::: Area #31 (Obsolete......... [General]) Message: #1122 (Read 15 times, has 0 replies, 1020 bytes) Date : Tue Sep 30 20:44:57 1997 From : Derek Irwin of fidonet#2:254/235.20 To : John Metcalf Subject: 8 Bit benchmarks Hi John, 13-Sep-97 20:28:15, John Metcalf wrote to All Subject: 8 Bit benchmarks /* 10 LET W=500:DIM F(W):LET P=1:LET A=3 20 LET F(P)=A:LET P=P+1:IF P>W THEN STOP 30 LET A=A+2:LET X=1 40 LET S=A/F(X):IF S=INT(S) THEN 30 50 LET X=X+1:IF X

CUL, Derek... ______________________ | Derek Irwin ) |\ | 2:254/235.20 __)~| |O\ |__________________/=___|_|__| " 'O O O` " 'O`----'O`=' --- Terminate 5.00/Pro [*Registered* 24/11/95] * Origin: Point 20 on PackLink +44(0)1812972486 (2:254/235.20) Message: #1123 (Read 15 times, has 0 replies, 2404 bytes) Date : Thu Oct 2 22:16:00 1997 From : John Metcalf of fidonet#2:257/71 To : All Subject: 8 bit benchmarks Thank-you to all who have either provided me with some benchmark figures or shown an interest. Here are the results I obtained on my own machines: Acorn Electron 2.0MHz 6502 138 seconds Amstrad CPC464 4.0 MHz Z80A 140 seconds Tandy 64k CoCo 2 ? MHz 6809E 197 seconds Commodore 64 1.0 MHz 6510 254 seconds Commodore Plus/4 1.0 MHz 8501 267 seconds Tandy 64k CoCo 2 0.895 MHz 6809E 271 seconds Atari 800XL 1.8 MHz 6502 316 seconds Spectrum 128k +3 3.55 MHz Z80A 388 seconds Corrections to the CPU details would be appreciated if any are required - is that CPU really 1.0 MHz, or is it 0.9 MHz or something. Is it a Zilog Z80A, or is the compatible NSC800 used? Does anyone know of a classic 8-bit computer faster than the Electron or slower than the Spectrum? You may notice there are two figures for the Tandy Colour Computer. The first is the result obtained when the infamous speed-up poke is used. For those who don't know, POKE 65495,0 will make your Dragon/Tandy run around 35% faster, unless you have an earlier model in which this poke apparently causes a crash. POKE 65494,0 returns the computer to normal speed. I suggest on faster computers a larger number of primes are calculated and when I originally wrote the program the INPUT statement was included to allow for this. On my 486sx25 in GWBasic, calculating 5000 primes takes 126 seconds compared to 5 seconds for 500 primes. For anyone who is curious, the 5000th prime is 48611. On the Acorn Electron, 5000 primes can be calculated if you set HIMEM to 32768 and don't mind the variables overwriting the screen. It takes 3880 seconds. (just under 65 minutes) All of the variables in the program are integer variables, except for S in line 30, though this could easily be altered. Remember though, the prime number stored in A will overflow a 16 bit signed integer after calculating 3500 or so primes. I look forward to hearing other comparisons of ancient computers, John. --- My other machine is a Pentium * Origin: Darren's Little Castle BBS - 33.6 - (+44-1462-621049) (2:257/71) Message: #1125 (Read 6 times, has 0 replies, 1228 bytes) Date : Sat Oct 4 01:30:43 1997 From : Chris Baillies of fidonet#2:256/651.12 To : Glenn Richards Subject: Re: 8 Bit benchmarks Hello Glenn! On Saturday, 27 September 1997 06:33:59, you scribbled to Chris Baillies About: Re: 8 Bit benchmarks CB> GR>> I just ran it on my RISC PC for a laugh, 0.88 seconds. :-) CB> > Pentium 90 (AMD K5) (qbasic) 0.97 secs GR> My RPC has a 30MHz ARM610 CPU. Say no more. :-) A faster BASIC helps a little. Whats on the RPC, BASIC VI? The only Arc I have round here is an ancient (upgraded) 305... GR> I don't have a StrongARM to hand, but I would guess at around GR> 0.05 seconds. Probably! Don't get me wrong, I'm no PC advocate. -=> BFN, Christian <=- Kind regards, Chris Baillies - Guisborough, Saturday, 4 October 1997 at 1:30. Fido : 2:256/651.12 - Internet : Chrisb@spuddy.mew.co.uk URL : Welcome at the TmNice Home Page http://www.beechwood.com .!. Mary had a little lamb. The doctor was surprised. --- Terminate 3.00/Pro * Origin: When did you last warm yourself with a Terminate! (2:256/651.12) Message: #1125 (Read 11 times, has 0 replies, 1228 bytes) Date : Sat Oct 4 01:30:43 1997 From : Chris Baillies of fidonet#2:256/651.12 To : Glenn Richards Subject: Re: 8 Bit benchmarks Hello Glenn! On Saturday, 27 September 1997 06:33:59, you scribbled to Chris Baillies About: Re: 8 Bit benchmarks CB> GR>> I just ran it on my RISC PC for a laugh, 0.88 seconds. :-) CB> > Pentium 90 (AMD K5) (qbasic) 0.97 secs GR> My RPC has a 30MHz ARM610 CPU. Say no more. :-) A faster BASIC helps a little. Whats on the RPC, BASIC VI? The only Arc I have round here is an ancient (upgraded) 305... GR> I don't have a StrongARM to hand, but I would guess at around GR> 0.05 seconds. Probably! Don't get me wrong, I'm no PC advocate. -=> BFN, Christian <=- Kind regards, Chris Baillies - Guisborough, Saturday, 4 October 1997 at 1:30. Fido : 2:256/651.12 - Internet : Chrisb@spuddy.mew.co.uk URL : Welcome at the TmNice Home Page http://www.beechwood.com .!. Mary had a little lamb. The doctor was surprised. --- Terminate 3.00/Pro * Origin: When did you last warm yourself with a Terminate! (2:256/651.12) Message: #1126 (Read 6 times, has 0 replies, 796 bytes) Date : Sat Oct 4 17:41:07 1997 From : Stephen Brookes of fidonet#2:259/2.15 To : John Metcalf Subject: 8 bit benchmarks Hi John, JM> Corrections to the CPU details would be appreciated if any are JM> required - is that CPU really 1.0 MHz, or is it 0.9 MHz or something. If you want to be that accurate, the C64 is 1.02MHz NTSC and 0.98 PAL ;) Since we had such a poor result from an Amiga, here's a better one. Amiga1200(Blitz Basic) does 500 in 0.16s and 5000 in 4.49s. (Blitz cheats though, it's a non-interpretive Basic...) Stephen --- Cyclone 1.25c #0016 * Origin: TPEC (2:259/2.15)