Received: from mailhost.viewlogic.com (mailhost.viewlogic.com [139.181.3.35]) by nacm.com (8.6.9/8.6.9) with SMTP id PAA08502 for <executor@nacm.com>; Tue, 8 Nov 1994 15:20:37 -0800
Received: from hari by mailhost.viewlogic.com (5.0/SMI-4.1)
id AA21603; Tue, 8 Nov 1994 18:17:06 +0500
From: sfoskett@viewlogic.com (Stephen Foskett)
Received: by hari (4.1) id AA07366; Tue, 8 Nov 94 18:20:03 EST
Message-Id: <9411082320.AA07366@hari>
Subject: 1.99 demo benchmarks and questions
To: executor@nacm.com
Date: Tue, 8 Nov 1994 18:20:02 -0500 (EST)
X-Mailer: ELM [version 2.4 PL23]
Content-Type: text
Content-Length: 3338
Sender: Executor-Owner@nacm.com
Precedence: bulk
I also downloaded 1.99b demo to see if executor has gotten far enough
to be a purchasable item... I have to support PCs, Suns and Macs here
at VIEWlogic, and I've only got enough desk space for a PC and a
Sparc, leaving the mac users cold.
1.99b does look great. Much improved over 1.2! The color is a nice
addition, and the speed is MUCH improved! (See speedo tests below...)
Now for the criticism... *grin*
The colors don't look "right". The Mac II used subdued, pastel tones
while executor, though the same colors, uses bright ones. It really
hurts the "look" of the product.
Also, I've been playing with the -*bpp switches. If I set it to 1, I
get monochrome. 4 gets 4-bit, 8 gets 8-bit (nice!). The most fun,
though, it 2. If -macbpp is set to 2, Speedometer (at least) can
switch pallettes like a real mac! It was able to do monochrome, 2, 4,
and 8 bit color tests! All the other settings locked executor into
that depth. Is this a bug or a feature? It would be nice to have the
thing in 8-bit color all of the time and still be able to self-switch
to other depths... Maybe a color setting in the prefs popup?
And what do those other prefs options do? Screen Updates seemed to
have no effect... Direct Disk Access slowed my performance down!
Flush Cache Often REALLY slowed the thing down! What's Refresh and
Cache Heuristics?
Also, at least in Speedometer (I used 3.1), the "display gauge" didn't
work except in 8-bit mode. Even in monochrome, it came through as a
big black block.
As for speed, here are my results:
Speedometer Report for ?????. Prepared December 31, 1969.
^^^^^^^^^^^^^^^^^
I knew the Mac II was old, but 1969??? Does Executor not read
the PCs clock?
Machine Record Version #: 5
System Information:
Computer: Mac II
CPU: MC68040
FPU: No FPU
MMU: Integral MMU
Color Quickdraw: 2.10 (32 Bit QD)
System Version: 6.0.7
Finder Version: Not Found
AppleTalk Version: Not Loaded
LaserWriter Version: Not Found
Laser Prep Version: Not Found
ImageWriter Version: Not Found
ROM Version: $0075
ROM Size: 128
Bit Depth: 2
Horizontal DPI: 72
Vertical DPI: 72
Primary Screen Size: 640 x 480
Physical RAM: 3072K
Logical RAM: 3072K
P.R. Results (Uses Mac Classic as 1.0):
CPU: 13.840
Graphics: 3.846
Disk: 102.066 Name of Hard Disk tested: /
Math: 30.661
Performance Rating (PR): 30.169
Benchmark Results (Uses Mac Classic as 1.0):
KWhetstones: 476.190 65.231
Dhrystones: 8498.583 8.722
Towers: 10.758
QuickSort: 11.976
Bubble Sort: 15.882
Queens: 15.266
Puzzle: 17.434
Permutations: 10.815
Fast Fourier: 24.308
F.P. Matrix Multiply: 24.295
Integer Matrix Multiply: 17.285
Sieve: 21.732
Benchmark
Average: 20.309
Color Tests (Uses Mac II as 1.0):
Black & White: 0.916
4 Colors: 1.115
16 Colors: 1.359
256 Colors: 2.267
Color Test Average: 1.414
So there you have it!
This computer was a Gateway P4D-66. 486DX2 66, PCI video, lots of
RAM. This test was done with -macbpp 2 and -applzone 4 under Executor
1.99b.
The color stuff is a bit faster than a Mac II. The processor stuff is
about 20 times the speed of a Mac Classic. (which sported an 8 MHz
MC68000)
Not bad at all!
For comparison, the processor rating on the same computer of Executor
