Apple M1 Microarchitecture Research by Dougall Johnson

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

SQRDMULH (by element, 4H)

Test 1: uops

Code:

  sqrdmulh v0.4h, v0.4h, 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)60696d6edispatch 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)a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
100430372308225482510001000100039831303018303730372415328951000100020003037303711100110000073216112630100030383038303830773038
100430372206125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830853038
100430372206125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372206125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372206125482510001000100039831313018303730372415328951000100020003037303711100110000073116112682100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
100430372306125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038

Test 2: Latency 1->2

Code:

  sqrdmulh v0.4h, v0.4h, 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)03080b18191e1f3a3f4e51inst 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)9fa1a8acc2cfd5d6dbddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372240000420061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296340100001003003830038300383003830038
10204300372250000390061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296340100001003003830038300383003830038
10204300372250000540061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296340100001003003830038300383003830038
10204300372250011420061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296340100001003003830038300383003830038
1020430037225000090061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296340100001003003830038300383003830038
1020430037225000060061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296340100001003003830038300383003830038
1020430037225000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296340100001003003830038300383003830038
10204300372240000180061295482510100100100081001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296345100001003003830038300383003830038
10204300372250000390061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296340100001003003830038300383003830038
1020430037225000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000710216022296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0307080b18191e1f3a3f4e51inst 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)9fa0a8acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722500000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
100243003722500000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000007212162229630010000103003830038300383003830038
100243003722500000000612954825100101010000101000050427731303001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
100243003722500000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
100243003722400000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
100243003722500000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
100243003722500000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
100243003722500000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
100243003722400000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000006402162229630010000103003830038300383003830038
100243003722500000900612953016410060101004811110438242854551302700303693031828315332889810908201016820223003036830367911002110910101000010411756327863942329882510000103036730370303683040730371

Test 3: Latency 1->3

Code:

  sqrdmulh v0.4h, v1.4h, 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)030b1e3f4e51inst 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)9fa8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250072629548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100037101161129634100001003003830038300383003830038
10204300372250061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001001667101161129634100001003003830038300383003830038
1020430037225006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100037101161129634100001003003830038300383003830038
10204300372250061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000547101161129634100001003003830038300383003830038
10204300372250061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000367101161129634100001003003830038300383003830038
1020430037224006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100007101161129634100001003003830038300383003830038
10204300372250025129548251010010010000100100005004277313030018300373008428265328745101002001000020020000300373003711102011009910010010000100037101161129634100001003003830038300383003830038
1020430037225006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100037101161129634100001003003830038300383003830038
1020430037241006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100037101161129634100001003003830038300383003830038
1020430037225006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100067101161129634100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03191e3f4e51inst 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)9fa8acc5cficache miss (d3)d5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722501261295482510010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001000064002162229630010000103003830038300383003830038
1002430037225002705295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000064002162229630010000103003830038300383003830038
10024300372250061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000064002162229630010000103003830038300383003830038
100243003722500569295482510010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001000064002162229630010000103003830038300383003830038
10024300372250061295482510010101000010100005042773130300183003730037282913287671001020100002020000300373003711100211091010100001000064002162229630010000103003830038300383003830038
10024300372250061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000164002162229630010000103003830038300383003830038
10024300372250061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000064002162229630010000103003830038300383003830038
10024300372250061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000064002162229630010000103003830038300383003830038
10024300372250061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000064002162229630010000103003830038300383003830038
10024300372251061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000064002162229630010000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  sqrdmulh v0.4h, v8.4h, v9.h[1]
  sqrdmulh v1.4h, v8.4h, v9.h[1]
  sqrdmulh v2.4h, v8.4h, v9.h[1]
  sqrdmulh v3.4h, v8.4h, v9.h[1]
  sqrdmulh v4.4h, v8.4h, v9.h[1]
  sqrdmulh v5.4h, v8.4h, v9.h[1]
  sqrdmulh v6.4h, v8.4h, v9.h[1]
  sqrdmulh v7.4h, v8.4h, 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)031e3f51inst 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)6061696b6d6edispatch 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)9fa8a9acbranch mispredict (cb)cfd0d5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8020420071150041258010010080000100800005006400000020020020039200399973399978010020080000200160000200392003911802011009910010080000100340960511004162220036800001002004020040200402004020040
802042003915004125801001008000010080000500640000002002002003920039997339997801002008000020016000020039200391180201100991001008000010042000511002162220036800001002004020040200402004020040
8020420039150041258010010080000100800005006400000020020020039200399973399978010020080000200160000200392003911802011009910010080000100360630511002162220036800001002004020040200402004020040
8020420039157048425801001008000010080000500640000002002002003920039997339997801002008000020016000020039200391180201100991001008000010021000511002162220036800001002004020040200402004020040
80204200391500412580100100800001008000050064000000200200200392003999733999780100200800002001600002003920039118020110099100100800001000000511002162220036800001002004020040200402004020040
8020420039150041258010010080000100800005006400000020020020039200399973399978010020080000200160000200392003911802011009910010080000100001260511002162320036800001002004020040200402004020040
80204200391500412580100100800001008000050064000010200200200392003999733999780100200800002001600002003920039118020110099100100800001000030511002162220036800001002004020040200402004020040
80204200391500412580100100800001008000050064000000200200200392003999733999780100200800002001600002003920039118020110099100100800001000000511002162220036800001002004020040200402004020040
802042003915004125801001008000010080000500640000102002002003920039997339997801002008000020016000020039200391180201100991001008000010051030511002162220036800001002004020040200402004020040
802042003915004125801001008000010080000500640000002002002003920039997339997801002008000020016000020039200391180201100991001008000010029030511002162220036800001002004020040200402004020040

1000 unrolls and 10 iterations

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

retire (01)cycle (02)03081e3a3f51inst 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)9fa7a8acc5cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420048150000402580010108000010800005064000012002020039200399996310019800102080000201600002003920039118002110910108000010000005020516112003680000102004020040200402004020040
8002420039150000402580010108000010800005064000002014120039200399996310019800102080000201600002003920039118002110910108000010000005020116222003680000102004020040200402004020040
80024200391500004025800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100015005020116112003680000102004020040200402004020040
800242003915000040258001010800001080000506400001200202003920039999631001980010208000020160000200392003911800211091010800001000129005020216222003680000102004020040200402004020040
80024200391500004025800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100090005020216222003680000102004020040200402004020040
8002420039150000402580010108000010800005064000012002020039200399996310019800102080000201600002003920039118002110910108000010000005020216222003680000102004020040200402004020040
80024200391500004025800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100015005020116112003680000102004020040200402004020040
80024200391500004025800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100018005020216222003680000102004020040200402004020040
80024200391500004025800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100015005020116112003680000102004020040200402004020040
8002420039150000402580010108000010800005064000012002020039200399996310019800102080000201600002003920039118002110910108000010003005020116112003680000102004020040200402004020040