Apple M1 Microarchitecture Research by Dougall Johnson

Firestorm: Overview | Base Instructions | SIMD and FP Instructions
Icestorm:  Overview | Base Instructions | SIMD and FP Instructions

SQDMULH (by element, H)

Test 1: uops

Code:

  sqdmulh h0, h0, v1.h[1]
  movi v0.16b, 1
  movi v1.16b, 2

(no loop instructions)

1000 unrolls and 1 iteration

Retires: 1.000

Issues: 1.000

Integer unit issues: 0.000

Load/store unit issues: 0.000

SIMD/FP unit issues: 1.000

retire (01)cycle (02)031e3f4e51inst issue (52)~issue fp/simd (54)~dispatch fp/simd (57)huge thing fp/simd (5b)696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op fp/simd (7e)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst neon or fp (9a)a0a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
1004303723019425482510001000100039831330183037303724153289510001000200030373037111001100000073316222630100030383038303830383038
100430372206125482510001000100039831330183037303724153289510001000200030373037111001100000073216222630100030383038303830383038
100430372206125482510001000100039831330183037303724153289510001000200030373037111001100000073216222630100030383038303830383038
100430372206125484510001000100039831330183037303724153289510001000200030373037111001100000073216222630100030383038303830383038
1004303723336125482510001000100039831330183037303724153289510001000200030373037111001100000073216222630100030383038303830383038
100430372206125482510001000100039831330183037303724153289510001000200030373037111001100000073216222630100030383038303830383038
100430372306125482510001000100039831330183037303724153289510001000200030373037111001100000073216222630100030383038303830383038
100430372306125482510001000100039831330183037303724153291410001000200030373037111001100000073216222630100030383038303830383038
100430372306125482510001000100039831330183037303724153289510001000200030373037111001100000073216222630100030383038303830383038
100430372206125482510001000100039831330183037303724153289510001000200030373037111001100000073216222630100030383038303830383038

Test 2: Latency 1->2

Code:

  sqdmulh h0, h0, v1.h[1]
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03070918191e1f3a3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa1a8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000071011611296340100001003003830038300383003830038
10204300372250000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000071011611296340100001003003830038300383003830038
10204300372250000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000071011611296340100001003003830038300383003830038
10204300372240000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000071011611296340100001003003830038300383003830038
10204300372250000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000071011611296340100001003003830038300383003830038
10204300372240000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000071011611296340100001003003830038300383003830038
10204300372240000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000071011611296340100001003003830038300383003830038
10204300372250000000612954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000071011611296340100001003003830038300383003830038
102043003722500000004412954825101001001000010010000500427731303001830037300372826532874510100200100002002000030037300371110201100991001001000010000071011611297060100001003003830038300863032630327
102043022722701170108211272954825101001001000011410447616427731303001830037300852826532874510100200100002002000030037300371110201100991001001000010000071011611296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03080b18191e1f3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696b6d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa0a6a8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225000000612954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000006403161229630010000103003830038300383003830038
10024300372240000001602954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731303016203003730037282873287671001020100002020000300373003711100211091010100001000006612162229630010000103003830038300383003830038
1002430037224000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
10024300372250000008322954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037232000000612954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731313012603003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
1002430037224000000612954825100101010000101000050427773503001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038

Test 3: Latency 1->3

Code:

  sqdmulh h0, h1, v0.h[1]
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03080b18191e1f3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa1a8a9accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250000006129548251010010010000100100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100000071011610296340100001003003830038300383003830038
10204300372250000006129548251010010010000100100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100000071011611296340100001003003830038300383003830038
10204300372250000006129548251010010010000100100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100000071011611296340100001003003830038300383003830038
102043003722500000061295482510113100100001001000050042786701300183003730037282653287451010020010000200200003003730037111020110099100100100001000201271011611296340100001003003830038300383003830038
102043003722500000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000550971011611296340100001003003830038300383003830038
1020430037225000003526129548251010010010000100100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100610071011611296340100001003003830038300383003830038
102043003722500003906129548251010010010000100100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100030071011611296340100001003003830038300383003830038
10204300372240000006129548251010010010000128100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100042071011611296340100001003022730038300383003830038
102043003722500000018929548251010010010000100100005004277313130018300373003728265328745101002001000020020000301333003711102011009910010010000100000071011611296340100001003003830038300383003830038
10204300372250100006129548251010010010000100100005004277313130018300373003728265328745101002001000020020000300373003711102011009910010010000100010071011611296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0307080b181e1f3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa7a8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372250000006129548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006403162229630010000103003830038300383003830038
10024300372250000006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372240000006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250000006129548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250000006129548251001010100001010000504277313130018300373003728287328767100102010000202033630037300371110021109101010000100006402162229630010000103003830038300383003830038
100243003722500000098229548251001010100001010000504278670030018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250000006129548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250000006129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250000006129548251001010100001010000504277313130018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038
10024300372250000606129548251001010100001010000504277313030018300373003728287328767100102010000202000030037300371110021109101010000100006402162229630010000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  sqdmulh h0, h8, v9.h[1]
  sqdmulh h1, h8, v9.h[1]
  sqdmulh h2, h8, v9.h[1]
  sqdmulh h3, h8, v9.h[1]
  sqdmulh h4, h8, v9.h[1]
  sqdmulh h5, h8, v9.h[1]
  sqdmulh h6, h8, v9.h[1]
  sqdmulh h7, h8, v9.h[1]
  movi v8.16b, 9
  movi v9.16b, 10

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)030b1e3f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa8acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
802042006015000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000511021611200360800001002004020040200902004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100200511011611200360800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000511011611200360800001002019620040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000511011611200360800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100130511011623200360800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000511011611200360800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000511011611200360800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000511011611200360800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000511011611200360800001002004020040200402004020040
802042003915010412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000511011611200360800001002004020040200402004020040

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)0307080a0b18191e1f3a3f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa1a6a7a8a9acc2cfd2d5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
800242004815000000000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001000021060502001229171720036080000102004020040200402004020040
800242003915600000000040258001010800001080000506400000200202003920039999631001980010208000020160000200392003911800211091010800001000036000502001716171720036080000102004020040200402004020040
800242008915000000000040258001010800001080000506400000200202003920111100143100198001020800002016000020039200391180021109101080000100002300050200616171720036080000102004020040200402004020040
8002420039150000000000402580010108000010800005064000002002020039200399996310019800102080000201600002003920039118002110910108000010000420186050200171617720036080000102004020040200402004020040
80024200391500000000004025800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100000000502001716161620036080000102004020040200402004020040
800242003915000000000082258001010800001080000506400000200202003920039999631001980010208000020160000200392003911800211091010800001000035030502001716221820036080000102024820040200402004020040
8002420039150000000000402580010108000010800005064000002002020039200399996310019800102080000201600002003920039118002110910108000010000250120502001416171720036080000102004020040200402004020040
80024202441500000000004025800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100002500050200171614620036080000102004020040200402004020040
80024200391500000000004025800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100003800050200716141720036080000102004020040200402004020040
800242003915000000000021025800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100002103050200191617820036080000102004020040200402004020040