This website is an archived copy of "Chris's Acorns", now hosted by The Centre for Computing History as part of The Chris Whytehead collection.
To view the page in it's original forms without this header, append "?cfch-archive=orig" to the URL.

RISC OS Performance Page

[Computers]  [Home] [Links] 

This page is to show the relative performance of RISC OS computers. Benchmarks may not mean a lot but this does at least show how performance has changed over the years. Running these benchmarks clearly shows that their performance depends in part on the screen size, the greater the scrren size and number of colours, the worse the performance measure. To try and ensure consistent results I have run the benckmark programs in mode 15 for the earlier Archimedes computers and 800x600 at 60MHz and 256 colours, any exceptions to this will be highlighted. To minimise the impact, if any, of other programs on !Alarm and !Edit will be on the icon bar. The results are also shown under the individual computer. My chosen measures are:

  1. Mathias Seiffert's !SICK 1.22 for Dhrystones/sec and kWhetstones/sec
  2. J.Abbott & Nick Smith ARMSi v4.0 (7 Aug-1996) for MIPS

Please note that these benchmarks were collected from the actual computers in my collection, with their unique combination of hardware, software and modifications. There is no guarantee that you will get the same results on your computer!

ComputerRISC OSScreenDhrystones/sec kWhetstones/secMIPS
A4402.00Mode 275383664.56
A310 (2)2.00Mode 154797n/a4.80
A30003.11Mode 154728444.82
A3020 (2)3.11Mode 285407357.09
A3020 (1)3.11Mode 285499367.18
A4000 (1)3.11Mode 285407367.19
A3010 (ARM250)3.11Mode 285500357.17
A3010 (Adelaide)3.11Mode 285508367.22
A4000S3.11Mode 285557367.61
A420/1 (2)3.10Mode 1517309214n/a
A310 (3)3.10Mode 151732621312.75
A5000 (1)3.19800x600x256 72Hz19557249913.75
A540 (2)2.01Mode 152174728414.14
A5000 (2)3.11Mode 152280925417
R2602.01Mode 152380531017.16
R1403.10Mode 152033726617.16
R260 (FPA)2.01Mode 15226212807n/a
A540 (1) (FPA)3.11Mode 15241523055n/a
A70003.60800x600x256 60Hz2425313717.87
RiscPC ACB25
ARM610 30MHz
3.50800x600x256 75Hz3774320317.95
RiscPC 600
Prototype
3.50800x600x256 75Hz3820320617.97
STB23.61TVn/an/a18.56
RiscPC ACB75B
ARM610 33MHz
3.60800x600x256 75Hz3826720419.39
RiscPC ACB75B
ARM710 40MHz
3.60800x600x256 75Hz4849930822.43
RiscPC ACB76
ARM710 40MHz
3.70800x600x256 75Hz4894630522.62
NetStationNCOS 1.06800x600x256 56Hz41146485623.67
Xemplar Matrix NCNCOS 1.11800x600x256 56Hz41001465423.68
A7000+3.71800x600x256 60Hz47548529728.41
Bush IBX200NCOS 5.13Mode 1561464699728.67
Bush IBX101NCOS 5.13Mode 1561460699128.75
A7000+ Odyssey3.71800x600x256 75Hz44200519432.63
Microdigital Mico4.03800x600x256 60Hz52581610132.92 
RiscStation NetWorxHD4.03800x600x256 60Hz63410737133.34
RiscStation R75004.39800x600x256 60Hz63566738133.34
StorngARM Risc PC 233MHz
(SRP26)
3.70800x600x256 75Hz4222772613156.2
Risc PC 700
233MHz StrongARM
4.02800x600x256 75Hz 4081422753159.2
Risc PC 600
287MHz StrongARM
6.10800x600x256 75Hz4990143356172.9
Risc PC 600
287MHz StronARM
6.14800x600x256 75Hz 4989373357184.6
RiscPC 600
287MHz StrongARM
4.39800x600x256 70Hz4928053457203.2

Notes
1. !SICK abends when run on an ARM7 CPU (ARM710 or ARM7500) which means I do not have Drystones/sec or kWhetstones/sec figures for these systems. This bug was fixed in V1.26 and I must retest the ARM710/7500 computers which failed.

[Computers]   [Home] [Links]