home *** CD-ROM | disk | FTP | other *** search
Wrap
Received: from relay2.UU.NET (relay2.UU.NET [192.48.96.7]) by nacm.com (8.6.9/8.6.9) with ESMTP id WAA13698 for <executor@nacm.com>; Thu, 8 Dec 1994 22:51:00 -0800 Received: from sloth.swcp.com by relay2.UU.NET with SMTP id QQxtmp00798; Fri, 9 Dec 1994 01:50:55 -0500 Received: from iclone.UUCP (uucp@localhost) by sloth.swcp.com (8.6.9/8.6.9) with UUCP id XAA07381; Thu, 8 Dec 1994 23:51:17 -0700 Received: from gwar.ardi.com by mailhost with smtp (nextstep Smail3.1.29.0 #11) id m0rFz2P-000YdkC; Thu, 8 Dec 94 23:42 MST Received: by gwar.ardi.com (linux Smail3.1.28.1 #5) id m0rFz2N-000GPyC; Thu, 8 Dec 94 23:42 MST Message-Id: <m0rFz2N-000GPyC@gwar.ardi.com> Date: Thu, 8 Dec 94 23:42 MST From: mat@ardi.com (Mat Hostetter) To: GILMOUR@BINAH.CC.BRANDEIS.EDU Cc: executor@nacm.com Subject: Re: Speedometer-3.23 In-Reply-To: <01HKEYBC07W09C0M95@BINAH.CC.BRANDEIS.EDU> References: <01HKEYBC07W09C0M95@BINAH.CC.BRANDEIS.EDU> Sender: owner-executor@nacm.com Precedence: bulk >>>>> "GILMOUR" == GILMOUR <GILMOUR@BINAH.CC.BRANDEIS.EDU> writes: GILMOUR> In fact, I saved the results I got for both Executor/Linux GILMOUR> 1.99b & c, compared the two, and was suprised to find GILMOUR> that 1.99b is as much as 57% faster than 1.99c in some GILMOUR> tests. Executor got slower? That is most puzzling! Which tests? Is it consistently reproducible? GILMOUR> The comparison chart shows that Executor / GILMOUR> Linux seems on about a par with a Mac LC III on my machine. OK, but what type of machine do you have? GILMOUR> I think this is a great program for testing out Executor. GILMOUR> If ARDI doesn't already use this prog for benchmarking GILMOUR> their product, then they should look into it. I haven't GILMOUR> seen this program on Internet, so if anyone is interested GILMOUR> let me know where I should upload it to. Please don't GILMOUR> ask me to mail it to you, as I can't spend my time GILMOUR> mailing it to hundreds of people. Speedometer has definitely been useful to us! Speedometer should be available on numerous ftp sites, so there shouldn't be any need for you to upload it anywhere. Users should beware that Executor has problems with Speedometer 4.0. The synthetic CPU paper we posted to the net has Executor's Speedometer CPU emulation numbers at the end. In case you missed it, here they are again. These numbers were computed with Executor/DOS, but the CPU emulator is the same for all 80x86 Executors. By definition, a Mac Classic scores 1.0 in all benchmarks: Quadra Pentium 486DX4 486DX/2 610 90MHz 75MHz 66MHz ------ ------ ------ ------ CPU 16.018 28.833 15.727 13.840 Dhrystones 19.586 21.886 12.084 9.424 Tower 18.909 27.130 12.235 11.556 Quicksort 17.759 27.105 15.606 13.919 Bubble sort 18.409 31.154 19.286 16.875 Queens 19.083 38.167 19.083 18.320 Puzzle 22.083 44.167 23.661 21.032 Permutations 21.019 28.564 11.604 12.242 Int. Matrix 24.200 26.469 19.369 16.608 Sieve 23.362 60.290 33.982 30.145 ------ ------ ------ ------ Average 20.490 33.881 18.582 16.680 In fairness, our graphics aren't nearly as fast as our CPU emulation, especially when we have to go through X windows or NEXTSTEP. -Mat