RDRAND (R64)

Operands

• Operand 1 (w): Register (RAX, RCX, RDX, RBX, RSP, RBP, RSI, RDI, R8, R9, R10, R11, R12, R13, R14, R15)
• Operand 2 (w, suppressed): Flags (AF: w, CF: w, OF: w, PF: w, SF: w, ZF: w)

Available performance data

• Arrow Lake-P
• Arrow Lake-E
• Meteor Lake-P
• Meteor Lake-E
• Emerald Rapids
• Alder Lake-P
• Alder Lake-E
• Rocket Lake
• Tiger Lake
• Ice Lake
• Cascade Lake
• Cannon Lake
• Skylake-X
• Coffee Lake
• Kaby Lake
• Skylake
• Broadwell
• Haswell
• Ivy Bridge
• Tremont
• Goldmont Plus
• Goldmont
• Airmont
• AMD Zen 5
• AMD Zen 4
• AMD Zen 3
• AMD Zen 2
• AMD Zen+

Arrow Lake-P

• Measurements
• Throughput
  • Computed from the port usage: 2.67
  • Measured (loop): 1573.80
  • Measured (unrolled): 1573.58
• Number of μops
  • Executed: 9
  • Retire slots: 14
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 15
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 13*ALU+1*JMP+1*LD+1*SHIFT+1*SLOW

Arrow Lake-E

• Measurements
• Throughput
  • Measured (loop): 1565.92
  • Measured (unrolled): 1459.58
• Number of μops
  • Executed: 13
  • Microcode Sequencer (MS): 13
  • Requires the complex decoder

Meteor Lake-P

• Measurements
• Throughput
  • Computed from the port usage: 7.00
  • Measured (loop): 1842.35
  • Measured (unrolled): 1844.38
• Number of μops
  • Executed: 21
  • Retire slots: 24
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 20
  • Requires the complex decoder (1 other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 2*p0+3*p0156B+2*p056+5*p06+7*p1+3*p15+1*p23A+2*p5

Meteor Lake-E

• Measurements
• Throughput
  • Measured (loop): 1457.13
  • Measured (unrolled): 1443.82
• Number of μops
  • Executed: 12
  • Microcode Sequencer (MS): 11
  • Requires the complex decoder

Emerald Rapids

• Measurements
• Throughput
  • Computed from the port usage: 7.00
  • Measured (loop): 118.65
  • Measured (unrolled): 118.90
• Number of μops
  • Executed: 21
  • Retire slots: 24
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 24
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 2*p0+3*p0156B+3*p056+4*p06+7*p1+3*p15+1*p23A+2*p5

Alder Lake-P

• Measurements
• Throughput
  • Computed from the port usage: 7.00
  • Measured (loop): 1379.67
  • Measured (unrolled): 1373.38
• Number of μops
  • Executed: 21
  • Retire slots: 24
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 24
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 2*p0+4*p0156B+2*p056+5*p06+7*p1+2*p15+1*p23A+2*p5

Alder Lake-E

• Measurements
• Throughput
  • Measured (loop): 1400.78
  • Measured (unrolled): 1224.35
• Number of μops
  • Executed: 19
  • Microcode Sequencer (MS): 18
  • Requires the complex decoder

Rocket Lake

• Measurements
• Throughput
  • Computed from the port usage: 7.75
  • Measured (loop): 1425.23
  • Measured (unrolled): 1425.33
• Number of μops
  • Executed: 32
  • Retire slots: 34
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 34
  • Requires the complex decoder (4 other instructions can be decoded with simple decoders in the same cycle)
• Port usage: 2*p0+4*p015+4*p0156+11*p06+7*p1+3*p5+1*p78

Tiger Lake

• Measurements
• Throughput
  • Computed from the port usage: 14.25
  • Measured (loop): 13541.75
  • Measured (unrolled): 13829.85
• Number of μops
  • Executed: 80
  • Retire slots: 64
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 91
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 4*p0+5*p015+2*p0156+24*p06+13*p1+1*p23+9*p49+9*p5+12*p78

Ice Lake

• Measurements
• Throughput
  • Computed from the port usage: 18.00
  • Measured (loop): 1548.77
  • Measured (unrolled): 1548.77
• Number of μops
  • Executed: 56
  • Retire slots: 54
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 62
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 9*p0+1*p01+17*p06+18*p1+10*p5+1*p78

Cascade Lake

• Measurements
• Throughput
  • Computed from the port usage: 4.00
  • Measured (loop): 606.03
  • Measured (unrolled): 607.43
• Number of μops
  • Executed: 16
  • Retire slots: 16
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 16
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 3*p015+1*p0156+8*p06+2*p1+1*p23

Cannon Lake

• Measurements
• Throughput
  • Computed from the port usage: 2.75
  • Measured (loop): 905.75
  • Measured (unrolled): 905.75
• Number of μops
  • Executed: 12
  • Retire slots: 14
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 14
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 1*p015+3*p0156+5*p06+2*p1+1*p23

Skylake-X

• Measurements
• Throughput
  • Computed from the port usage: 4.50
  • Measured (loop): 574.57
  • Measured (unrolled): 575.10
• Number of μops
  • Executed: 16
  • Retire slots: 16
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 16
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 3*p015+1*p0156+9*p06+2*p1+1*p23

Coffee Lake

• Measurements
• Throughput
  • Computed from the port usage: 1869.50
  • Measured (loop): 5718.93
  • Measured (unrolled): 5719.53
• Number of μops
  • Executed: 8459
  • Retire slots: 8416
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 8534
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 1808*p0+2*p015+2*p0156+1931*p06+1783*p1+1162*p23+12*p237+17*p4+1742*p5

Kaby Lake

• Measurements
• Throughput
  • Computed from the port usage: 3201.50
  • Measured (loop): 7241.02
  • Measured (unrolled): 7242.97
• Number of μops
  • Executed: 14407
  • Retire slots: 14350
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 14526
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 3102*p0+11*p01+6*p05+3301*p06+3003*p1+1997*p23+17*p4+2987*p5

Skylake

• Measurements
• Throughput
  • Computed from the port usage: 2460.00
  • Measured (loop): 6160.33
  • Measured (unrolled): 6160.10
• Number of μops
  • Executed: 11108
  • Retire slots: 11045
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 11216
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 2381*p0+8*p01+4*p05+2539*p06+2329*p1+1*p15+1533*p23+10*p237+18*p4+2293*p5

Broadwell

• Measurements
• Throughput
  • Computed from the port usage: 20.00
  • Measured (loop): 9223372036854775808.00
  • Measured (unrolled): 2088.27
• Number of μops
  • Executed: 75
  • Retire slots: 58
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 103
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 3*p0+35*p06+20*p1+7*p4+15*p5
• IACA 2.3
• Throughput
  • Computed from the port usage: 1.00
  • IACA: 8.57
• Number of μops: 5
• Port usage: 3*p0156+1*p06+1*p23
• IACA 3.0
• Throughput
  • Computed from the port usage: 1.00
  • IACA: 0.97
• Number of μops: 5
• Port usage: 3*p0156+1*p06+1*p23

Haswell

• Measurements
• Throughput
  • Computed from the port usage: 20.00
  • Measured (loop): 9223372036854775808.00
  • Measured (unrolled): 2381.62
• Number of μops
  • Executed: 73
  • Retire slots: 58
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 98
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 4*p0+36*p06+18*p1+7*p4+15*p5

Ivy Bridge

• Measurements
• Throughput
  • Computed from the port usage: 4.00
  • Measured (loop): 9223372036854775808.00
  • Measured (unrolled): 110.33
• Number of μops
  • Executed: 13
  • Retire slots: 13
  • Decoded (MITE): 0
  • Microcode Sequencer (MS): 13
  • Requires the complex decoder (no other instruction can be decoded with simple decoders in the same cycle)
• Port usage: 7*p015+1*p05+2*p1+1*p23+2*p5

Tremont

• Measurements
• Throughput
  • Measured (loop): 9223372036854775808.00
  • Measured (unrolled): 3539.98
• Number of μops
  • Executed: 58
  • Microcode Sequencer (MS): 58
  • Requires the complex decoder

Goldmont Plus

• Measurements
• Throughput
  • Measured (loop): 9223372036854775808.00
  • Measured (unrolled): 1940.43
• Number of μops
  • Executed: 16
  • Microcode Sequencer (MS): 16
  • Requires the complex decoder

Goldmont

• Measurements
• Throughput
  • Measured (loop): 9223372036854775808.00
  • Measured (unrolled): 3230.75
• Number of μops
  • Executed: 17
  • Microcode Sequencer (MS): 17
  • Requires the complex decoder

Airmont

• Measurements
• Throughput
  • Measured (loop): 9223372036854775808.00
  • Measured (unrolled): 1345.37
• Number of μops
  • Executed: 15
  • Microcode Sequencer (MS): 15
  • Requires the complex decoder

AMD Zen 5

• Measurements
• Throughput
  • Computed from the port usage: 8.50
  • Measured (loop): 67.40
  • Measured (unrolled): 66.97
• Number of μops
  • Executed: 56
• Port usage: 16*FP01+8*FP03+8*FP12+2*FP23+2*FP45
• Documentation
• Latency: NA
• Throughput: NA
• Number of μops: ucode
• Port usage: ucode

AMD Zen 4

• Measurements
• Throughput
  • Computed from the port usage: 7.33
  • Measured (loop): 65.33
  • Measured (unrolled): 65.00
• Number of μops
  • Executed: 58
• Port usage: 14*FP01+8*FP12+1*FP23+2*FP45
• Documentation
• Latency: variable
• Throughput: variable
• Number of μops: ucode

AMD Zen 3

• Measurements
• Throughput
  • Computed from the port usage: 7.00
  • Measured (loop): 72.25
  • Measured (unrolled): 72.00
• Number of μops
  • Executed: 55
• Port usage: 7*FP01+7*FP03+7*FP1+1*FP12+1*FP23+2*FP45
• Documentation
• Latency: variable
• Throughput: variable
• Number of μops: ucode

AMD Zen 2

• Measurements
• Throughput
  • Measured (loop): 3610.35
  • Measured (unrolled): 3610.62
• Number of μops
  • Executed: 17
• Documentation
• Latency: variable
• Throughput: variable
• Number of μops: ucode

AMD Zen+

• Measurements
• Throughput
  • Measured (loop): 2536.53
  • Measured (unrolled): 2535.90
• Number of μops
  • Executed: 19
• Documentation
• Latency: variable
• Throughput: variable
• Number of μops: ucode