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, 8H)

Test 1: uops

Code:

  sqdmulh v0.8h, v0.8h, 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)03081e3f4e51inst 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
1004303722006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723106125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303722006125482510001000100039831303018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
10043037230156125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303722006125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415728951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830383038
1004303723006125482510001000100039831313018303730372415328951000100020003037303711100110000073116112630100030383038303830853038

Test 2: Latency 1->2

Code:

  sqdmulh v0.8h, v0.8h, 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)030a1e1f3f4e51inst 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
10204300372240006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373008411102011009910010010000100007101161129634100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100007101161129634100001003003830038300383003830038
10204300372250008229548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100007101161129634100001003003830038300383003830038
10204300372240006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100007101161129634100001003003830038300383003830038
102043003722500025429539251010010010000100100005004277313130018300373003728265328745101002001000020020000300373017921102011009910010010000100007101161129634100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100007101161129634100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100007101161129634100001003003830038300383003830038
102043003722500074529548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100007101161129634100001003003830038300383003830038
10204300372250006129548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100007101161129634100001003003830038300383003830038
102043003722500023229548251010010010000100100005004277313030018300373003728265328745101002001000020020000300373003711102011009910010010000100007101161129634100001003003830038300383003830038

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)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)9fa0a1a6a8a9acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1002430037225000000156295482510010101000010100005042773131300183003730037282873287671001020100002020000300373003711100211091010100001000033016506404162229630010000103003830038300383003830038
1002430037225000000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000009606402162229630010000103003830038300383003830038
1002430037224000000612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010000002106682162229630010000103003830038300383003830038
10024300372250000006129548251001010100001010596504277313030018300373003728287102878610010201000020200003003730037111002110910101000010000003306402162229630010000103003830038300383003830038
10024300372250000005362954825100101010000101000050427731313001830037300372828732876710010201000020203443003730037111002110910101000010000550006402162229630010000103003830038300383003830038
1002430037224000000612954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000490306402162229630010000103003830038300383003830038
10024300372250000007262954825100101010000101000050427731313001830037300372830732876710010201000020200003003730037111002110910101000010000450006402162229630010000103003830038300383003830038
10024300372250000002512954825100101010000101000050427731303001830037300372828732876710010201000020200003003730037111002110910101000010000305106402172229630010000103003830038300383003830038
10024300372250000007262954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010000470306402162229630010000103003830038300383003830038
1002430037224000000612954825100101010000101000050427731313001830037300372828732876710010201000020200003003730037111002110910101000010000400006402162229630010000103003830038300383003830038

Test 3: Latency 1->3

Code:

  sqdmulh v0.8h, v1.8h, 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)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)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)9fa1a6a8a9accfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020430037224000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
10204300372250000000231295486310100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000000061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
10204300372250000000138295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830085300383003830038
1020430037225000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000000061295482510100100100001001000050042773130300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038
1020430037225000000061295482510100100100001001000050042773131300183003730037282653287451010020010000200200003003730037111020110099100100100001000000071011611296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0318191e3f4e51inst 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)696d6edispatch 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
10024300372250006129548251001010100001010000504277313300183003730037282923287671001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313300183003730037282873287671001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038
10024300372250006129548251001010100001010000504277313300183003730037282873287871001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038
10024300372250006129548251001010100001010000504278045300183003730037282873287671001020100002020000300373003711100211091010100001000640316332963010000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  sqdmulh v0.8h, v8.8h, v9.h[1]
  sqdmulh v1.8h, v8.8h, v9.h[1]
  sqdmulh v2.8h, v8.8h, v9.h[1]
  sqdmulh v3.8h, v8.8h, v9.h[1]
  sqdmulh v4.8h, v8.8h, v9.h[1]
  sqdmulh v5.8h, v8.8h, v9.h[1]
  sqdmulh v6.8h, v8.8h, v9.h[1]
  sqdmulh v7.8h, v8.8h, 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)03191e3f51inst 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)9fa0a9acc2cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8020420049150039622580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000051102161120036800001002004020040200402004020040
802042019415000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000251271161120036800001002004020040200402004020040
802042011215010412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100220051101161120036800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
80204200391500525412580100100800001008000050064000012002020039200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040
802042003915000412580100100800001008000050064000012002020089200399973399978010020080000200160000200392003911802011009910010080000100000051101161120036800001002004020040200402004020040

1000 unrolls and 10 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)696d6edispatch 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)9fc2c5branch mispredict (cb)cficache miss (d3)d5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420039150040258001010800001080000506400002002020039200399996310019800102080000201600002003920039118002110910108000010000502052162220036080000102004020040200402004020040
80024200391500256258001010800001080000506400002002020039200399996310019800102080000201600002003920039118002110910108000010000502052162220036080000102004020040200402004020040
80024200391500680258001010800001080000506400002002020039200399996310019800102080000201600002003920039118002110910108000010000502042162220036080000102004020040200402004020040
8002420039150040258001010800001080000506400002002020039200399996310019800102080000201600002003920039118002110910108000010000502042162220036080000102004020040200402004020040
8002420039150040258001010800001080000506400002002020039200399996310019800102080000201600002003920039118002110910108000010000502042162220036080000102004020040200402004020040
80024200391500135258001010800001080000506400002002020039200399996310019800102080000201600002003920039118002110910108000010000502042162220036080000102004020040200402004020040
80024200391500402580010108000010800005064000020020200392003999963100198001020800002016000020039200391180021109101080000100005020122162220036080000102004020040200402004020040
8002420039150040258001010800001080000506400002002020100200399996310019800102080000201600002003920039118002110910108000010000502042162220036080000102004020040200402004020040
8002420039150040258001010800001080000506400002002020039200399996310019800102080000201600002003920039118002110910108000010000502042162220036080000102004020040200402004020040
8002420039150040258001010800001080000506400002002020039200399996310019800102080000201600002003920039118002110910108000010000502042162220036080000102004020040200402004020040