + lowest for integer intrinsics
endif() endif()
endif() endif()
message(STATUS "Using CUDA architectures: ${CMAKE_CUDA_ARCHITECTURES}") message(STATUS "Using CUDA architectures: ${CMAKE_CUDA_ARCHITECTURES}")

View File

@ -164,16 +164,21 @@ ifdef LLAMA_CUBLAS
OBJS += ggml-cuda.o OBJS += ggml-cuda.o
NVCC = nvcc NVCC = nvcc
NVCCFLAGS = --forward-unknown-to-host-compiler -arch=native NVCCFLAGS = --forward-unknown-to-host-compiler -arch=native
ifdef LLAMA_CUDA_FORCE_DMMV
NVCCFLAGS += -DGGML_CUDA_FORCE_DMMV
endif # LLAMA_CUDA_FORCE_DMMV
ifdef LLAMA_CUDA_DMMV_X ifdef LLAMA_CUDA_DMMV_X
NVCCFLAGS += -DGGML_CUDA_DMMV_X=$(LLAMA_CUDA_DMMV_X) NVCCFLAGS += -DGGML_CUDA_DMMV_X=$(LLAMA_CUDA_DMMV_X)
else else
NVCCFLAGS += -DGGML_CUDA_DMMV_X=32 NVCCFLAGS += -DGGML_CUDA_DMMV_X=32
endif # LLAMA_CUDA_DMMV_X endif # LLAMA_CUDA_DMMV_X
ifdef LLAMA_CUDA_DMMV_Y ifdef LLAMA_CUDA_MMV_Y
NVCCFLAGS += -DGGML_CUDA_DMMV_Y=$(LLAMA_CUDA_DMMV_Y) NVCCFLAGS += -DGGML_CUDA_MMV_Y=$(LLAMA_CUDA_MMV_Y)
else ifdef LLAMA_CUDA_DMMV_Y
NVCCFLAGS += -DGGML_CUDA_MMV_Y=$(LLAMA_CUDA_DMMV_Y) # for backwards compatibility
else else
NVCCFLAGS += -DGGML_CUDA_DMMV_Y=1 NVCCFLAGS += -DGGML_CUDA_MMV_Y=1
endif # LLAMA_CUDA_DMMV_Y endif # LLAMA_CUDA_MMV_Y
ifdef LLAMA_CUDA_DMMV_F16 ifdef LLAMA_CUDA_DMMV_F16
NVCCFLAGS += -DGGML_CUDA_DMMV_F16 NVCCFLAGS += -DGGML_CUDA_DMMV_F16
endif # LLAMA_CUDA_DMMV_F16 endif # LLAMA_CUDA_DMMV_F16

View File

@ -345,8 +345,9 @@ Building the program with BLAS support may lead to some performance improvements
| Option | Legal values | Default | Description | | Option | Legal values | Default | Description |
|-------------------------|------------------------|---------|-------------| |-------------------------|------------------------|---------|-------------|
| LLAMA_CUDA_FORCE_DMMV | Boolean | false | Force the use of dequantization + matrix vector multiplication kernels instead of using kernels that do matrix vector multiplication on quantized data. By default the decision is made based on compute capability (MMVQ for 7.0/Turing/RTX 2000 or higher). Does not affect k-quants. |
| LLAMA_CUDA_DMMV_X | Positive integer >= 32 | 32 | Number of values in x direction processed by the CUDA dequantization + matrix vector multiplication kernel per iteration. Increasing this value can improve performance on fast GPUs. Power of 2 heavily recommended. Does not affect k-quants. | | LLAMA_CUDA_DMMV_X | Positive integer >= 32 | 32 | Number of values in x direction processed by the CUDA dequantization + matrix vector multiplication kernel per iteration. Increasing this value can improve performance on fast GPUs. Power of 2 heavily recommended. Does not affect k-quants. |
| LLAMA_CUDA_DMMV_Y | Positive integer | 1 | Block size in y direction for the CUDA dequantization + mul mat vec kernels. Increasing this value can improve performance on fast GPUs. Power of 2 recommended. Does not affect k-quants. | | LLAMA_CUDA_MMV_Y | Positive integer | 1 | Block size in y direction for the CUDA mul mat vec kernels. Increasing this value can improve performance on fast GPUs. Power of 2 recommended. Does not affect k-quants. |
| LLAMA_CUDA_DMMV_F16 | Boolean | false | If enabled, use half-precision floating point arithmetic for the CUDA dequantization + mul mat vec kernels. Can improve performance on relatively recent GPUs. | | LLAMA_CUDA_DMMV_F16 | Boolean | false | If enabled, use half-precision floating point arithmetic for the CUDA dequantization + mul mat vec kernels. Can improve performance on relatively recent GPUs. |
| LLAMA_CUDA_KQUANTS_ITER | 1 or 2 | 2 | Number of values processed per iteration and per CUDA thread for Q2_K and Q6_K quantization formats. Setting this value to 1 can improve performance for slow GPUs. | | LLAMA_CUDA_KQUANTS_ITER | 1 or 2 | 2 | Number of values processed per iteration and per CUDA thread for Q2_K and Q6_K quantization formats. Setting this value to 1 can improve performance for slow GPUs. |

View File

@ -70,9 +70,11 @@ typedef void (*ggml_cuda_op_t)(
// QK = number of values after dequantization // QK = number of values after dequantization
// QR = QK / number of values before dequantization // QR = QK / number of values before dequantization
// QI = number of 32 bit integers before dequantization
#define QK4_0 32 #define QK4_0 32
#define QR4_0 2 #define QR4_0 2
#define QI4_0 4
typedef struct { typedef struct {
half d; // delta half d; // delta
uint8_t qs[QK4_0 / 2]; // nibbles / quants uint8_t qs[QK4_0 / 2]; // nibbles / quants
@ -81,6 +83,7 @@ static_assert(sizeof(block_q4_0) == sizeof(ggml_fp16_t) + QK4_0 / 2, "wrong q4_0
#define QK4_1 32 #define QK4_1 32
#define QR4_1 2 #define QR4_1 2
#define QI4_1 4
typedef struct { typedef struct {
half d; // delta half d; // delta
half m; // min half m; // min
@ -90,6 +93,7 @@ static_assert(sizeof(block_q4_1) == sizeof(ggml_fp16_t) * 2 + QK4_1 / 2, "wrong
#define QK5_0 32 #define QK5_0 32
#define QR5_0 2 #define QR5_0 2
#define QI5_0 4
typedef struct { typedef struct {
half d; // delta half d; // delta
uint8_t qh[4]; // 5-th bit of quants uint8_t qh[4]; // 5-th bit of quants
@ -99,6 +103,7 @@ static_assert(sizeof(block_q5_0) == sizeof(ggml_fp16_t) + sizeof(uint32_t) + QK5
#define QK5_1 32 #define QK5_1 32
#define QR5_1 2 #define QR5_1 2
#define QI5_1 4
typedef struct { typedef struct {
half d; // delta half d; // delta
half m; // min half m; // min
@ -109,12 +114,25 @@ static_assert(sizeof(block_q5_1) == 2 * sizeof(ggml_fp16_t) + sizeof(uint32_t) +
#define QK8_0 32 #define QK8_0 32
#define QR8_0 1 #define QR8_0 1
#define QI8_0 8
typedef struct { typedef struct {
half d; // delta half d; // delta
int8_t qs[QK8_0]; // quants int8_t qs[QK8_0]; // quants
} block_q8_0; } block_q8_0;
static_assert(sizeof(block_q8_0) == sizeof(ggml_fp16_t) + QK8_0, "wrong q8_0 block size/padding"); static_assert(sizeof(block_q8_0) == sizeof(ggml_fp16_t) + QK8_0, "wrong q8_0 block size/padding");
#define QK8_1 32
#define QR8_1 1
#define QI8_1 8
typedef struct {
half d; // delta
half s; // unquantized sum
int8_t qs[QK8_0]; // quants
} block_q8_1;
static_assert(sizeof(block_q8_1) == 2*sizeof(ggml_fp16_t) + QK8_0, "wrong q8_1 block size/padding");
typedef float (*vec_dot_q_cuda_t)(const void * vbq, const block_q8_1 * bq8_1, const int iqs);
//================================= k-quants //================================= k-quants
#ifdef GGML_QKK_64 #ifdef GGML_QKK_64
@ -198,14 +216,15 @@ static_assert(sizeof(block_q6_K) == sizeof(ggml_fp16_t) + 13*QK_K/16, "wrong q6_
#define CUDA_SCALE_BLOCK_SIZE 256 #define CUDA_SCALE_BLOCK_SIZE 256
#define CUDA_ROPE_BLOCK_SIZE 256 #define CUDA_ROPE_BLOCK_SIZE 256
#define CUDA_DIAG_MASK_INF_BLOCK_SIZE 32 #define CUDA_DIAG_MASK_INF_BLOCK_SIZE 32
#define CUDA_QUANTIZE_BLOCK_SIZE 256
#define CUDA_DEQUANTIZE_BLOCK_SIZE 256 #define CUDA_DEQUANTIZE_BLOCK_SIZE 256
// dmmv = dequantize_mul_mat_vec // dmmv = dequantize_mul_mat_vec
#ifndef GGML_CUDA_DMMV_X #ifndef GGML_CUDA_DMMV_X
#define GGML_CUDA_DMMV_X 32 #define GGML_CUDA_DMMV_X 32
#endif #endif
#ifndef GGML_CUDA_DMMV_Y #ifndef GGML_CUDA_MMV_Y
#define GGML_CUDA_DMMV_Y 1 #define GGML_CUDA_MMV_Y 1
#endif #endif
#ifndef K_QUANTS_PER_ITERATION #ifndef K_QUANTS_PER_ITERATION
@ -270,7 +289,6 @@ static __global__ void rms_norm_f32(const float * x, float * dst, const int ncol
} }
// sum up partial sums // sum up partial sums
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -714,7 +732,6 @@ static __global__ void dequantize_mul_mat_vec_q2_k(const void * vx, const float
#endif #endif
// sum up partial sums and write back result // sum up partial sums and write back result
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -819,7 +836,6 @@ static __global__ void dequantize_mul_mat_vec_q3_k(const void * vx, const float
#endif #endif
// sum up partial sums and write back result // sum up partial sums and write back result
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -923,7 +939,6 @@ static __global__ void dequantize_mul_mat_vec_q4_k(const void * vx, const float
#endif #endif
// sum up partial sums and write back result // sum up partial sums and write back result
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -1028,7 +1043,6 @@ static __global__ void dequantize_mul_mat_vec_q5_k(const void * vx, const float
#endif #endif
// sum up partial sums and write back result // sum up partial sums and write back result
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -1139,7 +1153,6 @@ static __global__ void dequantize_mul_mat_vec_q6_k(const void * vx, const float
#endif #endif
// sum up partial sums and write back result // sum up partial sums and write back result
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -1158,6 +1171,41 @@ static __device__ void convert_f16(const void * vx, const int ib, const int iqs,
v.y = x[ib + iqs + 1]; v.y = x[ib + iqs + 1];
} }
static __global__ void quantize_q8_1(const float * x, void * vy, const int k) {
const int i = blockDim.x*blockIdx.x + threadIdx.x;
if (i >= k) {
return;
}
block_q8_1 * y = (block_q8_1 *) vy;
const int ib = i / QK8_0; // block index
const int iqs = i % QK8_0; // quant index
const float xi = x[i];
float amax = fabsf(xi);
float sum = xi;
#pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) {
amax = fmaxf(amax, __shfl_xor_sync(0xffffffff, amax, mask, 32));
sum += __shfl_xor_sync(0xffffffff, sum, mask, 32);
}
const float d = amax / 127;
const int8_t q = amax == 0.0f ? 0 : roundf(xi / d);
y[ib].qs[iqs] = q;
if (iqs > 0) {
return;
}
y[ib].d = d;
y[ib].s = sum;
}
template <int qk, int qr, dequantize_kernel_t dequantize_kernel> template <int qk, int qr, dequantize_kernel_t dequantize_kernel>
static __global__ void dequantize_block(const void * vx, float * y, const int k) { static __global__ void dequantize_block(const void * vx, float * y, const int k) {
const int i = blockDim.x*blockIdx.x + 2*threadIdx.x; const int i = blockDim.x*blockIdx.x + 2*threadIdx.x;
@ -1179,6 +1227,182 @@ static __global__ void dequantize_block(const void * vx, float * y, const int k)
y[iybs + iqs + y_offset] = v.y; y[iybs + iqs + y_offset] = v.y;
} }
static __device__ __forceinline__ float vec_dot_q4_0_q8_1(const void * vbq, const block_q8_1 * bq8_1, const int iqs) {
#if __CUDA_ARCH__ >= 600 // lowest compute capability for integer intrinsics
const block_q4_0 * bq4_0 = (const block_q4_0 *) vbq;
int vi;
memcpy(&vi, &bq4_0->qs[sizeof(int) * (iqs + 0)], sizeof(int));
const int ui0 = *((int *) &bq8_1->qs[sizeof(int) * (iqs + 0)]);
const int ui1 = *((int *) &bq8_1->qs[sizeof(int) * (iqs + QI4_0)]);
const float d = __half2float(bq4_0->d) * __half2float(bq8_1->d);
// subtract 8 from each quantized value
const int vi0 = __vsub4((vi >> 0) & 0x0F0F0F0F, 0x08080808);
const int vi1 = __vsub4((vi >> 4) & 0x0F0F0F0F, 0x08080808);
// SIMD dot product of quantized values
int sumi = __dp4a(vi0, ui0, 0);
sumi = __dp4a(vi1, ui1, sumi);
return sumi*d;
#else
return 0.0f; // only to satisfy the compiler
#endif // __CUDA_ARCH__ >= 600
}
static __device__ __forceinline__ float vec_dot_q4_1_q8_1(const void * vbq, const block_q8_1 * bq8_1, const int iqs) {
#if __CUDA_ARCH__ >= 600 // lowest compute capability for integer intrinsics
const block_q4_1 * bq4_1 = (const block_q4_1 *) vbq;
const int vi = *((int *) &bq4_1->qs[sizeof(int) * (iqs + 0)]);
const int ui0 = *((int *) &bq8_1->qs[sizeof(int) * (iqs + 0)]);
const int ui1 = *((int *) &bq8_1->qs[sizeof(int) * (iqs + QI4_1)]);
const float d = __half2float(bq4_1->d) * __half2float(bq8_1->d);
const float m = bq4_1->m;
const float s = bq8_1->s;
const int vi0 = (vi >> 0) & 0x0F0F0F0F;
const int vi1 = (vi >> 4) & 0x0F0F0F0F;
// SIMD dot product of quantized values
int sumi = __dp4a(vi0, ui0, 0);
sumi = __dp4a(vi1, ui1, sumi);
return sumi*d + m*s / QI4_1; // scale sum by QI4_1 because there are QI4_1 threads working on this block
#else
return 0.0f; // only to satisfy the compiler
#endif // __CUDA_ARCH__ >= 600
}
static __device__ __forceinline__ float vec_dot_q5_0_q8_1(const void * vbq, const block_q8_1 * bq8_1, const int iqs) {
#if __CUDA_ARCH__ >= 600 // lowest compute capability for integer intrinsics
const block_q5_0 * bq5_0 = (const block_q5_0 *) vbq;
int qs;
memcpy(&qs, &bq5_0->qs[sizeof(int) * (iqs + 0)], sizeof(int));
const int qh0 = bq5_0->qh[iqs/2 + 0] >> 4*(iqs%2);
const int qh1 = bq5_0->qh[iqs/2 + 2] >> 4*(iqs%2);
const int ui0 = *((int *) &bq8_1->qs[sizeof(int) * (iqs + 0)]);
const int ui1 = *((int *) &bq8_1->qs[sizeof(int) * (iqs + QI5_0)]);
const float d = __half2float(bq5_0->d) * __half2float(bq8_1->d);
int vi0 = (qs >> 0) & 0x0F0F0F0F; // lower 4 qs bits, still need qh0 as 5th bits
vi0 |= (qh0 << 4) & 0x00000010; // 1 -> 5
vi0 |= (qh0 << 11) & 0x00001000; // 2 -> 13
vi0 |= (qh0 << 18) & 0x00100000; // 3 -> 21
vi0 |= (qh0 << 25) & 0x10000000; // 4 -> 29
vi0 = __vsub4(vi0, 0x10101010); // subtract 16 from quantized values
int sumi = __dp4a(vi0, ui0, 0); // SIMD dot product of quantized values
int vi1 = (qs >> 4) & 0x0F0F0F0F; // upper 4 qs bits, still need qh1 as 5th bits
vi1 |= (qh1 << 4) & 0x00000010; // 1 -> 5
vi1 |= (qh1 << 11) & 0x00001000; // 2 -> 13
vi1 |= (qh1 << 18) & 0x00100000; // 3 -> 21
vi1 |= (qh1 << 25) & 0x10000000; // 4 -> 29
vi1 = __vsub4(vi1, 0x10101010); // subtract 16 from quantized values
sumi = __dp4a(vi1, ui1, sumi); // SIMD dot product of quantized values
return sumi*d;
#else
return 0.0f; // only to satisfy the compiler
#endif // __CUDA_ARCH__ >= 600
}
static __device__ __forceinline__ float vec_dot_q5_1_q8_1(const void * vbq, const block_q8_1 * bq8_1, const int iqs) {
#if __CUDA_ARCH__ >= 600 // lowest compute capability for integer intrinsics
const block_q5_1 * bq5_1 = (const block_q5_1 *) vbq;
const int qs = *((int *) &bq5_1->qs[sizeof(int) * (iqs + 0)]);
const int qh0 = bq5_1->qh[iqs/2 + 0] >> 4*(iqs%2);
const int qh1 = bq5_1->qh[iqs/2 + 2] >> 4*(iqs%2);
const int ui0 = *((int *) &bq8_1->qs[sizeof(int) * (iqs + 0)]);
const int ui1 = *((int *) &bq8_1->qs[sizeof(int) * (iqs + QI5_1)]);
const float d = __half2float(bq5_1->d) * __half2float(bq8_1->d);
const float m = bq5_1->m;
const float s = bq8_1->s;
int vi0 = (qs >> 0) & 0x0F0F0F0F; // lower 4 qs bits, still need qh0 as 5th bits
vi0 |= (qh0 << 4) & 0x00000010; // 1 -> 5
vi0 |= (qh0 << 11) & 0x00001000; // 2 -> 13
vi0 |= (qh0 << 18) & 0x00100000; // 3 -> 21
vi0 |= (qh0 << 25) & 0x10000000; // 4 -> 29
int sumi = __dp4a(vi0, ui0, 0); // SIMD dot product of quantized values
int vi1 = (qs >> 4) & 0x0F0F0F0F; // upper 4 qs bits, still need qh1 as 5th bits
vi1 |= (qh1 << 4) & 0x00000010; // 1 -> 5
vi1 |= (qh1 << 11) & 0x00001000; // 2 -> 13
vi1 |= (qh1 << 18) & 0x00100000; // 3 -> 21
vi1 |= (qh1 << 25) & 0x10000000; // 4 -> 29
sumi = __dp4a(vi1, ui1, sumi); // SIMD dot product of quantized values
return sumi*d + m*s / QI5_1; // scale sum by QI5_1 because there are QI5_1 threads working on this block
#else
return 0.0f; // only to satisfy the compiler
#endif // __CUDA_ARCH__ >= 600
}
static __device__ __forceinline__ float vec_dot_q8_0_q8_1(const void * vbq, const block_q8_1 * bq8_1, const int iqs) {
#if __CUDA_ARCH__ >= 600 // lowest compute capability for integer intrinsics
const block_q8_0 * bq8_0 = (const block_q8_0 *) vbq;
int vi;
memcpy(&vi, &bq8_0->qs[sizeof(int) * (iqs + 0)], sizeof(int));
const int ui = *((int *) &bq8_1->qs[sizeof(int) * (iqs + 0)]);
const float d = __half2float(bq8_0->d) * __half2float(bq8_1->d);
// SIMD dot product of quantized values
int sumi = __dp4a(vi, ui, 0);
return sumi*d;
#else
return 0.0f; // only to satisfy the compiler
#endif // __CUDA_ARCH__ >= 600
}
template <int qk, int qi, typename block_q_t, vec_dot_q_cuda_t vec_dot_q_cuda>
static __global__ void mul_mat_vec_q(const void * vx, const void * vy, float * dst, const int ncols, const int nrows) {
const int row = blockIdx.y*blockDim.y + threadIdx.y;
if (row >= nrows) {
return;
}
const int blocks_per_row = ncols / qk;
const int blocks_per_warp = WARP_SIZE / qi;
// partial sum for each thread
float tmp = 0.0f;
const block_q_t * x = (const block_q_t *) vx;
const block_q8_1 * y = (const block_q8_1 *) vy;
for (int i = 0; i < blocks_per_row; i += blocks_per_warp) {
const int ibx = row*blocks_per_row + i + threadIdx.x / qi; // x block index
const int iby = i + threadIdx.x / qi; // y block index
const int iqs = threadIdx.x % qi; // x block quant index when casting the quants to int
tmp += vec_dot_q_cuda(&x[ibx], &y[iby], iqs);
}
// sum up partial sums and write back result
#pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
}
if (threadIdx.x == 0) {
dst[row] = tmp;
}
}
template <int qk, int qr, dequantize_kernel_t dequantize_kernel> template <int qk, int qr, dequantize_kernel_t dequantize_kernel>
static __global__ void dequantize_mul_mat_vec(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows) { static __global__ void dequantize_mul_mat_vec(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows) {
// qk = quantized weights per x block // qk = quantized weights per x block
@ -1233,7 +1457,6 @@ static __global__ void dequantize_mul_mat_vec(const void * vx, const dfloat * y,
} }
// sum up partial sums and write back result // sum up partial sums and write back result
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -1284,7 +1507,6 @@ static __global__ void mul_mat_p021_f16_f32(const void * vx, const float * y, fl
const int idst = channel*nrows_dst + row_dst; const int idst = channel*nrows_dst + row_dst;
// sum up partial sums and write back result // sum up partial sums and write back result
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -1330,7 +1552,6 @@ static __global__ void mul_mat_vec_nc_f16_f32( // nc == non-contiguous
} }
// sum up partial sums and write back result // sum up partial sums and write back result
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -1440,7 +1661,6 @@ static __global__ void soft_max_f32(const float * x, float * dst, const int ncol
} }
// sum up partial sums // sum up partial sums
__syncthreads();
#pragma unroll #pragma unroll
for (int mask = 16; mask > 0; mask >>= 1) { for (int mask = 16; mask > 0; mask >>= 1) {
tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32); tmp += __shfl_xor_sync(0xffffffff, tmp, mask, 32);
@ -1494,6 +1714,11 @@ static void rms_norm_f32_cuda(const float * x, float * dst, const int ncols, con
rms_norm_f32<<<nrows, block_dims, 0, stream>>>(x, dst, ncols); rms_norm_f32<<<nrows, block_dims, 0, stream>>>(x, dst, ncols);
} }
static void quantize_row_q8_1_cuda(const float * x, void * vy, const int k, cudaStream_t stream) {
const int num_blocks = (k + CUDA_QUANTIZE_BLOCK_SIZE - 1) / CUDA_QUANTIZE_BLOCK_SIZE;
quantize_q8_1<<<num_blocks, CUDA_QUANTIZE_BLOCK_SIZE, 0, stream>>>(x, vy, k);
}
static void dequantize_row_q4_0_cuda(const void * vx, float * y, const int k, cudaStream_t stream) { static void dequantize_row_q4_0_cuda(const void * vx, float * y, const int k, cudaStream_t stream) {
const int num_blocks = (k + CUDA_DEQUANTIZE_BLOCK_SIZE - 1) / CUDA_DEQUANTIZE_BLOCK_SIZE; const int num_blocks = (k + CUDA_DEQUANTIZE_BLOCK_SIZE - 1) / CUDA_DEQUANTIZE_BLOCK_SIZE;
dequantize_block<QK4_0, QR4_0, dequantize_q4_0><<<num_blocks, CUDA_DEQUANTIZE_BLOCK_SIZE, 0, stream>>>(vx, y, k); dequantize_block<QK4_0, QR4_0, dequantize_q4_0><<<num_blocks, CUDA_DEQUANTIZE_BLOCK_SIZE, 0, stream>>>(vx, y, k);
@ -1562,45 +1787,45 @@ static void dequantize_row_q6_K_cuda(const void * vx, float * y, const int k, cu
static void dequantize_mul_mat_vec_q4_0_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) { static void dequantize_mul_mat_vec_q4_0_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0); GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_DMMV_Y - 1) / GGML_CUDA_DMMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1); const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_DMMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
dequantize_mul_mat_vec<QK4_0, QR4_0, dequantize_q4_0> dequantize_mul_mat_vec<QK4_0, QR4_0, dequantize_q4_0>
<<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows);
} }
static void dequantize_mul_mat_vec_q4_1_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) { static void dequantize_mul_mat_vec_q4_1_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0); GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_DMMV_Y - 1) / GGML_CUDA_DMMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1); const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_DMMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
dequantize_mul_mat_vec<QK4_1, QR4_1, dequantize_q4_1> dequantize_mul_mat_vec<QK4_1, QR4_1, dequantize_q4_1>
<<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows);
} }
static void dequantize_mul_mat_vec_q5_0_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) { static void dequantize_mul_mat_vec_q5_0_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0); GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_DMMV_Y - 1) / GGML_CUDA_DMMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1); const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_DMMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
dequantize_mul_mat_vec<QK5_0, QR5_0, dequantize_q5_0> dequantize_mul_mat_vec<QK5_0, QR5_0, dequantize_q5_0>
<<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows);
} }
static void dequantize_mul_mat_vec_q5_1_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) { static void dequantize_mul_mat_vec_q5_1_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0); GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_DMMV_Y - 1) / GGML_CUDA_DMMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1); const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_DMMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
dequantize_mul_mat_vec<QK5_1, QR5_1, dequantize_q5_1> dequantize_mul_mat_vec<QK5_1, QR5_1, dequantize_q5_1>
<<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows);
} }
static void dequantize_mul_mat_vec_q8_0_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) { static void dequantize_mul_mat_vec_q8_0_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0); GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_DMMV_Y - 1) / GGML_CUDA_DMMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1); const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_DMMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
dequantize_mul_mat_vec<QK8_0, QR8_0, dequantize_q8_0> dequantize_mul_mat_vec<QK8_0, QR8_0, dequantize_q8_0>
<<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows);
} }
@ -1647,6 +1872,51 @@ static void dequantize_mul_mat_vec_q6_K_cuda(const void * vx, const float * y, f
dequantize_mul_mat_vec_q6_k<<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows); dequantize_mul_mat_vec_q6_k<<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows);
} }
static void mul_mat_vec_q4_0_q8_1_cuda(const void * vx, const void * vy, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<QK4_0, QI4_0, block_q4_0, vec_dot_q4_0_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows);
}
static void mul_mat_vec_q4_1_q8_1_cuda(const void * vx, const void * vy, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<QK4_0, QI4_1, block_q4_1, vec_dot_q4_1_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows);
}
static void mul_mat_vec_q5_0_q8_1_cuda(const void * vx, const void * vy, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<QK5_0, QI5_0, block_q5_0, vec_dot_q5_0_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows);
}
static void mul_mat_vec_q5_1_q8_1_cuda(const void * vx, const void * vy, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<QK5_1, QI5_1, block_q5_1, vec_dot_q5_1_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows);
}
static void mul_mat_vec_q8_0_q8_1_cuda(const void * vx, const void * vy, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
mul_mat_vec_q<QK8_0, QI8_0, block_q8_0, vec_dot_q8_0_q8_1>
<<<block_nums, block_dims, 0, stream>>>(vx, vy, dst, ncols, nrows);
}
static void convert_fp16_to_fp32_cuda(const void * vx, float * y, const int k, cudaStream_t stream) { static void convert_fp16_to_fp32_cuda(const void * vx, float * y, const int k, cudaStream_t stream) {
const int num_blocks = (k + CUDA_DEQUANTIZE_BLOCK_SIZE - 1) / CUDA_DEQUANTIZE_BLOCK_SIZE; const int num_blocks = (k + CUDA_DEQUANTIZE_BLOCK_SIZE - 1) / CUDA_DEQUANTIZE_BLOCK_SIZE;
dequantize_block<1, 1, convert_f16><<<num_blocks, CUDA_DEQUANTIZE_BLOCK_SIZE, 0, stream>>>(vx, y, k); dequantize_block<1, 1, convert_f16><<<num_blocks, CUDA_DEQUANTIZE_BLOCK_SIZE, 0, stream>>>(vx, y, k);
@ -1654,9 +1924,9 @@ static void convert_fp16_to_fp32_cuda(const void * vx, float * y, const int k, c
static void convert_mul_mat_vec_f16_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) { static void convert_mul_mat_vec_f16_cuda(const void * vx, const dfloat * y, float * dst, const int ncols, const int nrows, cudaStream_t stream) {
GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0); GGML_ASSERT(ncols % GGML_CUDA_DMMV_X == 0);
const int block_num_y = (nrows + GGML_CUDA_DMMV_Y - 1) / GGML_CUDA_DMMV_Y; const int block_num_y = (nrows + GGML_CUDA_MMV_Y - 1) / GGML_CUDA_MMV_Y;
const dim3 block_nums(1, block_num_y, 1); const dim3 block_nums(1, block_num_y, 1);
const dim3 block_dims(WARP_SIZE, GGML_CUDA_DMMV_Y, 1); const dim3 block_dims(WARP_SIZE, GGML_CUDA_MMV_Y, 1);
dequantize_mul_mat_vec<1, 1, convert_f16> dequantize_mul_mat_vec<1, 1, convert_f16>
<<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows); <<<block_nums, block_dims, 0, stream>>>(vx, y, dst, ncols, nrows);
} }
@ -1822,6 +2092,7 @@ static size_t g_scratch_offset = 0;
static int g_device_count = -1; static int g_device_count = -1;
static int g_main_device = 0; static int g_main_device = 0;
static int g_compute_capabilities[GGML_CUDA_MAX_DEVICES];
static float g_tensor_split[GGML_CUDA_MAX_DEVICES] = {0}; static float g_tensor_split[GGML_CUDA_MAX_DEVICES] = {0};
static cublasHandle_t g_cublas_handles[GGML_CUDA_MAX_DEVICES] = {nullptr}; static cublasHandle_t g_cublas_handles[GGML_CUDA_MAX_DEVICES] = {nullptr};
@ -1839,9 +2110,12 @@ void ggml_init_cublas() {
for (int id = 0; id < g_device_count; ++id) { for (int id = 0; id < g_device_count; ++id) {
cudaDeviceProp prop; cudaDeviceProp prop;
CUDA_CHECK(cudaGetDeviceProperties(&prop, id)); CUDA_CHECK(cudaGetDeviceProperties(&prop, id));
fprintf(stderr, " Device %d: %s\n", id, prop.name); fprintf(stderr, " Device %d: %s, compute capability %d.%d\n", id, prop.name, prop.major, prop.minor);
g_tensor_split[id] = total_vram; g_tensor_split[id] = total_vram;
total_vram += prop.totalGlobalMem; total_vram += prop.totalGlobalMem;
g_compute_capabilities[id] = 100*prop.major + 10*prop.minor;
} }
for (int id = 0; id < g_device_count; ++id) { for (int id = 0; id < g_device_count; ++id) {
g_tensor_split[id] /= total_vram; g_tensor_split[id] /= total_vram;
@ -2057,7 +2331,7 @@ inline void ggml_cuda_op_rms_norm(
(void) i1; (void) i1;
} }
inline void ggml_cuda_op_dequantize_mul_mat_vec( inline void ggml_cuda_op_mul_mat_vec(
const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst, char * src0_ddq_i, const ggml_tensor * src0, const ggml_tensor * src1, ggml_tensor * dst, char * src0_ddq_i,
float * src0_ddf_i, float * src1_ddf_i, float * dst_ddf_i, int64_t i02, int64_t i01_low, int64_t i01_high, int i1, float * src0_ddf_i, float * src1_ddf_i, float * dst_ddf_i, int64_t i02, int64_t i01_low, int64_t i01_high, int i1,
cudaStream_t & cudaStream_main){ cudaStream_t & cudaStream_main){
@ -2069,69 +2343,116 @@ inline void ggml_cuda_op_dequantize_mul_mat_vec(
const int64_t ne00 = src0->ne[0]; const int64_t ne00 = src0->ne[0];
const int64_t nrows = i01_high - i01_low; const int64_t nrows = i01_high - i01_low;
// on some GPUs it is faster to convert src1 to half and to use half precision intrinsics #ifdef GGML_CUDA_FORCE_DMMV
#ifdef GGML_CUDA_DMMV_F16 const bool use_mul_mat_vec_q = false;
size_t ash;
dfloat * src1_dfloat = nullptr; // dfloat == half
bool src1_convert_f16 = src0->type == GGML_TYPE_Q4_0 || src0->type == GGML_TYPE_Q4_1 ||
src0->type == GGML_TYPE_Q5_0 || src0->type == GGML_TYPE_Q5_1 ||
src0->type == GGML_TYPE_Q8_0 || src0->type == GGML_TYPE_F16;
if (src1_convert_f16) {
src1_dfloat = (half *) ggml_cuda_pool_malloc(ne00*sizeof(half), &ash);
ggml_cpy_f32_f16_cuda((char *) src1_ddf_i, (char *) src1_dfloat, ne00,
ne00, 1, sizeof(float), 0, 0,
ne00, 1, sizeof(half), 0, 0, cudaStream_main);
}
#else #else
dfloat * src1_dfloat = src1_ddf_i; // dfloat == float, no conversion int id;
CUDA_CHECK(cudaGetDevice(&id));
const bool mul_mat_vec_q_implemented = src0->type == GGML_TYPE_Q4_0 ||
src0->type == GGML_TYPE_Q4_1 ||
src0->type == GGML_TYPE_Q5_0 ||
src0->type == GGML_TYPE_Q5_1 ||
src0->type == GGML_TYPE_Q8_0;
// The integer intrinsics used in mul_mat_vec_q are available with compute capability 6.
// However, they have bad performance with Pascal cards.
// Therefore, in a multi GPU setting decide at runtime which GPUs should use mul_mat_vec_q.
const bool use_mul_mat_vec_q = g_compute_capabilities[id] >= 700 && mul_mat_vec_q_implemented;
#endif
if (use_mul_mat_vec_q) {
size_t as;
void * src1_q8_1 = ggml_cuda_pool_malloc(ne00*sizeof(block_q8_1)/QK8_1, &as);
quantize_row_q8_1_cuda(src1_ddf_i, src1_q8_1, ne00, cudaStream_main);
switch (src0->type) {
case GGML_TYPE_Q4_0:
mul_mat_vec_q4_0_q8_1_cuda(src0_ddq_i, src1_q8_1, dst_ddf_i, ne00, nrows, cudaStream_main);
break;
case GGML_TYPE_Q4_1:
mul_mat_vec_q4_1_q8_1_cuda(src0_ddq_i, src1_q8_1, dst_ddf_i, ne00, nrows, cudaStream_main);
break;
case GGML_TYPE_Q5_0:
mul_mat_vec_q5_0_q8_1_cuda(src0_ddq_i, src1_q8_1, dst_ddf_i, ne00, nrows, cudaStream_main);
break;
case GGML_TYPE_Q5_1:
mul_mat_vec_q5_1_q8_1_cuda(src0_ddq_i, src1_q8_1, dst_ddf_i, ne00, nrows, cudaStream_main);
break;
case GGML_TYPE_Q8_0:
mul_mat_vec_q8_0_q8_1_cuda(src0_ddq_i, src1_q8_1, dst_ddf_i, ne00, nrows, cudaStream_main);
break;
default:
GGML_ASSERT(false);
break;
}
ggml_cuda_pool_free(src1_q8_1, as);
} else {
// on some GPUs it is faster to convert src1 to half and to use half precision intrinsics
#ifdef GGML_CUDA_DMMV_F16
size_t ash;
dfloat * src1_dfloat = nullptr; // dfloat == half
bool src1_convert_f16 = src0->type == GGML_TYPE_Q4_0 || src0->type == GGML_TYPE_Q4_1 ||
src0->type == GGML_TYPE_Q5_0 || src0->type == GGML_TYPE_Q5_1 ||
src0->type == GGML_TYPE_Q8_0 || src0->type == GGML_TYPE_F16;
if (src1_convert_f16) {
src1_dfloat = (half *) ggml_cuda_pool_malloc(ne00*sizeof(half), &ash);
ggml_cpy_f32_f16_cuda((char *) src1_ddf_i, (char *) src1_dfloat, ne00,
ne00, 1, sizeof(float), 0, 0,
ne00, 1, sizeof(half), 0, 0, cudaStream_main);
}
#else
dfloat * src1_dfloat = src1_ddf_i; // dfloat == float, no conversion
#endif // GGML_CUDA_DMMV_F16 #endif // GGML_CUDA_DMMV_F16
switch (src0->type) { switch (src0->type) {
case GGML_TYPE_Q4_0: case GGML_TYPE_Q4_0:
dequantize_mul_mat_vec_q4_0_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q4_0_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_Q4_1: case GGML_TYPE_Q4_1:
dequantize_mul_mat_vec_q4_1_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q4_1_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_Q5_0: case GGML_TYPE_Q5_0:
dequantize_mul_mat_vec_q5_0_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q5_0_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_Q5_1: case GGML_TYPE_Q5_1:
dequantize_mul_mat_vec_q5_1_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q5_1_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_Q8_0: case GGML_TYPE_Q8_0:
dequantize_mul_mat_vec_q8_0_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q8_0_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_Q2_K: case GGML_TYPE_Q2_K:
dequantize_mul_mat_vec_q2_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q2_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_Q3_K: case GGML_TYPE_Q3_K:
dequantize_mul_mat_vec_q3_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q3_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_Q4_K: case GGML_TYPE_Q4_K:
dequantize_mul_mat_vec_q4_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q4_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_Q5_K: case GGML_TYPE_Q5_K:
dequantize_mul_mat_vec_q5_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q5_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_Q6_K: case GGML_TYPE_Q6_K:
dequantize_mul_mat_vec_q6_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main); dequantize_mul_mat_vec_q6_K_cuda(src0_ddq_i, src1_ddf_i, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
case GGML_TYPE_F16: case GGML_TYPE_F16:
convert_mul_mat_vec_f16_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main); convert_mul_mat_vec_f16_cuda(src0_ddq_i, src1_dfloat, dst_ddf_i, ne00, nrows, cudaStream_main);
break; break;
default: default:
GGML_ASSERT(false); GGML_ASSERT(false);
break; break;
} }
#ifdef GGML_CUDA_DMMV_F16 #ifdef GGML_CUDA_DMMV_F16
if (src1_convert_f16) { if (src1_convert_f16) {
ggml_cuda_pool_free(src1_dfloat, ash); ggml_cuda_pool_free(src1_dfloat, ash);
} }
#endif // GGML_CUDA_DMMV_F16 #endif // GGML_CUDA_DMMV_F16
}
(void) src1; (void) src1;
(void) dst; (void) dst;
@ -2701,8 +3022,8 @@ void ggml_cuda_mul_mat(const ggml_tensor * src0, const ggml_tensor * src1, ggml_
}else if (src0->type == GGML_TYPE_F32) { }else if (src0->type == GGML_TYPE_F32) {
ggml_cuda_op(src0, src1, dst, ggml_cuda_op_mul_mat_cublas, true, false); ggml_cuda_op(src0, src1, dst, ggml_cuda_op_mul_mat_cublas, true, false);
} else if (ggml_is_quantized(src0->type) || src0->type == GGML_TYPE_F16) { } else if (ggml_is_quantized(src0->type) || src0->type == GGML_TYPE_F16) {
if (src1->ne[1] == 1 && src0->ne[0] % GGML_CUDA_DMMV_X == 0 && src0->ne[1] % GGML_CUDA_DMMV_Y == 0) { if (src1->ne[1] == 1 && src0->ne[0] % GGML_CUDA_DMMV_X == 0) {
ggml_cuda_op(src0, src1, dst, ggml_cuda_op_dequantize_mul_mat_vec, false, false); ggml_cuda_op(src0, src1, dst, ggml_cuda_op_mul_mat_vec, false, false);
} else { } else {
ggml_cuda_op(src0, src1, dst, ggml_cuda_op_mul_mat_cublas, true, false); ggml_cuda_op(src0, src1, dst, ggml_cuda_op_mul_mat_cublas, true, false);
} }