bryce@zen.UUCP (09/17/87)
Memory | Type | Workbench | HAM picture |
-----------------------------------------------------------
Amiga 1000 chip | chip | reference*| -22%
Amiga 500 chip | chip | 0% | -22%
Amiga 2000 chip | chip | 0% | -22%
A501 internal (A500) | half**| 0% | -22%
M501 internal (A500) | half | 0% | -22%
S501 internal (A500) | half | 0% | -22%
"internal 512K hack" | half | 0% | -22%
Defective A500 | chip | -8% | n/a
--------------------------------------------------
Known defective 1.5 Meg | fast | -7.00% | same ****
Cltd / aMEGA | fast | -6.53% | same
Kickstart WCS RAM | fast | +7.52% | same
Spirit 1.5 Meg $C00000 | fast | +5.13% | same
Insider $C00000 | fast | +5.13% | same
Allegra | fast |? +2% ?| same
Microbotics Starboard 2 | fast | +1.0% | same
--------------------------------------------------
ASDG | fast | | same
Comspec AX1000 | fast | | same
RAM-BOard-1 | fast | | same
Academy 1 Meg. | fast | | same
PAL internal | fast | | same
PAL jr internal | fast | | same
Byte-By-Byte Zorro | fast | | same
Xpander II | fast | | same
RS Data systems | fast | | same
Tecmar T card | fast | | same
Supra | fast | | same
MaxiRAM | fast | | same
Aminetics | ????
Cardinal Jumbo RAM | ????
* The chip memory in an Amiga 1000 is the reference point for
all measurements.
** "half" means "half-fast" memory. To learn it's qualities, say
it's name 10 times quickly. The A501 and friends can be changed
into chip memory by future upgrade, eliminating the derogitory name)
*** Blitter speed is not affected by your choice of memory expansion.
The blitter, bimmer and copper speed is the same for the A500, A1000
and A2000. This is a test of CPU throughput.
**** True fast memory is not affected by even fancy graphics modes.
You are welcome to add to the summary. Here is the program used:
begin 644 SpeedT
M```#\P`````````!```````````````L```#Z0```"PL>``$3J[_B`AY``$`H
MO^`!$_P`C`"_X0$3_`#_`+_C`2(\````$"`\``#__TYQ4<C__%')__((>0`!U
M`+_@`1/\`(,`O^$!(CP``!'`(#P``/__3G%1R/_\4<G_\@AY``$`O^`!(CP`Z
M```0(#P``/__3G%1R/_\4<G_\@AY``$`O^`!$_P`C`"_X0%.KO^"3G53<&5ES
@9"!T97-T+"!B>2!"<GEC92!.97-B:71T`````````_*_M
``
end
size 212
Be sure that you are displaying the standard 642*200 workbench. Make sure
that this program loads into the type of memory under test. To get a time
base, run "noFASTmem" from the Amiga 500/2000 Workbench disk. This will
time chip memory. Remove noFASTmem to time your expansion memory.
When started the LED will dim. In 2 seconds the LED will go bright again;
start your stopwatch now! Go away for 9 1/2 minuites. After that observe
the LED. Within 90 seconds or so it will go dim. Two seconds later it
will go bright as the signal to stop the watch. The machine is locked up
for the duration.
Record the base time (for internal chip memory), RAM time, computer type, and
the manufactuer and model of RAM under test. Send them to:
|\ /| . Ack! (NAK, ENQ, SYN)
{o O} .
(") bryce@hoser.berkeley.EDU -or- ucbvax!hoser!bryce
U How can you go back if you have not yet gone forth?spg@well.UUCP (Stephen P. Grant) (09/22/87)
In Article <3796@zen.berkeley.edu> bryce@hoser.berkeley.edu (Bryce Nesbitt) writes: > | PERCENT FASTER WITH *** | > | 642*200 | same size | >Memory | Type | Workbench | HAM picture | >----------------------------------------------------------- >Allegra | fast |? +2% ?| same What do you mean by this? ^^^^^^^^^^^ I have run "SPEEDT" on my A1000 at home with a 2 meg Alegra attached. Here are my results which I would have mailed to you if I was swift enough to figure out Usenet addresses. As suggested, these results were obtained on a standard 640 x 200 CLI window. I used "MemGrab" to disable Alegra for the CHIP timing. 2 MEG Alegra 09:44:72 09:44:83 CHIP ram 10:30:43 10:30:41 >|\ /| . Ack! (NAK, ENQ, SYN) >{o O} . > (") bryce@hoser.berkeley.EDU -or- ucbvax!hoser!bryce > U How can you go back if you have not yet gone forth? Stephen P. Grant Engineering Manager Access Associates