What this comparison shows

One processor or two

One machine, two configurations, so almost everything about them is held constant. The clock is identical, which the 1× multiplier states plainly, and the memory capacity doubles.

The SPEC rows are the reason this preset exists. Four published results appear here, and not one of them pairs with another. Each result is a separate comparability group. Its suite, version and base or peak variant are part of the group’s identity. The two configurations were submitted under different variants, so no column ever meets another inside one group. The figures are shown apart, with their benchmark identity next to each, and no multiplier is produced.

This is the correct outcome even though it is a frustrating one. A SPEC CFP result and a SPEC CINT result measure different work; a base result and a peak result allow different compiler settings. Dividing across those lines is exactly what SPEC’s own fair-use rules exist to prevent, and it is what most “N× faster” claims on the internet are made of.

At a glance

This card lists what each record states. These are the six questions every machine in the catalog is asked. A row with answers from different methods has no bar or multiplier; the line beneath it explains the difference.

Question	Silicon Graphics OctaneSilicon Graphics, Inc.Workstation · 1997Octane 1X 250MHz R10k	Silicon Graphics OctaneSilicon Graphics, Inc.Workstation · 1997Octane 2X 250MHz R10k
CPU clock	Silicon Graphics Octane (Octane 1X 250MHz R10k)250 MHzConfirmed	Silicon Graphics Octane (Octane 2X 250MHz R10k)250 MHzConfirmed
Methods, caveats and sources for cpu clock Where these sit among 65 machines recorded this way 200 MHzSega Dreamcast 250 MHzSilicon Graphics Octane 294.912 MHzSony PlayStation 2 Silicon Graphics Octane — 36th-lowest of the 65 figures recorded this way, and 5th-lowest of the 7 from 1994–2000. Positions on a linear scale. See all 65 over time.
GPU clock	Silicon Graphics Octane (Octane 1X 250MHz R10k)not applicableno gpu component is fitted to this record	Silicon Graphics Octane (Octane 2X 250MHz R10k)not applicableno gpu component is fitted to this record
Memory capacity	Silicon Graphics Octane (Octane 1X 250MHz R10k)128 MiBConfirmed	Silicon Graphics Octane (Octane 2X 250MHz R10k)256 MiBConfirmed
Methods, caveats and sources for memory capacity Where these sit among 68 machines recorded this way 67 108 864 BMicrosoft Xbox 128 MiBSilicon Graphics Octane 128 MiBATI Radeon 9700 Pro Silicon Graphics Octane — 33rd-lowest of the 68 figures recorded this way, and the highest of the 9 from 1994–2000. Positions on a linear scale. See all 68 over time.
Memory bandwidth	Silicon Graphics Octane (Octane 1X 250MHz R10k)unknownThe cited documents are published SPEC results, which describe the machine’s configuration and its scores and state no memory bandwidth.	Silicon Graphics Octane (Octane 2X 250MHz R10k)unknownThe cited documents are published SPEC results, which describe the machine’s configuration and its scores and state no memory bandwidth.
Peak FP32	Silicon Graphics Octane (Octane 1X 250MHz R10k)500 MFLOP/sMIPS R10000Confirmed	Silicon Graphics Octane (Octane 2X 250MHz R10k)500 MFLOP/sMIPS R10000Confirmed
Methods, caveats and sources for peak fp32 Where these sit among 4 machines recorded this way 8 MFLOP/sCommodore Amiga 4000 500 MFLOP/sSilicon Graphics Octane 6.2 GFLOP/sSony PlayStation 2 Silicon Graphics Octane — 2nd-lowest of the 4 figures recorded this way. Positions on a linear scale. See all 4 over time.
Power	Silicon Graphics Octane (Octane 1X 250MHz R10k)unknownA SPEC result reports the hardware and software configuration of the machine under test, not its power rating. SGI’s own hardware documentation for the Octane is not held by any archive this project may cite: what survives of SGI on the Internet Archive is brochures and price lists, and the Octane brochure of December 1999 states the supply voltage and no wattage at all.	Silicon Graphics Octane (Octane 2X 250MHz R10k)unknownA SPEC result reports the hardware and software configuration of the machine under test, not its power rating. SGI’s own hardware documentation for the Octane is not held by any archive this project may cite: what survives of SGI on the Internet Archive is brochures and price lists, and the Octane brochure of December 1999 states the supply voltage and no wattage at all.

Silicon Graphics Octane (Octane 1X 250MHz R10k), the first record, is the multiplier baseline. Each section covers its own metrics. This comparison has no overall score. Bars are drawn to scale within a row and carry no number of their own.

Processing

Instruction rates, floating-point peaks, benchmark results and clocks. None of these converts into another.

Clock frequencycpu

Method	Silicon Graphics OctaneOctane 1X 250MHz R10k	Silicon Graphics OctaneOctane 2X 250MHz R10k
Nominal clock	Silicon Graphics Octane (Octane 1X 250MHz R10k)250 MHzConfirmed baseline Recorded as Vendor-rated. Stated by the vendor. Sources spec-cpu95-octane-int	Silicon Graphics Octane (Octane 2X 250MHz R10k)250 MHzConfirmed 1×the baselinecaveat Recorded as Vendor-rated. Stated by the vendor. Caveat Two processors at 250 MHz is not 500 MHz of anything. Clock frequency does not add across processors, which is exactly why the floating-point result obtained on this machine is recorded separately from the single-processor integer one. Multiplier `a ÷ b` (ratio v1, to 3 significant digits) Silicon Graphics Octane: 250 MHz Silicon Graphics Octane: 250 MHz Compares clock-frequency at cpu scope, obtained by nominal-clock. Valid only within this comparability group, and reported to 3 significant digits — the precision of the least precise input. Sources spec-cpu95-octane-fp

Peak FP32 ratecpu

Method	Silicon Graphics OctaneOctane 1X 250MHz R10k	Silicon Graphics OctaneOctane 2X 250MHz R10k
Theoretical peak	Silicon Graphics Octane (Octane 1X 250MHz R10k)500 MFLOP/sConfirmed baselinecaveat Recorded as MIPS R10000: Derived. Computed here from figures the vendor stated. Caveat Computed here, not published anywhere. Nobody rated this processor in floating-point operations per second: a processor of 1996 was documented by what its units retire per cycle, and the manual says the adder and the multiplier each accept one operation every cycle and stops there. The multiplication is the derived claim this record names, with both constants cited to that manual. It is an upper bound twice over — every unit fed on every cycle, nothing stalling — and the same figure holds at double precision, because the manual gives the multiplier one repeat rate for both. Sources mips-r10000-users-manual spec-cpu95-octane-int	Silicon Graphics Octane (Octane 2X 250MHz R10k)500 MFLOP/sConfirmed 1×the baselinecaveat Recorded as MIPS R10000: Derived. Computed here from figures the vendor stated. Caveat Computed here, not published anywhere. Nobody rated this processor in floating-point operations per second: a processor of 1996 was documented by what its units retire per cycle, and the manual says the adder and the multiplier each accept one operation every cycle and stops there. The multiplication is the derived claim this record names, with both constants cited to that manual. It is an upper bound twice over — every unit fed on every cycle, nothing stalling — and the same figure holds at double precision, because the manual gives the multiplier one repeat rate for both. Multiplier `a ÷ b` (ratio v1, to 3 significant digits) Silicon Graphics Octane: 500 MFLOP/s Silicon Graphics Octane: 500 MFLOP/s Compares peak-fp32-rate at cpu scope, obtained by theoretical-peak. Valid only within this comparability group, and reported to 3 significant digits — the precision of the least precise input. Sources mips-r10000-users-manual spec-cpu95-octane-int

Sustained FP64 ratewhole-system

Both records address this quantity, but neither states a value. The figures are recorded as an explicit absence, so there is nothing to compare.

Method	Silicon Graphics OctaneOctane 1X 250MHz R10k	Silicon Graphics OctaneOctane 2X 250MHz R10k
Benchmark run	Silicon Graphics Octane (Octane 1X 250MHz R10k)unknownSGIStuff reports a LINPACK result for an Octane with an R10000, but names neither the LINPACK revision nor the floating-point precision. The row cannot establish an FP64 benchmark result. baseline Recorded as Published independently. Sources sgistuff-octane-linpack	Silicon Graphics Octane (Octane 2X 250MHz R10k)unknownSGIStuff reports a LINPACK result for an Octane with an R10000, but names neither the LINPACK revision nor the floating-point precision. The row cannot establish an FP64 benchmark result. baseline states none Recorded as Published independently. Multiplier No multiplier: Silicon Graphics Octane (Octane 1X 250MHz R10k) records no value here, so there is nothing to divide by. Sources sgistuff-octane-linpack

SPEC CPU resultwhole-system

These figures belong to different comparability groups (published spec result under spec-cpu 95 cint-base, published spec result under spec-cpu 95 cint-peak, published spec result under spec-cpu 95 cfp-base and published spec result under spec-cpu 95 cfp-peak). They are shown separately, and no multiplier is produced.

Method	Silicon Graphics OctaneOctane 1X 250MHz R10k	Silicon Graphics OctaneOctane 2X 250MHz R10k
Published SPEC resultspec-cpu 95, cint-base	Silicon Graphics Octane (Octane 1X 250MHz R10k)13.0Confirmed baselinecaveat Recorded as Measured. Stated by the vendor. Caveat SPEC fair use applies to any quotation of this figure. Base and peak are different comparability groups: base results are run with one flag set across all benchmarks, peak results allow per-benchmark tuning, and the two are never interchangeable. Comparable only with other CINT95 base results, never with CPU2000 or later suites. SPEC publishes no conversion between suite generations. Sources spec-cpu95-octane-int	Silicon Graphics Octane (Octane 2X 250MHz R10k)not recorded
Published SPEC resultspec-cpu 95, cint-peak	Silicon Graphics Octane (Octane 1X 250MHz R10k)13.6Confirmed baselinecaveat Recorded as Measured. Stated by the vendor. Caveat SPEC fair use applies. A peak result is tuned per benchmark and is not comparable with a base result, including the base result for this same machine. Sources spec-cpu95-octane-int	Silicon Graphics Octane (Octane 2X 250MHz R10k)not recorded
Published SPEC resultspec-cpu 95, cfp-base	Silicon Graphics Octane (Octane 1X 250MHz R10k)not recorded	Silicon Graphics Octane (Octane 2X 250MHz R10k)25.5Confirmed baseline states nonecaveat Recorded as Measured. Stated by the vendor. Caveat SPEC fair use applies. A floating-point result is not comparable with an integer one: different benchmarks, different code, different comparability group. This result was also obtained on a two-processor machine with an auto-parallelizing compiler, so it is not a single-processor figure. Multiplier No multiplier: Silicon Graphics Octane (Octane 1X 250MHz R10k) records no value here, so there is nothing to divide by. Sources spec-cpu95-octane-fp
Published SPEC resultspec-cpu 95, cfp-peak	Silicon Graphics Octane (Octane 1X 250MHz R10k)not recorded	Silicon Graphics Octane (Octane 2X 250MHz R10k)26.6Confirmed baseline states nonecaveat Recorded as Measured. Stated by the vendor. Caveat SPEC fair use applies. Peak and base are separate comparability groups, and this is a two-processor result. Multiplier No multiplier: Silicon Graphics Octane (Octane 1X 250MHz R10k) records no value here, so there is nothing to divide by. Sources spec-cpu95-octane-fp

Memory

Capacity, bandwidth, bus width and timing. Binary and decimal prefixes are kept as the source wrote them.

Memory capacitywhole-system

Method	Silicon Graphics OctaneOctane 1X 250MHz R10k	Silicon Graphics OctaneOctane 2X 250MHz R10k
Design capacity	Silicon Graphics Octane (Octane 1X 250MHz R10k)128 MiBConfirmed baselinecaveat Recorded as Vendor-rated. Stated by the vendor. Caveat The result file writes "128MB", which in this period and context means 128 MiB. It describes the tested configuration, not the maximum the Octane supported. Sources spec-cpu95-octane-int	Silicon Graphics Octane (Octane 2X 250MHz R10k)256 MiBConfirmed 2×the baseline Recorded as Vendor-rated. Stated by the vendor. Multiplier `a ÷ b` (ratio v1, to 3 significant digits) Silicon Graphics Octane: 256 MiB Silicon Graphics Octane: 128 MiB Compares memory-capacity at whole-system scope, obtained by design-capacity. Valid only within this comparability group, and reported to 3 significant digits — the precision of the least precise input. Sources spec-cpu95-octane-fp

Memory bandwidthmemory

Both records address this quantity, but neither states a value. The figures are recorded as an explicit absence, so there is nothing to compare.

Method	Silicon Graphics OctaneOctane 1X 250MHz R10k	Silicon Graphics OctaneOctane 2X 250MHz R10k
Theoretical peak	Silicon Graphics Octane (Octane 1X 250MHz R10k)unknownThe cited documents are published SPEC results, which describe the machine’s configuration and its scores and state no memory bandwidth. baselinecaveat Recorded as Published independently. Caveat Unknown because of what this record is sourced from. Silicon Graphics quoted a bandwidth for the Octane’s crossbar in its own literature, and none of that literature is cited here. The machine is in the catalog on the strength of two SPEC submissions, which say what was fitted and how fast it ran the suite. Sources spec-cpu95-octane-int	Silicon Graphics Octane (Octane 2X 250MHz R10k)unknownThe cited documents are published SPEC results, which describe the machine’s configuration and its scores and state no memory bandwidth. baseline states nonecaveat Recorded as Published independently. Caveat Unknown because of what this record is sourced from. Silicon Graphics quoted a bandwidth for the Octane’s crossbar in its own literature, and none of that literature is cited here. The machine is in the catalog on the strength of two SPEC submissions, which say what was fitted and how fast it ran the suite. Multiplier No multiplier: Silicon Graphics Octane (Octane 1X 250MHz R10k) records no value here, so there is nothing to divide by. Sources spec-cpu95-octane-int

Energy

A cooling-design target, a specified rating and a measured wall draw are separate quantities, and only the last supports a multiplier.

Rated power consumptionwhole-system

Both records address this quantity, but neither states a value. The figures are recorded as an explicit absence, so there is nothing to compare.

Method	Silicon Graphics OctaneOctane 1X 250MHz R10k	Silicon Graphics OctaneOctane 2X 250MHz R10k
Rated power	Silicon Graphics Octane (Octane 1X 250MHz R10k)unknownA SPEC result reports the hardware and software configuration of the machine under test, not its power rating. SGI’s own hardware documentation for the Octane is not held by any archive this project may cite: what survives of SGI on the Internet Archive is brochures and price lists, and the Octane brochure of December 1999 states the supply voltage and no wattage at all. baselinecaveat Recorded as Published independently. Caveat The same limitation as the bandwidth, from the same two documents. A workstation of this size has a supply rating printed on it; no source here states it. Sources spec-cpu95-octane-int	Silicon Graphics Octane (Octane 2X 250MHz R10k)unknownA SPEC result reports the hardware and software configuration of the machine under test, not its power rating. SGI’s own hardware documentation for the Octane is not held by any archive this project may cite: what survives of SGI on the Internet Archive is brochures and price lists, and the Octane brochure of December 1999 states the supply voltage and no wattage at all. baseline states nonecaveat Recorded as Published independently. Caveat The same limitation as the bandwidth, from the same two documents. A workstation of this size has a supply rating printed on it; no source here states it. Multiplier No multiplier: Silicon Graphics Octane (Octane 1X 250MHz R10k) records no value here, so there is nothing to divide by. Sources spec-cpu95-octane-int

Change it

This preset is a starting point, not a fixed set. Open it in the comparison builder to swap a record in or out. The builder needs JavaScript; this page does not.