[net.micro.trs-80] Benchmarks on ][e, Mac, "PC", & 2000

2141smh@aluxe.UUCP (henning) (04/20/84)

80MICRO published an interesting comparison of benchmark results for
the Apple ][e, MacIntosh, IBM-PC, and Tandy 2000.  The results are:

TEST		   	   ][e		   Mac		   PC		   2000
=================	=======		=======		=======		=======
For loop x 5000		  5 sec		  5 sec		  7 sec		  2 sec
Goto loop x 5000	 41 sec		 17 sec		 24 sec		 10 sec
Variable math		 79 sec		 46 sec		 61 sec		 20 sec
Numeric math		 87 sec		 55 sec		 59 sec		 21 sec
Gosub + goto loop	 95 sec		 58 sec		 66 sec		 23 sec
Above + for loop	140 sec		 99 sec		117 sec		 43 sec
Above + string values	220 sec		142 sec		186 sec		 66 sec
Sieve of Eratosthenes	211 sec		109 sec		177 sec		 58 sec
Mid$ function		 31 sec		 22 sec		 23 sec		  9 sec
Bubble sort		 19 sec		 11 sec		 13 sec		  4 sec
Disc read & write	 ?		 ?		 20 sec		  9 sec
Screen write		 ?		 ?		 25 sec		  7 sec
=================	=======		=======		=======		=======
Overall rating		3.7 sec		2.3 sec		2.8 sec		1.0 sec

Conclusion:  The Tandy 2000 is over twice as fast as the competition.

betsy@dartvax.UUCP (Betsy Hanes Perry) (04/20/84)

>Conclusion:  The Tandy 2000 is over twice as fast as the competition.
 
Ummm... Well, my conclusion would be that Microsoft Basic for the
Mac is godawful slow.   The tests in question WERE run in Microsoft
Basic on the Mac, no?  

I question the portability of language-implementation-dependent benchmarks 
(like mid$) in any case.  When every micro was based on BASIC, and
very similar BASICS at that, it made some sense.  Nowadays, fewer
machines have resident BASICS, and there are zillions of different
BASICS available for any given machine.  (Check out the April Byte!)
Under the circumstances, I don't think you can count 'running the
same Basic program' as an independent variable anymore.

-- 
Betsy Perry
UUCP: {decvax|linus|cornell}!dartvax!betsy
CSNET: betsy@dartmouth
ARPA:  betsy%dartmouth@csnet-relay

nazgul@apollo.uucp (Kee Hinckley) (04/27/84)

---

    Cute benchmarks.  WHAT LANGUAGE WERE THEY USING?

                                -nazgul

2141smh@aluxe.UUCP (henning) (04/30/84)

****                                                                 ****
From the keys of Steve Henning, AT&T Bell Labs, Reading, PA aluxe!2141smh

The benchmarks were based on identical Micro-Soft Basic programs.