Apple M1 Microarchitecture Research by Dougall Johnson

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

UQSHRN2 (8H)

Test 1: uops

Code:

  uqshrn2 v0.16b, v1.8h, #3
  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)03191e1f3f4e51inst 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
10043037230001072548251000100010003983131301830373037241532895100010002000303730371110011000473116112630100030383038303830383038
10043037220002592548251000100010003983130301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037230002562548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037220002222548251000100010003983130301830373037241532895100010002000303730372110011000173116112630100030383038303830383038
10043037220001952539251000100010003983130301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037230001702548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037220002542548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
1004303723000612548251000100010003983131301830373037241532895100010002000303730371110011000173116112630100030383038303830383038
10043037220003092548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038
10043037220004412548251000100010003983131301830373037241532895100010002000303730371110011000073116112630100030383038303830383038

Test 2: Latency 1->1

Code:

  uqshrn2 v0.16b, v1.8h, #3
  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)696d6edispatch stall (70)72scheduler 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)9fa8branch mispredict (cb)cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10204300372250000300061295482510100100100001001000050042773133001830037300372826503287451010020010000200200003003730037111020110099100100100001000071021611296340100001003003830038300383003830038
1020430037224000024264061295482510100100100001001000050042788153001830037300372826503287451010020010000200200003003730037111020110099100100100001000071011611296340100001003003830038300383003830038
102043003722500003150061295482510100100100001001000050042773133001830037300372826503287451010020010000200200003003730037111020110099100100100001000071011611296340100001003003830038300383003830038
102043003722500004650061295482510100100100001001000050042773133001830037300372826503287451010020010000200200003003730037111020110099100100100001000071011611296340100001003003830038300383003830038
102043003722500003720061295482510100100100001001000050042773133001830037300372826503287451010020010000200200003003730037111020110099100100100001000071011611296340100001003003830038300383003830038
102043003722500003630061295482510100100100001001000050042773133001830037300372826503287451010020010000200200003003730037111020110099100100100001000071011611296340100001003003830038300383003830038
1020430037225000030264061295482510100100100001001000050042773133001830037300372826503287451010020010000200200003003730037111020110099100100100001000071011611296340100001003003830038300383003830038
102043003722500004290061295482510100100100001001000050042773133001830037300372826503287451010020010000200200003008530037111020110099100100100001000071011611296340100001003003830038300383003830038
10204300372250000270061295482510100100100001001000050042773133001830037300372826503287451010020010000200200003003730037111020110099100100100001000071011611296340100001003003830038300383003830038
10204300372250000210061295482510100100100001001000050042773133001830037300372826503287451010020010000200200003003730037111020110099100100100001000071011611296340100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)0307080a0b18191e1f3a3f4e51inst 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)9fa0a1a6a7a8acc2cfd5d6ddinst fetch restart (de)e0? int output thing (e9)? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
100243003722400000000061295482510010101000010100006042773131300183003730037282873287671001020100002020000300373003711100211091010100001000001007052162229630010000103003830038300383003830038
100243003722400000000061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
100243003722400000000061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
1002430037225000000000726295482510010121000012100005042773130300183003730037282873287671016320100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
1002430131224000000000536295482510010101000010100005042773130300183003730037282873287671001220100002020000300373003711100211091010100001000000006402166229632210000103003830038300383003830038
100243003722500000000061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
100243003722500000000061295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
100243003722500000000061295482510010101002410100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
100243003722500000000061295482510012101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001000000006402162229630010000103003830038300383003830038
1002430037225000000000103295482510010101000010100005042773130300183003730037282873287671001020100002020000300373003711100211091010100001002000006402162229630210000103003830038300383003830038

Test 3: Latency 1->2

Code:

  uqshrn2 v0.16b, v0.8h, #3
  movi v0.16b, 1

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)03191e1f3f4e51inst 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)cdcfd6e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1020430037225030029329547251010010010000100100005004277160130018030037300372825262874010100200100082002001630037300371110201100991001001000010001001117181629646100001003003830038300383003830038
1020430037224018072629547251010010010000100100005004277160030018030037300372827162874110100200100082002001630037300371110201100991001001000010001001117171629646100001003003830038300383003830038
102043003722503006129547251010010010000100100005004278507130018030037300372827162874010100200100082002001630037300371110201100991001001000010000031117181629645100001003003830038300383003830038
102043003722501806129547251010010010000100100005004277160130018030037300372827172874010100200100082002001630084300852110201100991001001000010000001117181629645100001003018130038300383003830038
102043003722506906129547251010010010000100100005004277160130018030037300372827162874110100200100082002001630037300371110201100991001001000010000001117181629646100001003003830038300383003830038
102043003722402706129547251010010010000100100005004277160130018030037300372827172874010100200100082002001630037300371110201100991001001000010000001117181629645100001003003830038300383003830038
10204300372250906129547251010010010000100100005004277160130018030037300372827172874110100200100082002001630037300371110201100991001001000010000001117181629646100001003003830038300383003830038
10204300372250006129547251010010010000100100005004277160130018030037300372827162874010100200100082002001630037300371110201100991001001000010000001117181629645100001003003830038300383003830038
102043003722501206129547251010010010000100100005004277160130018030037300372827162874110100200100082002001630037300371110201100991001001000010000001117181629645100001003003830038300383003830038
102043003722501506129547251010010010000100100005004277160130018030037300372827172874010100200100082002001630037300371110201100991001001000010000001117181629646100001003003830038300383003830038

1000 unrolls and 10 iterations

Result (median cycles for code): 3.0037

retire (01)cycle (02)031e1f3f4e51inst 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)9fa8cfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024300372259006129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640416342962910000103003830038300383003830038
10024300372251506129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640316442962910000103003830038300383003830038
1002430037225906129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640416442962910000103003830038300383003830038
1002430037225006129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640416442962910000103003830038300383003830038
1002430037225006129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640416432962910000103003830038300383003830038
1002430037225006129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640316342962910000103003830038300383003830038
10024300372251506129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640416432962910000103003830038300383003830038
10024300372255406129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640416432962910000103003830038300383003830038
1002430037224006129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640416432962910000103003830038300383003830038
1002430037225006129547251001010100001010000504277160030018300373003728286328767100102010000202000030037300371110021109101010000100640416432962910000103003830038300383003830038

Test 4: throughput

Count: 8

Code:

  movi v0.16b, 0
  uqshrn2 v0.16b, v8.8h, #3
  movi v1.16b, 0
  uqshrn2 v1.16b, v8.8h, #3
  movi v2.16b, 0
  uqshrn2 v2.16b, v8.8h, #3
  movi v3.16b, 0
  uqshrn2 v3.16b, v8.8h, #3
  movi v4.16b, 0
  uqshrn2 v4.16b, v8.8h, #3
  movi v5.16b, 0
  uqshrn2 v5.16b, v8.8h, #3
  movi v6.16b, 0
  uqshrn2 v6.16b, v8.8h, #3
  movi v7.16b, 0
  uqshrn2 v7.16b, v8.8h, #3
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2508

retire (01)cycle (02)0318191e3f51inst 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)60696d6escheduler 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)9fa0a8c5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0ea? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1602042008815100129292580116100800161008002850064019612004520065200656128012820080028200160268200652006521160201100991001001600001000011110119016002006211600001002006620066200662006620066
16020420065150000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000011110119016002006201600001002006620066200662006620066
16020420065151000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000011110119016002006201600001002006620066200662006620066
16020420065150000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000011110119016002006201600001002006620066200662006620066
16020420065150000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000011110119016002006201600001002006620066200662006620066
1602042006515100237292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001002011110119016002006201600001002006620066200662006620066
16020420065150000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000011110119016002006201600001002006620066200662006620066
160204200651500018292580116106800161008002850064019612004520065200656128012820080028200160056201312006511160201100991001001600001000011110119016002006201600001002006620066200662006620066
160204200651500015292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000011110119016002006201600001002006620066200662006620066
16020420065151000292580116100800161008002850064019612004520065200656128012820080028200160056200652006511160201100991001001600001000011110119016002006201600001002006620066200662006620066

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2507

retire (01)cycle (02)03040818191e1f3a3f51inst 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)5f6061696d6escheduler 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)9fa7a8accfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0eaec? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1600242007715000004201572980010108000010800005064000000020041200712006032280010208000020160000200602006011160021109101016000010000100356222034322141820050201160000102005220054200542005220052
160024200511500000600452980010108000010800005064000001520043200602006032280010208000020160000200512005111160021109101016000010000100413512027111191820050201160000102005420052200522005220052
1600242005115000004500452780010108000010800005064000011520032200512005132280010208000020160000200532005311160021109101016000010000100408511825211171220048211160000102005220052200522005220054
160024200531510000900452780010108000010800005064000011520034200622005332280010208000020160000200512005111160021109101016000010000100388511325111161720048201160000102005220052200522005220054
1600242005315000002400452780010108000010800005064000010520032200602005132280010208000020160000200512005111160021109101016000010000100393111725111181920050201160000102005220052200542005420052
160024200531500000000452780010108000010800005064000010020034200512005132280010208000020160000200512005311160021109101016000010000100383111527211131620048201160000102005220052200522005220054
16002420051150000072004527800101080000108000050640000110200322005120051322800102080000201600002005120051111600211091010160000100069100368511025111162120048201160000102005220052200522005420052
1600242005115000001500682780010108000010800005064000010520032200622005132280010208000020160000200532005311160021109101016000010000100393111727211151720048211160000102005220054200542005220052
1600242005115000005100452780010118000010800005064000010520032200512005132280010208000020160000200532005311160021109101016000010000100413111225111171620050201160000102005220052200522005220052
160024200511500000000452780010108000010800005064000010520134200532005132280010208000020160000200512005111160021109101016000010000100348511727111151720048201160000102005220052200522005220052

Test 5: throughput

Count: 16

Code:

  uqshrn2 v0.16b, v16.8h, #3
  uqshrn2 v1.16b, v16.8h, #3
  uqshrn2 v2.16b, v16.8h, #3
  uqshrn2 v3.16b, v16.8h, #3
  uqshrn2 v4.16b, v16.8h, #3
  uqshrn2 v5.16b, v16.8h, #3
  uqshrn2 v6.16b, v16.8h, #3
  uqshrn2 v7.16b, v16.8h, #3
  uqshrn2 v8.16b, v16.8h, #3
  uqshrn2 v9.16b, v16.8h, #3
  uqshrn2 v10.16b, v16.8h, #3
  uqshrn2 v11.16b, v16.8h, #3
  uqshrn2 v12.16b, v16.8h, #3
  uqshrn2 v13.16b, v16.8h, #3
  uqshrn2 v14.16b, v16.8h, #3
  uqshrn2 v15.16b, v16.8h, #3
  movi v16.16b, 17

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2502

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)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)9fa0a1a6a8a9acc5branch mispredict (cb)cdcfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
16020440059300030251601081001600081001600205001280132040020400394003919977619990160120200160032200320064400394003911160201100991001001600001000000001111011801600400361600001004004040040400404004040040
16020440039299030251601081001600081001600205001280132040020400394003919977619990160120200160032200320064400394003911160201100991001001600001000003001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132040020400394003919977619990160120200160032200320064400394003911160201100991001001600001000000001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132040020400394003919977619990160120200160032200320064400394003911160201100991001001600001000000031111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132040020400394003919977619990160120200160032200320064400394003911160201100991001001600001000000001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132040020400394003919977619990160120200160032200320064400394003911160201100991001001600001000000001111011801600400361600001004004040040400404004040040
16020440039300030251601081001600081001600205001280132040020400394010219977619990160120200160032202320064400394020311160201100991001001600001000001031111011803900400361600001004004040040400404004040040
160204400392994830251601081001600081001600205001280132040020400394003919977620024160120200160032200320064400394003911160201100991001001600001000000001111011801600400361600001004004040040400404004040040
16020440039299030251601081001600081001600205001280132040020400394003919977619990160120200160032200320064400394003911160201100991001001600001000000001111011801600400361600001004004040040400404004040040
16020440039299030251601081001600081001600205001280132040020400394003919977619990160120200160032200320064400394003911160201100991001001600001000000001111011801600400361600001004004040040400404004040040

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2502

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)5f6061696d6edispatch 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)9fa8accfd0d2icache miss (d3)d5d6d9dadbddinst fetch restart (de)e0eaeb? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
1600244003930004625160010101600001016000050128000011540020400394003919996320019160010201600002032000040039400391116002110910101600001000100228212816221241740036305160000104004040040400404004040040
160024400393000466016010810160000101600005012800001110400204003940039199963200191600102016000020320000400394003911160021109101016000010001002213712316211242440036155160000104004040040400404004040040
1600244003930002162516001010160000101600005012800001110400204003940039199963200191601222016000020320000400394003911160021109101016000010001002213712516211242540036155160000104004040040400404004040040
160024400393000462516001010160000101600005012800001110400204003940039199963200191600102016000020320000400394003911160021109101016000010001002213712616211232340036155160000104004040040400404004040040
160024400393000462516001010160000101600005012800001110400204003940039199963200191600102016000020320000400394003911160021109101016000010001002213712416211242540036155160000104004040040400404004040040
160024400393000462516001010160000101600005012800001110400204003940039199963200191600102016000020320000400394003911160021109101016000010001002213712516211232340036155160000104004040040400404004040040
160024400393000462516001010160000101600005012800001110400204003940039199963200191600102016000020320000400394003911160021109101016000010101002213712516211242540036155160000104004040040400404004040040
160024400392990462516001010160000101600005012800001110400204003940039199963200191600102016000020320000400394003911160021109101016000010001002213712416211242440036155160000104004040040400404004040040
160024400393000462516001010160000101600005012800001110400204003940039199963200191600102016000020320000400394003911160021109101016000010101002213712316211232340036155160000104004040040400404004040145
160024400393000522516001010160000101600005012800001110400204003940039199963200191600102016000020320000400394003911160021109101016000010001002213712216211232340036155160000104004040040400404004040040