es1o+@andrew.cmu.edu (Eric Mitchell Snider) (12/13/90)
I just bought a IIsi 40/2. I had the 1 meg of 256K SIMMs from the si put into my Plus. I want to put my 4 megs of RAM from my Plus into the IIsi. The memory from the Plus consists of two 70ns 1 meg SIMMs and two 100ns 1 meg SIMMs. Is there any problem with doing this? MacWorld says "The machine requires 100 nanosecond RAM chips, but Apple is providing 80ns ones. Thus users who have the IIsi and IIci can swap SIMMs between the two." So 100ns should be fast enough but is it ok to mix 2 70ns and 2 100ns? The technician that was swapping the memory wrote on the work order "100ns do not work in SI's." Eric Snider es1o@andrew.cmu.edu
woods@convex.com (Darrin Woods) (12/13/90)
In article <UbNemAq00WBK43LV9y@andrew.cmu.edu> es1o+@andrew.cmu.edu (Eric Mitchell Snider) writes: >I just bought a IIsi 40/2. I had the 1 meg of 256K SIMMs from the si >put into my Plus. I want to put my 4 megs of RAM from my Plus into the >IIsi. The memory from the Plus consists of two 70ns 1 meg SIMMs and two >100ns 1 meg SIMMs. Is there any problem with doing this? MacWorld says >"The machine requires 100 nanosecond RAM chips, but Apple is providing >80ns ones. Thus users who have the IIsi and IIci can swap SIMMs between >the two." So 100ns should be fast enough but is it ok to mix 2 70ns and >2 100ns? > 100ns chips work just fine - go tell the tech to go back to school. I don't know if I would recommend mixing the two. The si works in banks of 4 like the II, IIx, cx, ci SE/30 fx... All 4 SIMMS should be the same speed. It usually works just fine to mix speeds across two different banks, but Apple (and I) do not recommend mixing in a bank because of the way that the mac uses the SIMMS in the bank. One SIMM will be receiving (sorry, working with) info at 70ns and the other at 100ns, and it will probably cause problems. Blacksheep Senior Systems Engineer -- Darrin R. Woods woods@convex.com This is a guest account. Convex knows nothing about what I'm saying, or even that I'm saying it.
amanda@visix.com (Amanda Walker) (12/15/90)
In article <110992@convex.convex.com> woods@convex.com (Darrin Woods) writes: >100ns chips work just fine - go tell the tech to go back to school. >[...] All 4 SIMMS should be the same speed. >[...] Apple (and I) do not recommend mixing in a bank because of the >way that the mac uses the SIMMS in the bank. One SIMM will be receiving >(sorry, working with) info at 70ns and the other at 100ns, and it will >probably cause problems. > >Blacksheep >Senior Systems Engineer Bzzzzzt. Maybe you should go back to school too... :) This myth seems amazingly persistent in the Mac community. It makes no difference to anything except your pocketbook if a SIMM is faster than it needs to be. You can mix and match to your heart's content as long as each SIMM meets the *minimum* speed requirements. They just all have to be the same *size*. Pushing a SIMM faster than its rated speed can sometimes cause flakiness. Slower is not a problem. If you don't believe me, go to the closest engineering library or electronics distributor, find a manufacturer's databook, and read the data sheet for a dynamic RAM chip. The access time of a chip is simply the maximum amount of time that the chip is guaranteed to respond within. If a machine's memory circuitry is designed for a given speed of RAM, it just allows that much time before it reads the data off of the bus. If it's ready before that, it doesn't evaporate or anything :). -- Amanda Walker amanda@visix.com Visix Software Inc. ...!uunet!visix!amanda -- If we were meant to fly, we wouldn't keep losing our luggage.
russotto@eng.umd.edu (Matthew T. Russotto) (12/15/90)
In article <twsd2n13pa@visix.com> amanda@visix.com (Amanda Walker) writes: > >This myth seems amazingly persistent in the Mac community. It makes no >difference to anything except your pocketbook if a SIMM is faster than >it needs to be. You can mix and match to your heart's content as long as >each SIMM meets the *minimum* speed requirements. They just all have to >be the same *size*. Talk to the Apple.people who wrote Tech Note 176... They are the promulgators of the myth, and the people I always cite (as flame retardant) when I mention it. Does anyone know WHY Apple says this? -- Matthew T. Russotto russotto@eng.umd.edu russotto@wam.umd.edu .sig under construction, like the rest of this campus.
woods@convex.com (Darrin Woods) (12/17/90)
In article <1990Dec15.035554.15172@eng.umd.edu> russotto@eng.umd.edu (Matthew T. Russotto) writes: >In article <twsd2n13pa@visix.com> amanda@visix.com (Amanda Walker) writes: >> >>This myth seems amazingly persistent in the Mac community. It makes no >>difference to anything except your pocketbook if a SIMM is faster than >>it needs to be. You can mix and match to your heart's content as long as >>each SIMM meets the *minimum* speed requirements. They just all have to >>be the same *size*. > >Talk to the Apple.people who wrote Tech Note 176... They are the promulgators >of the myth, and the people I always cite (as flame retardant) when I >mention it. > >Does anyone know WHY Apple says this? >-- >Matthew T. Russotto russotto@eng.umd.edu russotto@wam.umd.edu Thank you. I have nothing against putting faster chips in a machine when there are chips already in the machine that are slower. As long as they all meet the required speed or faster this is acceptable. What I said didn't work was mixing speeds within a Bank. The mac requires that all chips within a given bank be the same speed. As to the earlier rebutal to my posting - you can sit in a lab all you want and test the SIMMS to your hearts desire, and you will probably acheive the same answers as stated earlier - it doesn't matter. Now sit in your lab with a Mac, and put different speed SIMMS within a bank and test out the circuits, that Mac cannot handle it. As to WHY Apple says this.. In this case it is because it is true. And I can testify that I have had the same results as Apple listed, as I have tried it. You get ID 01's and 03's. Blacksheep Senior Systems Engineer -- Darrin R. Woods woods@convex.com This is a guest account. Convex knows nothing about what I'm saying, or even that I'm saying it.
n67786@lehtori.tut.fi (Nieminen Tero) (12/17/90)
> Thank you. I have nothing against putting faster chips in a machine > when there are chips already in the machine that are slower. As long > as they all meet the required speed or faster this is acceptable. This is not exactly true now is it? I just went to move some 80ns SIMMs from my IIfx into my IIx. Guess What? They do not fit. Sockets are different sizes. Am I missing something here? Different speed SIMMs have different socket dimensions, right? Just the fact that fx RAMs are so called dual port rams. Dual port RAMs are only used in Mac fx, so it won't make much of a hassle (at least at the time being). For the curious and technically minded: dualport RAMs are such that provided certain conditions are met dualport rams can be written to and read from at the same time. The larger sockets are needed for this kinds of RAMs to operate. tim. -- Tero Nieminen Tampere University of Technology n67786@cc.tut.fi Tampere, Finland, Europe
time@tbomb.ice.com (Tim Endres) (12/17/90)
In article <111547@convex.convex.com>, woods@convex.com (Darrin Woods) writes: > Thank you. I have nothing against putting faster chips in a machine > when there are chips already in the machine that are slower. As long > as they all meet the required speed or faster this is acceptable. This is not exactly true now is it? I just went to move some 80ns SIMMs from my IIfx into my IIx. Guess What? They do not fit. Sockets are different sizes. Am I missing something here? Different speed SIMMs have different socket dimensions, right? tim.
bruner@sp15.csrd.uiuc.edu (John Bruner) (12/18/90)
In article <111547@convex.convex.com> woods@convex.com (Darrin Woods) writes: > Now sit in your > lab with a Mac, and put different speed SIMMS within a bank and test > out the circuits, that Mac cannot handle it. Vendors will sometimes supply parts that exceed their ratings (particularly as their process technology matures). Hence, a 100ns RAM chip might really be an 80ns chip. If the Mac really can't handle mixing RAM's of different speeds, then perhaps the only safe way to replace or add memory is to use SIMM's from the same lot. (Even then there may be speed variations among the chips.) Maybe this is the reason Apple puts 80ns memory in the new machines -- they can't count on all of the 100ns chips being the same speed...? -- John Bruner Center for Supercomputing R&D, University of Illinois bruner@csrd.uiuc.edu (217) 244-4476
hamilton@kickapoo.cs.iastate.edu (Jon Hamilton) (12/18/90)
time@tbomb.ice.com (Tim Endres) writes: [stuff deleted...] >I just went to move some 80ns SIMMs from my IIfx into my IIx. >Guess What? They do not fit. Sockets are different sizes. >Am I missing something here? Different speed SIMMs have different >socket dimensions, right? Wrong. The //fx SIMMs are different beasts than the rest of the Mac line uses. Different speeds of the same kind of SIMM all have the same socket size. jon >tim.
amanda@visix.com (Amanda Walker) (12/18/90)
In article <111547@convex.convex.com> woods@convex.com (Darrin Woods) writes: >As to WHY Apple says this.. In this case it is because it is true. >And I can testify that I have had the same results as Apple listed, >as I have tried it. You get ID 01's and 03's. Well, I'm not going to go so far as to dispute your experience, but I must say it's a very good trick, since SIMMs marked for the same speed have no guarantees of actually responding at the same speed (you'd have to do very careful die selection in order to do this, actually). Since a 100ns SIMM may well respond within 80ns, it's hard to see how putting an 80ns SIMM in will make any operational difference. I dunno about a IIsi, but I've run a Mac II with mixed speed SIMMs, and as long as they were all at least as fast as the machine's requirements, they worked just fine. Maybe there's an optical scanner inside the cases of newer Macs that lets the ROM read the markings off of the chips. That's about the only way I can think of to make a difference. Someone what to quote chapter and verse from the Tech Note? My copy is evidently out of date, since all it mentions is the size... -- Amanda Walker amanda@visix.com Visix Software Inc. ...!uunet!visix!amanda -- Remember: this isn't reality. This is Usenet.
evelyn@uts.amdahl.com (Evelyn Mast) (12/18/90)
In article <1CE00001.0b2um2@tbomb.ice.com> time@tbomb.ice.com writes: > >In article <111547@convex.convex.com>, woods@convex.com (Darrin Woods) writes: >> Thank you. I have nothing against putting faster chips in a machine >> when there are chips already in the machine that are slower. As long >> as they all meet the required speed or faster this is acceptable. > >This is not exactly true now is it? > >I just went to move some 80ns SIMMs from my IIfx into my IIx. >Guess What? They do not fit. Sockets are different sizes. > >Am I missing something here? Different speed SIMMs have different >socket dimensions, right? > >tim. My understanding of this situation (without having my documentation right in front of me) is that the SIMMs in the IIfx are TOTALLY different than those used in all the other current Macs. So I would not expect what you tried to work. Sorry. -- ------------------------------------------------------------------------------- Evelyn Mast evelyn@amdahl.com I didn't mean it - I swear evelyn@key.amdahl.com -------------------------------------------------------------------------------
woods@convex.com (Darrin Woods) (12/18/90)
>> as they all meet the required speed or faster this is acceptable. > >This is not exactly true now is it? > >I just went to move some 80ns SIMMs from my IIfx into my IIx. >Guess What? They do not fit. Sockets are different sizes. > >Am I missing something here? Different speed SIMMs have different >socket dimensions, right? > >tim. OK, so there is an exclusion. The fx is the only mac currently on the market that uses a different SIMM than the rest. It is designed similarly to ones on the Laserwriter IINTX - not the same speed, (untill now) but same design. It has more pin connections than the 'regular' SIMMS for the rest of the mac line. Different speed SIMMS are not different sizes. It is just an engineering anomaly in the IIfx, but will probably become a standard as apple produces faster macs. Blacksheep Senior Systems Engineer v: -- Darrin R. Woods woods@convex.com This is a guest account. Convex knows nothing about what I'm saying, or even that I'm saying it.
amanda@visix.com (Amanda Walker) (12/19/90)
In article <hamilton.661460834@kickapoo.cs.iastate.edu> hamilton@kickapoo.cs.iastate.edu (Jon Hamilton) writes: >time@tbomb.ice.com (Tim Endres) writes: >>Am I missing something here? Different speed SIMMs have different >>socket dimensions, right? > >Wrong. The //fx SIMMs are different beasts than the rest of the Mac line >uses. Different speeds of the same kind of SIMM all have the same socket >size. In particular, I believe that the IIfx uses "static column" memory, which when combined with the 68030's burst mode, ekes out some extra performance from a given basic speed of memory, at the cost of requiring extra support circuitry. -- Amanda Walker amanda@visix.com Visix Software Inc. ...!uunet!visix!amanda -- If you know what you're doing, how long it will take, or how much it will cost, it isn't research.