alpaka
Abstraction Library for Parallel Kernel Acceleration
SyclSubgroupSize.hpp
Go to the documentation of this file.
1
/* Copyright 2023 Andrea Bocci, Aurora Perego
2
* SPDX-License-Identifier: MPL-2.0
3
*/
4
5
#ifdef ALPAKA_ACC_SYCL_ENABLED
6
7
# ifdef __SYCL_DEVICE_ONLY__
8
9
# if(__SYCL_TARGET_INTEL_GPU_BDW__) ||
/* Broadwell Intel graphics architecture */
\
10
(__SYCL_TARGET_INTEL_GPU_SKL__) ||
/* Skylake Intel graphics architecture */
\
11
(__SYCL_TARGET_INTEL_GPU_KBL__) ||
/* Kaby Lake Intel graphics architecture */
\
12
(__SYCL_TARGET_INTEL_GPU_CFL__) ||
/* Coffee Lake Intel graphics architecture */
\
13
(__SYCL_TARGET_INTEL_GPU_APL__) ||
/* Apollo Lake Intel graphics architecture */
\
14
(__SYCL_TARGET_INTEL_GPU_GLK__) ||
/* Gemini Lake Intel graphics architecture */
\
15
(__SYCL_TARGET_INTEL_GPU_WHL__) ||
/* Whiskey Lake Intel graphics architecture */
\
16
(__SYCL_TARGET_INTEL_GPU_AML__) ||
/* Amber Lake Intel graphics architecture */
\
17
(__SYCL_TARGET_INTEL_GPU_CML__) ||
/* Comet Lake Intel graphics architecture */
\
18
(__SYCL_TARGET_INTEL_GPU_ICLLP__) ||
/* Ice Lake Intel graphics architecture */
\
19
(__SYCL_TARGET_INTEL_GPU_EHL__) ||
/* Elkhart Lake or Jasper Lake Intel graphics architecture */
\
20
(__SYCL_TARGET_INTEL_GPU_TGLLP__) ||
/* Tiger Lake Intel graphics architecture */
\
21
(__SYCL_TARGET_INTEL_GPU_RKL__) ||
/* Rocket Lake Intel graphics architecture */
\
22
(__SYCL_TARGET_INTEL_GPU_ADL_S__) ||
/* Alder Lake S or Raptor Lake S Intel graphics architecture */
\
23
(__SYCL_TARGET_INTEL_GPU_ADL_P__) ||
/* Alder Lake P Intel graphics architecture */
\
24
(__SYCL_TARGET_INTEL_GPU_ADL_N__) ||
/* Alder Lake N Intel graphics architecture */
\
25
(__SYCL_TARGET_INTEL_GPU_DG1__) ||
/* DG1 Intel graphics architecture */
\
26
(__SYCL_TARGET_INTEL_GPU_ACM_G10__) ||
/* Alchemist G10 Intel graphics architecture */
\
27
(__SYCL_TARGET_INTEL_GPU_ACM_G11__) ||
/* Alchemist G11 Intel graphics architecture */
\
28
(__SYCL_TARGET_INTEL_GPU_ACM_G12__) ||
/* Alchemist G12 Intel graphics architecture */
\
29
(__SYCL_TARGET_INTEL_GPU_MTL_U__) ||
/* Meteor Lake U/S or Arrow Lake U/S Intel graphics architecture */
\
30
(__SYCL_TARGET_INTEL_GPU_MTL_H__) ||
/* Meteor Lake H Intel graphics architecture */
\
31
(__SYCL_TARGET_INTEL_GPU_ARL_H__) ||
/* Arrow Lake H Intel graphics architecture */
\
32
(__SYCL_TARGET_INTEL_GPU_BMG_G21__) ||
/* Battlemage G21 Intel graphics architecture */
\
33
(__SYCL_TARGET_INTEL_GPU_LNL_M__)
/* Lunar Lake Intel graphics architecture */
34
35
# define SYCL_SUBGROUP_SIZE (8 | 16 | 32)
36
37
# elif(__SYCL_TARGET_INTEL_GPU_PVC__) ||
/* Ponte Vecchio Intel graphics architecture */
\
38
(__SYCL_TARGET_INTEL_GPU_PVC_VG__)
/* Ponte Vecchio VG Intel graphics architecture */
39
40
# define SYCL_SUBGROUP_SIZE (16 | 32)
41
42
# elif(__SYCL_TARGET_INTEL_X86_64__)
/* generate code ahead of time for x86_64 CPUs */
43
44
# define SYCL_SUBGROUP_SIZE (4 | 8 | 16 | 32 | 64)
45
46
# elif(__SYCL_TARGET_NVIDIA_GPU_SM50__) ||
/* NVIDIA Maxwell architecture (compute capability 5.0) */
\
47
(__SYCL_TARGET_NVIDIA_GPU_SM52__) ||
/* NVIDIA Maxwell architecture (compute capability 5.2) */
\
48
(__SYCL_TARGET_NVIDIA_GPU_SM53__) ||
/* NVIDIA Jetson TX1 / Nano (compute capability 5.3) */
\
49
(__SYCL_TARGET_NVIDIA_GPU_SM60__) ||
/* NVIDIA Pascal architecture (compute capability 6.0) */
\
50
(__SYCL_TARGET_NVIDIA_GPU_SM61__) ||
/* NVIDIA Pascal architecture (compute capability 6.1) */
\
51
(__SYCL_TARGET_NVIDIA_GPU_SM62__) ||
/* NVIDIA Jetson TX2 (compute capability 6.2) */
\
52
(__SYCL_TARGET_NVIDIA_GPU_SM70__) ||
/* NVIDIA Volta architecture (compute capability 7.0) */
\
53
(__SYCL_TARGET_NVIDIA_GPU_SM72__) ||
/* NVIDIA Jetson AGX (compute capability 7.2) */
\
54
(__SYCL_TARGET_NVIDIA_GPU_SM75__) ||
/* NVIDIA Turing architecture (compute capability 7.5) */
\
55
(__SYCL_TARGET_NVIDIA_GPU_SM80__) ||
/* NVIDIA Ampere architecture (compute capability 8.0) */
\
56
(__SYCL_TARGET_NVIDIA_GPU_SM86__) ||
/* NVIDIA Ampere architecture (compute capability 8.6) */
\
57
(__SYCL_TARGET_NVIDIA_GPU_SM87__) ||
/* NVIDIA Jetson/Drive AGX Orin (compute capability 8.7) */
\
58
(__SYCL_TARGET_NVIDIA_GPU_SM89__) ||
/* NVIDIA Ada Lovelace arch. (compute capability 8.9) */
\
59
(__SYCL_TARGET_NVIDIA_GPU_SM90__)
/* NVIDIA Hopper architecture (compute capability 9.0) */
60
61
# define SYCL_SUBGROUP_SIZE (32)
62
63
# elif(__SYCL_TARGET_AMD_GPU_GFX700__) ||
/* AMD GCN 2.0 Sea Islands architecture (gfx 7.0) */
\
64
(__SYCL_TARGET_AMD_GPU_GFX701__) ||
/* AMD GCN 2.0 Sea Islands architecture (gfx 7.0) */
\
65
(__SYCL_TARGET_AMD_GPU_GFX702__) ||
/* AMD GCN 2.0 Sea Islands architecture (gfx 7.0) */
\
66
(__SYCL_TARGET_AMD_GPU_GFX801__) ||
/* AMD GCN 3.0 Volcanic Islands architecture (gfx 8.0) */
\
67
(__SYCL_TARGET_AMD_GPU_GFX802__) ||
/* AMD GCN 3.0 Volcanic Islands architecture (gfx 8.0) */
\
68
(__SYCL_TARGET_AMD_GPU_GFX803__) ||
/* AMD GCN 4.0 Arctic Islands architecture (gfx 8.0) */
\
69
(__SYCL_TARGET_AMD_GPU_GFX805__) ||
/* AMD GCN 3.0 Volcanic Islands architecture (gfx 8.0) */
\
70
(__SYCL_TARGET_AMD_GPU_GFX810__) ||
/* AMD GCN 3.0 Volcanic Islands architecture (gfx 8.1) */
\
71
(__SYCL_TARGET_AMD_GPU_GFX900__) ||
/* AMD GCN 5.0 Vega architecture (gfx 9.0) */
\
72
(__SYCL_TARGET_AMD_GPU_GFX902__) ||
/* AMD GCN 5.0 Vega architecture (gfx 9.0) */
\
73
(__SYCL_TARGET_AMD_GPU_GFX904__) ||
/* AMD GCN 5.0 Vega architecture (gfx 9.0) */
\
74
(__SYCL_TARGET_AMD_GPU_GFX906__) ||
/* AMD GCN 5.1 Vega II architecture (gfx 9.0) */
\
75
(__SYCL_TARGET_AMD_GPU_GFX908__) ||
/* AMD CDNA 1.0 Arcturus architecture (gfx 9.0) */
\
76
(__SYCL_TARGET_AMD_GPU_GFX909__) ||
/* AMD GCN 5.0 Raven 2 architecture (gfx 9.0) */
\
77
(__SYCL_TARGET_AMD_GPU_GFX90A__) ||
/* AMD CDNA 2.0 Aldebaran architecture (gfx 9.0) */
\
78
(__SYCL_TARGET_AMD_GPU_GFX90C__) ||
/* AMD GCN 5.1 Renoir architecture (gfx 9.0) */
\
79
(__SYCL_TARGET_AMD_GPU_GFX940__) ||
/* AMD CDNA 3.0 Aqua Vanjaram architecture (gfx 9.4) */
\
80
(__SYCL_TARGET_AMD_GPU_GFX941__) ||
/* AMD CDNA 3.0 Aqua Vanjaram architecture (gfx 9.4) */
\
81
(__SYCL_TARGET_AMD_GPU_GFX942__)
/* AMD CDNA 3.0 Aqua Vanjaram architecture (gfx 9.4) */
82
83
# define SYCL_SUBGROUP_SIZE (64)
84
85
# elif(__SYCL_TARGET_AMD_GPU_GFX1010__) ||
/* AMD RDNA 1.0 Navi 10 architecture (gfx 10.1) */
\
86
(__SYCL_TARGET_AMD_GPU_GFX1011__) ||
/* AMD RDNA 1.0 Navi 12 architecture (gfx 10.1) */
\
87
(__SYCL_TARGET_AMD_GPU_GFX1012__) ||
/* AMD RDNA 1.0 Navi 14 architecture (gfx 10.1) */
\
88
(__SYCL_TARGET_AMD_GPU_GFX1013__) ||
/* AMD RDNA 2.0 Oberon architecture (gfx 10.1) */
\
89
(__SYCL_TARGET_AMD_GPU_GFX1030__) ||
/* AMD RDNA 2.0 Navi 21 architecture (gfx 10.3) */
\
90
(__SYCL_TARGET_AMD_GPU_GFX1031__) ||
/* AMD RDNA 2.0 Navi 22 architecture (gfx 10.3) */
\
91
(__SYCL_TARGET_AMD_GPU_GFX1032__) ||
/* AMD RDNA 2.0 Navi 23 architecture (gfx 10.3) */
\
92
(__SYCL_TARGET_AMD_GPU_GFX1033__) ||
/* AMD RDNA 2.0 Van Gogh architecture (gfx 10.3) */
\
93
(__SYCL_TARGET_AMD_GPU_GFX1034__) ||
/* AMD RDNA 2.0 Navi 24 architecture (gfx 10.3) */
\
94
(__SYCL_TARGET_AMD_GPU_GFX1035__) ||
/* AMD RDNA 2.0 Rembrandt Mobile architecture (gfx 10.3) */
\
95
(__SYCL_TARGET_AMD_GPU_GFX1036__) ||
/* AMD RDNA 2.0 Raphael architecture (gfx 10.3) */
\
96
(__SYCL_TARGET_AMD_GPU_GFX1100__) ||
/* AMD RDNA 3.0 Navi 31 architecture (gfx 11.0) */
\
97
(__SYCL_TARGET_AMD_GPU_GFX1101__) ||
/* AMD RDNA 3.0 Navi 32 architecture (gfx 11.0) */
\
98
(__SYCL_TARGET_AMD_GPU_GFX1102__) ||
/* AMD RDNA 3.0 Navi 33 architecture (gfx 11.0) */
\
99
(__SYCL_TARGET_AMD_GPU_GFX1103__) ||
/* AMD RDNA 3.0 Phoenix mobile architecture (gfx 11.0) */
\
100
(__SYCL_TARGET_AMD_GPU_GFX1150__) ||
/* AMD RDNA 3.5 Strix Point architecture (gfx 11.5) */
\
101
(__SYCL_TARGET_AMD_GPU_GFX1151__) ||
/* AMD RDNA 3.5 Strix Halo architecture (gfx 11.5) */
\
102
(__SYCL_TARGET_AMD_GPU_GFX1200__) ||
/* AMD RDNA 4.0 Navi 44 architecture (gfx 12.0) */
\
103
(__SYCL_TARGET_AMD_GPU_GFX1201__)
/* AMD RDNA 4.0 Navi 48 architecture (gfx 12.0) */
104
105
// starting from gfx10, HIP supports only wavefront size 32
106
# define SYCL_SUBGROUP_SIZE (32)
107
108
# else
// __SYCL_TARGET_*
109
110
# define SYCL_SUBGROUP_SIZE (0)
/* unknown target */
111
112
# endif
// __SYCL_TARGET_*
113
114
# else
115
116
# define SYCL_SUBGROUP_SIZE (0)
/* host compilation */
117
118
# endif
// __SYCL_DEVICE_ONLY__
119
120
#endif
// ALPAKA_ACC_SYCL_ENABLED
include
alpaka
kernel
SyclSubgroupSize.hpp
Generated on Thu Dec 12 2024 11:51:45 for alpaka by
1.9.1
1.9.1