Apple M1 Microarchitecture Research by Dougall Johnson

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

PMUL (8B)

Test 1: uops

Code:

  pmul v0.8b, v0.8b, v1.8b
  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)a8cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
10043037230612548251000100010003983130301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037230612548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037223612548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037220612548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037230612548251000100010003983130301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037230822548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037230612548251000100010003983130301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037220612548251000100010003983130301830373037241532895100010002000303730371110011000173116112630100030383038303830383038
100430372275822548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037220612548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038

Test 2: Latency 1->2

Code:

  pmul v0.8b, v0.8b, v1.8b
  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)03080b1e3f4e51inst 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)9fa1a8a9acc5branch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250006129548251010010010000100100005004277313130018030037300372826532874510100200100002002000030037301331110201100991001001000010000001075711613296340100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313130018030037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
10204300372240006129548251010010010000100100005004277313130018030037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313130018030037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313130018030037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313130018030037300372826532874510100200103352002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313130018030037300372826532874510100200100002002000030037300371110201100991001001000010000000071011615296340100001003003830038300383003830038
102043003722510186129548251010010010000100100005004277313130018030037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
102043003722500310329548251010010010000100100005004277313130018330037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313130018030037300372826532874510100200100002002000030037300371110201100991001001000010000000071011611296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0308090b18191e1f3f4e51inst 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)9fa0a1a6a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372240000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
100243003722500000001602954825100101010000101000050427731313001803003730037282873287671001020103232020000300373003711100211091010100001000000006402162229630010000103003830085301823008630038
10024300372250000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731303001803003730037282873287671001022100002020000300833003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372240000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
10024300372250000000612954825100101010000101000050427731313001803003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038

Test 3: Latency 1->3

Code:

  pmul v0.8b, v1.8b, v0.8b
  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)9fa6a9c2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102043003722400719295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
10204300372250061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
10204300372250061295482510100100100001001000050042773130300183003730037282653287451010021210000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
10204300372240061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722506061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
10204300372250061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
10204300372250061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
10204300372250061295482510100100100001001000050042773130300183003730037282653287451010020010000214200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
102043003722503661295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038
10204300372250061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000007101161129634100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03080918191e1f3f4e51inst 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)9fa0a1a6a8acc2cdcfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722500000061295482510010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001000000006403163329630010000103003830038300383003830038
100243003722500000061295482510010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001000000006403163329630010000103003830038300863017930085
100243003722501220061295482510010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001000000006403163329630010000103003830038300383003830038
100243003722500000061295482510010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001000000006403163329630010000103003830038300383003830038
10024300842250122156104103294941431006915100481110298654286812130270303203036628311322889610906221098524222763032230366811002110910101000010002016888207453726729884410000103041730371304163036930369
100243036822701779366164640294851581007212100481711043884286812130270303683036528320342880211059221114224206683036730369811002110910101000010002119398007734405529882210000103041330367303203037230368
10024303692271177807616339229485161100691210056161089478428692413027030406304172832032876711206201130524229423046130461211002110910101000010460011133007667643429883110000103041530416304173032330320
1002430037228016700251295482510010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001000000006403163329630010000103003830038300383003830038
100243003722500000061295484410010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001000000006403163329630010000103003830038300383003830038
1002430037225000000726295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000000006403163329630010000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  pmul v0.8b, v8.8b, v9.8b
  pmul v1.8b, v8.8b, v9.8b
  pmul v2.8b, v8.8b, v9.8b
  pmul v3.8b, v8.8b, v9.8b
  pmul v4.8b, v8.8b, v9.8b
  pmul v5.8b, v8.8b, v9.8b
  pmul v6.8b, v8.8b, v9.8b
  pmul v7.8b, v8.8b, v9.8b
  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)03080b18191e1f3f51inst 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)9fa1a6a8a9acc5cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
802042005915000000041258010010080000100800005006400000200202003920039997339997801002008000020016000020039200391180201100991001008000010000005750511021611200360800001002004020040200402004020040
8020420039150000000412580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000000511011611200360800001002004020040200402004020040
8020420039150000000412580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000000511011611200360800001002004020040200402004020040
8020420039150000000412580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000000511011611200360800001002004020040200402004020040
8020420039150000000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000000511011611200360800001002004020040200402004020040
8020420039150000000412580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000000511011611200360800001002004020040200402004020040
80204200391500000180832580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000000511011611200360800001002004020040200402004020040
80204200391500000210412580100100800001008000050064000002002020039200399973399978010020080000200160000200392003911802011009910010080000100000000511011611200360800001002004020040200402004020040
8020420039150000000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000000511011611200360800001002004020040200402004020040
8020420039150000000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000001511021611200360800001002004020040200402004020040

1000 unrolls and 10 iterations

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

retire (01)cycle (02)0307080a0b18191e3a3f51inst 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)9fa0a1a6a8accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420048150101000024725800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100000050242216232320036080000102004020040200402004020040
8002420039150101000024725800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100000050242316232220036080000102004020040200402004020040
8002420039150101000024725800101080000108000050640000020062200392003999963100198001020800002016000020039200391180021109101080000100001350242316222220036080000102004020040200402004020040
80024201101501020000247258001010800001280000506400000200202003920039999631001980010208000020160000200392003911800211091010800001000008750242216192120036080000102004020040200402004020040
8002420039150101000024725800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100000350242216232320036080000102004020040200402004020040
8002420039150101000024725800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100000050241616221920077080000102004020040200402004020040
8002420039150101000024725800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100000050242316222220036080000102004020040200402004020040
8002420039150101000024725800101080000108000050640000120020200392003999963100198001020800002016000020039200391180021109101080000100000050242316242420036080000102004020040200402004020040
8002420039150101000024725800101080000108000050640000020020200392003999967100198001020800002016000020039200391180021109101080000100000050242216222320036080000102004020040200402009420040
8002420039150101000024725800101080000108000050640000020020200392003999963100198001020800002016000020039200391180021109101080000100000050242316212420036080000102004020040200402004020040