The Official MP problems page

The Gurus Certify
this Site is Cool.

The Official MP Problems Page

Get Email Chg Notices | Updates | Email Us | Upgrades Discussion Forum!

Recent Updates :

12/26/97-MacWorks is selling Genesis systems starting at $1499 (with single CPU card).
12/06/97- MacWorks has Genesis stripped and MP computers for sale. Supply is limited. Beware the many bargain basement MP cards for sale on the net - most are not returnable unless for the same card (DOA returns only)
10/27/97 - Tweek has new Daystar MP400 Cards for $699 (w/30 day defects warranty). Click here for details.
10/15/97 - Beware the Fire Sale on Newer MP cards - as reported in my Upgrades Forum.
8/12/97 - I throw in the towel on the Newer MP400 card - returned to Newer. It just does not work in any Mac I can find except with no cache and only 1 pair of dimms.
8/6/97 - Another user reports cache problems with Newer's MP cards.
8/6/97 - A developer reports cache problems with Newer's MP cards.
8/6/97 - A 9500 MP user writes that the Bravado 1000 capture card software prevents the MP libraries from loading at startup.
7/04/97 - David Burkhalter writes "I'd like to pass along a known problem with the MP system software and Metrowerks CodeWarrior development software. The MetroNub init will cause random crashes on machines with MP cards. Metrowerks is aware of the problem (they have delayed the introduction of the new version of the IDE that supports MP until the problem can be fixed.)." Thanks David.
7/03/97 - MacCPU says that a hardware fix to the MP cards is coming
(Quoting from MacInTouch page):
"Some customers have reported hardware problems with some of the MP processor cards. Daystar and Newer have identified the problem and cured it. Customers experiencing problems with the cards will have replacements sent shortly."
I did not see any reference to this at the MacCPU, Newer, or Daystar web sites. Maybe this page did some Good after all!
6/29/97 - An Italian reader sends a note of his 8500/MP fix.
6/28/97 - Apple replies MP only supported on 95/9600's - see updates.
6/27/97 - Apple TS person says there may be a display/database bug at the TS Online forum. This could prevent messages from displaying while browsing - but that searches should find all posts. They also say my last "missing" post was evidently never received. Whatever the reason for missing posts, at least it may get some attention now.
6/24/97- Users with MP problems - try this.
6/24/97- 8600 owner says new MP card stable only w/L2 cache removed.
6/23/97 - New Media's Summary of hi-perf Mac systems is worth reading.
6/23/97 - Added my Apple experiences during this ordeal.

6/21/97 - Extensive testing raises more issues than answers.

6/20/97 - Matched 64meg dimms (only) fail to work.

6/20/97 - Umax MP user reports many Type 1,2,3 errors & lockups w/7.6.1. However many more users report no problems.
6/19/97 - Good reports with OS7.6/MP outweight bad 10:1 so far.

6/19/97 - This User reports post OS 7.6 install problems.

6/19/97 - Another User duplicates our problems.
(the "blinking box" brings back memories...bad ones).
6/19/97 - Many MP users reply that OS 7.6/MP is running fine.
Memory configuration now suspect (again). Thanks to all who responded!

Purpose of this page:

Like other pages on this site, we hope that publishing the problems we've encountered will result in the following benefits:

Report specific MP issues/problems (& solutions hopefully).
Listing of workarounds or solutions from other users.
Provide a focused area where others can share their results.
Serve as a warning to other users about the problem.
Document the configurations that have the problem.
Document what has been attempted to solve the problem.
List updates to related system software.
If you have a MP system running under OS 7.6/7.6.1, please Contact Us.
Your name and email address will not be posted on the page if you prefer.

Multiprocessing under OS 7.6 ?

Over the last month or so, I have been unsuccessful in attempting to obtain reliable performance with OS 7.6/7.6.1 and a dual 200mhz 604E cpu card in my Mac 8500. I could fill several pages with the nightmare this project has become. At best the card would run for an hour or so, and then either fail to boot, generate type 1, 2, 3, and 10 errors, cause applications to "suddenly quit", corrupt data files, cause disks not to mount, and lock the machine (many times so hard I had to power off the outlet strip). Replacing cards, reformatting drives, clean OS installs, moving memory, changing cache, disabling extensions, zapping the pram, rebuilding desktop files, and updating MP software was not a solution. Each time replacing the MP card with a single cpu card restored stable operation. This was required in many cases to repair damaged disks or reinstall OS software without errors. Cache sensitivity was also a hallmark of the card, as it would only run with the Newer Tech. 256k cache, whereas the Newer 225 card ran with every cache I had. I'm currently awaiting a matched pair of 2k refresh, 64meg dimms from MacGurus to see if this will help solve the problem - I pray it will.

Until today I kept quiet, hoping that my mixed vendor memory may have been the cause. Since the MP card was very touchy about cache, I thought maybe it also was picky about memory timing as well. (The memory mixing and interleaving problems and oddities of the Apple 8500 motherboards are well known, as reported at MacInTouch on their PCI Ram problems page.) However after receiving several emails from other MP400 users experiencing similar problems I began to wonder if something else was the root cause.

Today I received the email below from another suffering MP owner that may shed some light on the problem:

A Reader with OS7.6/MP problems (like ours) wrote:

" Morning Mike,

Just spent another interesting 1 1/2 hours on the phone with tech support: so...let me tell you about yesterday first: clean installed 7.6.1, reinitialized the other hard drive with the latest version of the APS software, reinstalled software, nortoned and techtooled everything: switched cards: it started up perfectly! Amazing! it worked ...for three minutes: tried various rebooting methods: no extensions, from the CD: nothing but freezes and error messages. Shut down went back to the old card: another clean install, got everything set up, booted up from the CD used the second drive to run norton : I can not begin to tell you the damage that the 400 caused: software that I hadn't even tried to use was damaged: old system files shredded: hideous! It was as if a combine harvester had been through the machine. Have everything up and running on the old card now: probably a good idea to reinitialize my main drive (I'll get to that at the weekend unless necessity forces it upon me earlier).

So... tech support told me some interesting stuff: apparently the three people that he is aware of that have had the same gamut of problems are you, me and some guy in Australia: his problem was apparently solved by removing a couple of Dimms. One thing we all share is running system 7.6.1: apparently the mass of purchasers are using 7.5.5 with no problems. More news: Apple (supposedly only this last week) have informed Daystar and Newer that the way system software is written it will allow the second processor to run at only 48% of it's capacity: interesting? 7.6.1. is NOT multiprocessing compliant, but MP aware.(Note: Daystar informs me this is a misunderstanding, Apple has made no such statement) I would love for this thing to work but I fear I am giving up on this card and will go for one of the single processor cards. The odds seem stacked against having this thing up and fully functioning in the near future. Will one fully functioning, faster cpu be better than an iffy mp? It sure looks like it right now.

I am a photographer with a digital camera back and for the latest software I need to be running 7.6.1: with photoshop it would be great to have the mp capability, but not like this.(Tech Support) told me that as this is cutting edge technology there is bound to be some fine tuning involved but I am not prepared to go to the lengths of finding out the star sign of the packing manager to make sure that everything is just so. If everything on my current system is in fine working order now, then there is something wrong if this new technology is not compliant with that: either make a new machine and call it that, or make an upgrade that is just that; rather than this sorry state. Sorry, starting to gripe. It's just that over the last 2 years the number of dud components, software packages and peripherals that I have had to return has left a bad taste in my mouth: I'm getting fed up with having to pay to be a beta tester."

6/19/97 - A reader attempts to duplicate our problems (and succeeds).
He writes:

"I have an 8600 which after seeing your list of problems w/newers dual 200 card decided to see if I could duplicate. Original setup:8600/200 with 148mb RAM (all interleaved), 2 - 2gb seagate drives striped and split in two 2gb partitions (1st as startup), 2 - 2gb external drives, one partitioned into 2 1gb volumes, external CDR, external Syjet drive, ims twin turbo video card w/ 4.01 of drivers, system 7.6.1 and speed doubler 1.33

Installed newer MP software, shutdown, held in reset button, put in dual 200. restarted - got date and time incorrect message, and clicked OK. flashing blank white box-locked up.
restarted, extensions off. got past date message tried to drag finder prefs to trash-flashing box w/ error message hit reset button on CPU - flashing box doubled in size, hit keyboard restart and then cpu button to shut off.

Pulled out dual 200 and put in original cpu. everything is fine again.-"

6/19/97 - Another user reports post 7.6 install problems:

I have a PM 9500/180MP running 7.6.1 and since installing the system I regularly get a -192 error which corrupts both the Finder preference file and the desktop file.

You cannot escape the -192 error and the machine must be restarted. After the 192 error the machine will freeze at startup, crash, etc, etc.

The only solution I have found is a complete System rebuild - trashing the Finder preference and at least a daily rebuild of the desktop files will give limited relief.

The net result is I need to install a clean System about once a week - this only happened after going to 7.6.

Well at least there are 5 people in this unlucky club now...

Daystar/Newer Tech. News/Comments:

Daystar licenses the designs used by all major MP card manufacturers, including Apple and Newer. I have been impressed with their willingness to help, even with another vendors card. Their Tech Support has stated to me that my system software and hardware (OS 7.6, Xclaim VR, QT 2.5, QDraw3d 1.5.1) should have no problems with the MP cards. Daystar has informed me that OS 7.6 is shipping with their newest Genesis systems, and they recommend OS 7.6.1 as well. As stated above, Daystar says the user reported Apple comments on OS 7.6 MP efficiency were obviously a misunderstanding, as Apple has made no such statement to them.

Note:
There is a new set of floppy images for their MP software version 2.6 at this page. I'm waiting for the matched dimms to arrive to do further testing with the new software.

Newer Technology has gone the extra mile in helping determine the cause also. They replaced the initial card on the chance it was defective, and have offered many suggestions on possible solutions. The number of variations of software and hardware make debugging this problem a difficult task.
Update: 8/7/97
I've changed my mind about Newer's support. I've not heard WORD ONE from them in many weeks. I guess this 2nd bad card is somehow my system's fault, even though the detailed tests on the MP card prove it either defective or extremely poor design. Based on 5 Newer tech cards we've seen only ONE of them did not have problems. I call that more than just bad luck wouldn't you?

6/23/97: What about Apple?
Apple has commented that MultiProcessing is not supported on any machines except the 9500/9600. Thay also said problems regarding 3rd party hardware are not reasons for any warranty service (i.e. Swap my Rev A motherboard with a Rev B one). I've bought many Mac machines since 1984 (with the first 128k model) and this is the only time I've ever asked anything from them. I guess they've got their hands full with other issues.

Updates - breaking news will be posted here:

6/28/97 - Apple replies that MP is not supported on any system but the 95/9600. They also state problems involving 3rd party components, memory, cache and other addons are not reasons for requesting a motherboard swap, even though my motherboard is a rev A version. Daystar, however has gone the extra mile and offered to troubleshoot the system at their facility. If Apple provided the kind of support that companies like Daystar and Newer routinely do, then maybe they would not be in the situation they are now. You reap what you sow I guess.
During a new test run, with no L2 cache installed, the MP card failed all nPower memory tests except with only a matched set of 16meg dimms in slots A4/B4. Adding any other memory (even a matched pair) causes the memory test to fail. The MP card still passes all nPower processor tests.
The nPower control panel has proved to be invaluable in helping determine what the causes of the erratic behavior was, as other utilities did not detect the memory problems. nPower Control is a very impressive piece of software, proof positive that Daystar really knows the Mac inside and out (much better than Apple themselves I suspect). My gut still says the MP card is bad.
6/24/97 - It appears that my MP card is not satisfied with even the Newer cache, or else I've got a 2nd bad card in a row. Although the Newer ram gauge utility shows memory passing fine, the nPower control extensive memory test fails miserably, with several bad addresses and literally "getting stuck" at one point (screen test status does not scroll, and even begins to back up a line). Removing the MP card and retesting with the XLR8 card at the same bus/cpu speed passes the same extensive memory test. (BTW - the card passes all nPower processor card tests with flying colors.) MP users who have reported intermittent problems should perform the nPower control panel memory test (set to "extensive" and at least 5 minutes). I'd appreciate feedback on what the results are for those with problems.
6/21/97 - The matched 64meg dimms were received. They were identical TI 60ns dimms. They failed to solve the problem, generating type 10/11 errors and locking up the machine. Once it booted from the cdrom, and passed Newer's Ram Test gauge 80 times straight, then locked up when launching disk first aid. Dimms were tried in A4/B4, A1/B1, and A2/B2 with no luck. At this point I selected a matched pair of 16meg dimms (exact same mfr and type) and installed them in A4/B4. Same cache (Newer 256k) was used, and Daystar MP 2.6 update was installed ( we removed the older versions prior to installing, and threw away MP pref's also). Initial testing of the MP card vs. XLR8 card with this configuration (Newer 256k cache, latest MP software, only one set of matched 16meg dimms, clean OS 7.6 install) is done. I am awaiting replies from Daystar/Newer before publishing the results I saw, as they raise more questions than answers. Like the old saying, the more you stir it...
6/19/97 - The vast majority of users are running MP cards fine under OS7.6. The only variable in my system that changes it from a rock solid machine into a shaky one is replacing the CPU card. Since OS7.6 does not seem to be the culprit, then obviously the MP card *can* have problems in some systems with memory/cache that run fine with a single cpu card. I've proven that with the cards I have, beyond any doubt. Keeping that in mind may spare you the problems reported here. I just wish I had known that at the start of this exercise.
6/19/97 - Many Users have contacted me today reporting they are running OS 7.6/7.6.1 without problems. Most users were using 9500's but a few were running 8500's as well (one with a 1 meg cache!). One user reported rock solid performance with over 300meg of a mix of Dimms installed (8's,16's,64's), however most suggested that memory quality/matching was a prime consideration when populating their machines. Some users reported that their MP Mac's had run for months without errors. It appears the issue of mixed memory and/or a potential 8500 motherboard problem is now the prime suspect, however I can only speak for my case. (This was my opinion before I received the user feedback above.) Tomorrow I will be able to verify this, as a matched pair of 64meg dimms are due to arrive. I will test the 8500 with only that memory pair installed and report the results late friday/early Saturday (say a prayer for me will you?).
6/19/97 - Daystar called me, assuring me that OS 7.6/7.6.1 is running very well on both their Genesis systems and Mac's upgraded with MP processor cards. User Feedback also backs this up.
Daystar 2.6 MP update (while there get the Movie Pak update also).

Our Configurations:

Apple PM8500, OS 7.6/7.6.1 (clean installs)
Builtin video only (no other PCI cards)
64megs 60ns Dimms (four 16meg dimms)
Apple (Seagate) 2gig HD (reformatted/zero'd)
Daystar Dual processor Hal (not Apple's)
Newer Tech 256k Cache Dimm (others failed)
Dimms installed in A4/B4,A3/B3 (best bet)
No external SCSI devices connected
All usual tricks tried (Pram zap,desktop rebuild,clean installs,no extensions,etc.)
Other users configurations will be posted soon.

What's your Experience?:

Have experience with a MP product you'd like to share?
Want more information or have questions about the tests or results?
Contact Us with your feedback.

Enter your e-mail address to receive e-mail when this page is updated.
Your Internet e-mail address:

= Top of Page =
= Back To WWW.XLR8YOURMAC.COM =

2 Visitors since 7/01/97

Copyright © Michael Breeden, 1997.

No part of this sites content is to be reproduced in any form without written permission. All brand or product names mentioned here are properties of their respective companies.
Users of this web site must read and are bound by the terms and conditions of use.

Note! This page may cause a Type 1,2,3,10, or 11 error.
If it does, it must be your ram, cache or motherboard.
Please remove or replace them.
(4 weeks of hell - and I've still got a sense of humor!)

= Top of Page =
= Back To WWW.XLR8YOURMAC.COM =

	The Gurus Certify this Site is Cool.
The Official MP Problems Page
Get Email Chg Notices \| Updates \| Email Us \| Upgrades Discussion Forum!
Recent Updates :
12/26/97-MacWorks is selling Genesis systems starting at $1499 (with single CPU card). 12/06/97- MacWorks has Genesis stripped and MP computers for sale. Supply is limited. Beware the many bargain basement MP cards for sale on the net - most are not returnable unless for the same card (DOA returns only) 10/27/97 - Tweek has new Daystar MP400 Cards for $699 (w/30 day defects warranty). Click here for details. 10/15/97 - Beware the Fire Sale on Newer MP cards - as reported in my Upgrades Forum. 8/12/97 - I throw in the towel on the Newer MP400 card - returned to Newer. It just does not work in any Mac I can find except with no cache and only 1 pair of dimms. 8/6/97 - Another user reports cache problems with Newer's MP cards. 8/6/97 - A developer reports cache problems with Newer's MP cards. 8/6/97 - A 9500 MP user writes that the Bravado 1000 capture card software prevents the MP libraries from loading at startup. 7/04/97 - David Burkhalter writes "I'd like to pass along a known problem with the MP system software and Metrowerks CodeWarrior development software. The MetroNub init will cause random crashes on machines with MP cards. Metrowerks is aware of the problem (they have delayed the introduction of the new version of the IDE that supports MP until the problem can be fixed.)." Thanks David. 7/03/97 - MacCPU says that a hardware fix to the MP cards is coming (Quoting from MacInTouch page): "Some customers have reported hardware problems with some of the MP processor cards. Daystar and Newer have identified the problem and cured it. Customers experiencing problems with the cards will have replacements sent shortly." I did not see any reference to this at the MacCPU, Newer, or Daystar web sites. Maybe this page did some Good after all! 6/29/97 - An Italian reader sends a note of his 8500/MP fix. 6/28/97 - Apple replies MP only supported on 95/9600's - see updates. 6/27/97 - Apple TS person says there may be a display/database bug at the TS Online forum. This could prevent messages from displaying while browsing - but that searches should find all posts. They also say my last "missing" post was evidently never received. Whatever the reason for missing posts, at least it may get some attention now. 6/24/97- Users with MP problems - try this. 6/24/97- 8600 owner says new MP card stable only w/L2 cache removed. 6/23/97 - New Media's Summary of hi-perf Mac systems is worth reading. 6/23/97 - Added my Apple experiences during this ordeal. 6/21/97 - Extensive testing raises more issues than answers. 6/20/97 - Matched 64meg dimms (only) fail to work. 6/20/97 - Umax MP user reports many Type 1,2,3 errors & lockups w/7.6.1. However many more users report no problems. 6/19/97 - Good reports with OS7.6/MP outweight bad 10:1 so far. 6/19/97 - This User reports post OS 7.6 install problems. 6/19/97 - Another User duplicates our problems. (the "blinking box" brings back memories...bad ones). 6/19/97 - Many MP users reply that OS 7.6/MP is running fine. Memory configuration now suspect (again). Thanks to all who responded!
Purpose of this page:
Like other pages on this site, we hope that publishing the problems we've encountered will result in the following benefits: Report specific MP issues/problems (& solutions hopefully). Listing of workarounds or solutions from other users. Provide a focused area where others can share their results. Serve as a warning to other users about the problem. Document the configurations that have the problem. Document what has been attempted to solve the problem. List updates to related system software. If you have a MP system running under OS 7.6/7.6.1, please Contact Us. Your name and email address will not be posted on the page if you prefer.
Multiprocessing under OS 7.6 ?
Over the last month or so, I have been unsuccessful in attempting to obtain reliable performance with OS 7.6/7.6.1 and a dual 200mhz 604E cpu card in my Mac 8500. I could fill several pages with the nightmare this project has become. At best the card would run for an hour or so, and then either fail to boot, generate type 1, 2, 3, and 10 errors, cause applications to "suddenly quit", corrupt data files, cause disks not to mount, and lock the machine (many times so hard I had to power off the outlet strip). Replacing cards, reformatting drives, clean OS installs, moving memory, changing cache, disabling extensions, zapping the pram, rebuilding desktop files, and updating MP software was not a solution. Each time replacing the MP card with a single cpu card restored stable operation. This was required in many cases to repair damaged disks or reinstall OS software without errors. Cache sensitivity was also a hallmark of the card, as it would only run with the Newer Tech. 256k cache, whereas the Newer 225 card ran with every cache I had. I'm currently awaiting a matched pair of 2k refresh, 64meg dimms from MacGurus to see if this will help solve the problem - I pray it will. Until today I kept quiet, hoping that my mixed vendor memory may have been the cause. Since the MP card was very touchy about cache, I thought maybe it also was picky about memory timing as well. (The memory mixing and interleaving problems and oddities of the Apple 8500 motherboards are well known, as reported at MacInTouch on their PCI Ram problems page.) However after receiving several emails from other MP400 users experiencing similar problems I began to wonder if something else was the root cause. Today I received the email below from another suffering MP owner that may shed some light on the problem:
A Reader with OS7.6/MP problems (like ours) wrote:
" Morning Mike, Just spent another interesting 1 1/2 hours on the phone with tech support: so...let me tell you about yesterday first: clean installed 7.6.1, reinitialized the other hard drive with the latest version of the APS software, reinstalled software, nortoned and techtooled everything: switched cards: it started up perfectly! Amazing! it worked ...for three minutes: tried various rebooting methods: no extensions, from the CD: nothing but freezes and error messages. Shut down went back to the old card: another clean install, got everything set up, booted up from the CD used the second drive to run norton : I can not begin to tell you the damage that the 400 caused: software that I hadn't even tried to use was damaged: old system files shredded: hideous! It was as if a combine harvester had been through the machine. Have everything up and running on the old card now: probably a good idea to reinitialize my main drive (I'll get to that at the weekend unless necessity forces it upon me earlier). So... tech support told me some interesting stuff: apparently the three people that he is aware of that have had the same gamut of problems are you, me and some guy in Australia: his problem was apparently solved by removing a couple of Dimms. One thing we all share is running system 7.6.1: apparently the mass of purchasers are using 7.5.5 with no problems. More news: Apple (supposedly only this last week) have informed Daystar and Newer that the way system software is written it will allow the second processor to run at only 48% of it's capacity: interesting? 7.6.1. is NOT multiprocessing compliant, but MP aware.(Note: Daystar informs me this is a misunderstanding, Apple has made no such statement) I would love for this thing to work but I fear I am giving up on this card and will go for one of the single processor cards. The odds seem stacked against having this thing up and fully functioning in the near future. Will one fully functioning, faster cpu be better than an iffy mp? It sure looks like it right now. I am a photographer with a digital camera back and for the latest software I need to be running 7.6.1: with photoshop it would be great to have the mp capability, but not like this.(Tech Support) told me that as this is cutting edge technology there is bound to be some fine tuning involved but I am not prepared to go to the lengths of finding out the star sign of the packing manager to make sure that everything is just so. If everything on my current system is in fine working order now, then there is something wrong if this new technology is not compliant with that: either make a new machine and call it that, or make an upgrade that is just that; rather than this sorry state. Sorry, starting to gripe. It's just that over the last 2 years the number of dud components, software packages and peripherals that I have had to return has left a bad taste in my mouth: I'm getting fed up with having to pay to be a beta tester." 6/19/97 - A reader attempts to duplicate our problems (and succeeds). He writes: "I have an 8600 which after seeing your list of problems w/newers dual 200 card decided to see if I could duplicate. Original setup:8600/200 with 148mb RAM (all interleaved), 2 - 2gb seagate drives striped and split in two 2gb partitions (1st as startup), 2 - 2gb external drives, one partitioned into 2 1gb volumes, external CDR, external Syjet drive, ims twin turbo video card w/ 4.01 of drivers, system 7.6.1 and speed doubler 1.33 Installed newer MP software, shutdown, held in reset button, put in dual 200. restarted - got date and time incorrect message, and clicked OK. flashing blank white box-locked up. restarted, extensions off. got past date message tried to drag finder prefs to trash-flashing box w/ error message hit reset button on CPU - flashing box doubled in size, hit keyboard restart and then cpu button to shut off. Pulled out dual 200 and put in original cpu. everything is fine again.-" 6/19/97 - Another user reports post 7.6 install problems: I have a PM 9500/180MP running 7.6.1 and since installing the system I regularly get a -192 error which corrupts both the Finder preference file and the desktop file. You cannot escape the -192 error and the machine must be restarted. After the 192 error the machine will freeze at startup, crash, etc, etc. The only solution I have found is a complete System rebuild - trashing the Finder preference and at least a daily rebuild of the desktop files will give limited relief. The net result is I need to install a clean System about once a week - this only happened after going to 7.6. Well at least there are 5 people in this unlucky club now...
Daystar/Newer Tech. News/Comments:
Daystar licenses the designs used by all major MP card manufacturers, including Apple and Newer. I have been impressed with their willingness to help, even with another vendors card. Their Tech Support has stated to me that my system software and hardware (OS 7.6, Xclaim VR, QT 2.5, QDraw3d 1.5.1) should have no problems with the MP cards. Daystar has informed me that OS 7.6 is shipping with their newest Genesis systems, and they recommend OS 7.6.1 as well. As stated above, Daystar says the user reported Apple comments on OS 7.6 MP efficiency were obviously a misunderstanding, as Apple has made no such statement to them. Note: There is a new set of floppy images for their MP software version 2.6 at this page. I'm waiting for the matched dimms to arrive to do further testing with the new software. Newer Technology has gone the extra mile in helping determine the cause also. They replaced the initial card on the chance it was defective, and have offered many suggestions on possible solutions. The number of variations of software and hardware make debugging this problem a difficult task. Update: 8/7/97 I've changed my mind about Newer's support. I've not heard WORD ONE from them in many weeks. I guess this 2nd bad card is somehow my system's fault, even though the detailed tests on the MP card prove it either defective or extremely poor design. Based on 5 Newer tech cards we've seen only ONE of them did not have problems. I call that more than just bad luck wouldn't you? 6/23/97: What about Apple? Apple has commented that MultiProcessing is not supported on any machines except the 9500/9600. Thay also said problems regarding 3rd party hardware are not reasons for any warranty service (i.e. Swap my Rev A motherboard with a Rev B one). I've bought many Mac machines since 1984 (with the first 128k model) and this is the only time I've ever asked anything from them. I guess they've got their hands full with other issues.
Updates - breaking news will be posted here:
6/28/97 - Apple replies that MP is not supported on any system but the 95/9600. They also state problems involving 3rd party components, memory, cache and other addons are not reasons for requesting a motherboard swap, even though my motherboard is a rev A version. Daystar, however has gone the extra mile and offered to troubleshoot the system at their facility. If Apple provided the kind of support that companies like Daystar and Newer routinely do, then maybe they would not be in the situation they are now. You reap what you sow I guess. During a new test run, with no L2 cache installed, the MP card failed all nPower memory tests except with only a matched set of 16meg dimms in slots A4/B4. Adding any other memory (even a matched pair) causes the memory test to fail. The MP card still passes all nPower processor tests. The nPower control panel has proved to be invaluable in helping determine what the causes of the erratic behavior was, as other utilities did not detect the memory problems. nPower Control is a very impressive piece of software, proof positive that Daystar really knows the Mac inside and out (much better than Apple themselves I suspect). My gut still says the MP card is bad. 6/24/97 - It appears that my MP card is not satisfied with even the Newer cache, or else I've got a 2nd bad card in a row. Although the Newer ram gauge utility shows memory passing fine, the nPower control extensive memory test fails miserably, with several bad addresses and literally "getting stuck" at one point (screen test status does not scroll, and even begins to back up a line). Removing the MP card and retesting with the XLR8 card at the same bus/cpu speed passes the same extensive memory test. (BTW - the card passes all nPower processor card tests with flying colors.) MP users who have reported intermittent problems should perform the nPower control panel memory test (set to "extensive" and at least 5 minutes). I'd appreciate feedback on what the results are for those with problems. 6/21/97 - The matched 64meg dimms were received. They were identical TI 60ns dimms. They failed to solve the problem, generating type 10/11 errors and locking up the machine. Once it booted from the cdrom, and passed Newer's Ram Test gauge 80 times straight, then locked up when launching disk first aid. Dimms were tried in A4/B4, A1/B1, and A2/B2 with no luck. At this point I selected a matched pair of 16meg dimms (exact same mfr and type) and installed them in A4/B4. Same cache (Newer 256k) was used, and Daystar MP 2.6 update was installed ( we removed the older versions prior to installing, and threw away MP pref's also). Initial testing of the MP card vs. XLR8 card with this configuration (Newer 256k cache, latest MP software, only one set of matched 16meg dimms, clean OS 7.6 install) is done. I am awaiting replies from Daystar/Newer before publishing the results I saw, as they raise more questions than answers. Like the old saying, the more you stir it... 6/19/97 - The vast majority of users are running MP cards fine under OS7.6. The only variable in my system that changes it from a rock solid machine into a shaky one is replacing the CPU card. Since OS7.6 does not seem to be the culprit, then obviously the MP card can have problems in some systems with memory/cache that run fine with a single cpu card. I've proven that with the cards I have, beyond any doubt. Keeping that in mind may spare you the problems reported here. I just wish I had known that at the start of this exercise. 6/19/97 - Many Users have contacted me today reporting they are running OS 7.6/7.6.1 without problems. Most users were using 9500's but a few were running 8500's as well (one with a 1 meg cache!). One user reported rock solid performance with over 300meg of a mix of Dimms installed (8's,16's,64's), however most suggested that memory quality/matching was a prime consideration when populating their machines. Some users reported that their MP Mac's had run for months without errors. It appears the issue of mixed memory and/or a potential 8500 motherboard problem is now the prime suspect, however I can only speak for my case. (This was my opinion before I received the user feedback above.) Tomorrow I will be able to verify this, as a matched pair of 64meg dimms are due to arrive. I will test the 8500 with only that memory pair installed and report the results late friday/early Saturday (say a prayer for me will you?). 6/19/97 - Daystar called me, assuring me that OS 7.6/7.6.1 is running very well on both their Genesis systems and Mac's upgraded with MP processor cards. User Feedback also backs this up. Daystar 2.6 MP update (while there get the Movie Pak update also).
Our Configurations:
Apple PM8500, OS 7.6/7.6.1 (clean installs) Builtin video only (no other PCI cards) 64megs 60ns Dimms (four 16meg dimms) Apple (Seagate) 2gig HD (reformatted/zero'd) Daystar Dual processor Hal (not Apple's) Newer Tech 256k Cache Dimm (others failed) Dimms installed in A4/B4,A3/B3 (best bet) No external SCSI devices connected All usual tricks tried (Pram zap,desktop rebuild,clean installs,no extensions,etc.) Other users configurations will be posted soon.
What's your Experience?:
Have experience with a MP product you'd like to share? Want more information or have questions about the tests or results? Contact Us with your feedback.
Enter your e-mail address to receive e-mail when this page is updated. Your Internet e-mail address:

= Top of Page = = Back To WWW.XLR8YOURMAC.COM =


2 Visitors since 7/01/97 Copyright © Michael Breeden, 1997. No part of this sites content is to be reproduced in any form without written permission. All brand or product names mentioned here are properties of their respective companies. Users of this web site must read and are bound by the terms and conditions of use.
Note! This page may cause a Type 1,2,3,10, or 11 error. If it does, it must be your ram, cache or motherboard. Please remove or replace them. (4 weeks of hell - and I've still got a sense of humor!)
= Top of Page = = Back To WWW.XLR8YOURMAC.COM =