regan@cs.Buffalo.EDU (Kenneth Regan) (02/07/90)
In article <660@tci.bell-atl.com> kempf@tci.bell-atl.com (Cory Kempf) writes: >mitchell@tci.bell-atl.com (Rob Mitchell) writes: > >> Basically, subject says it all. > >Has anyone heard anything about when we can get 4MB SIMMS firthe mac >in quantity? (for a reasonable price?) > I'm putting four 70ns 1Mb x 9 SIMMs in my Mac IIcx, to augment the four Apple 120ns 1Mb x 9s already there. Technology Works assured me that the speed difference wouldn't matter, and that having the parity chip wouldn't hurt -- the IIcx would just ignore it. [Why spend more for 70ns? Maybe someone will come up with an add-in that will exploit faster RAM, or the chips will find another useful home someday. It's not much more.] Perhaps more interesting, TW told me they now have 80ns 4Mb x 9 SIMMs in stock, for only $590 each!!. We're ordering four for the SparcStation1; again, TW assured me they'll work in either the high-end Macs or the Sun workstations or etc. [Again, I'm a RAM neophyte, and even need to ask: "Once you specify (speed)(chip size) x (8 or 9), a SIMM is a SIMM in any machine--right??] Kenneth W. Regan Assistant Professor Computer Science Dept. (Opinions not < SUNYaB) SUNY at Buffalo, 226 Bell Hall Tel.: (716) 636-3189, -3180 Buffalo, NY 14260 regan@cs.buffalo.edu
yuan@uhccux.uhcc.hawaii.edu (Yuan 'Hacker' Chang) (02/10/90)
In article <17118@eerie.acsu.Buffalo.EDU> regan@castor.cs.Buffalo.EDU.UUCP (Kenneth Regan) writes:
-
-I'm putting four 70ns 1Mb x 9 SIMMs in my Mac IIcx, to augment the four
-Apple 120ns 1Mb x 9s already there. Technology Works assured me that the
-speed difference wouldn't matter, and that having the parity chip wouldn't
-hurt -- the IIcx would just ignore it. [ . . . ]
FYI -- The IIci (not the IIcx) will do parity detection when fed
x9 SIMMs. That's the good news. The bad news is that all it does with
that information is to ask you to restart (according to the manual). I'd
much prefer to be given a chance to recover from the error. If I'm working
on a 2M document on a 32M machine, I wouldn't want a flipped bit where it
doesn't count to cause me to loose my work. I think we'll stick with x8
SIMMs...
--
Yuan Chang "What can go wrong, did"
UUCP: {uunet,ucbvax,dcdwest}!ucsd!nosc!uhccux!yuan
ARPA: uhccux!yuan@nosc.MIL "Wouldn't you like to
INTERNET: yuan@uhccux.uhcc.Hawaii.Edu be an _A_m_i_g_o_i_d too?!?"russ@key.COM (Russell Donnan) (02/14/90)
In article <6530@uhccux.uhcc.hawaii.edu> yuan@uhccux.uhcc.hawaii.edu (Yuan 'Hacker' Chang) writes: {In article <17118@eerie.acsu.Buffalo.EDU> regan@castor.cs.Buffalo.EDU.UUCP (Kenneth Regan) writes: {- { FYI -- The IIci (not the IIcx) will do parity detection when fed {x9 SIMMs. That's the good news. The bad news is that all it does with {that information is to ask you to restart (according to the manual). I'd {much prefer to be given a chance to recover from the error. If I'm working {on a 2M document on a 32M machine, I wouldn't want a flipped bit where it {doesn't count to cause me to loose my work. I think we'll stick with x8 {SIMMs... {-- {Yuan Chang "What can go wrong, did" First of all, let me say that I hate parity memory as much as the next guy... BUT, I understand its usefulness. A flipped bit where it doesn't count will not make you lose work in another area. Parity is checked on a READ of memory. So if you get a parity error, it is either your current application and associated data, or the Mac OS which had troubles. What I would like to see is an opportunity to save my document if the parity error happens in the application. I realize you can't do this for the OS. How about it Apple? -Russ -- Russ Donnan (415) 623-2121 Amdahl Corporation, Key Computer Laboratories, Fremont, CA, USA russ@key.amdahl.com, ...!{pacbell,sgi,amdahl}!key!russ -To capture the essence of an opinion takes but one lawyer.
yuan@uhccux.uhcc.hawaii.edu (Yuan 'Hacker' Chang) (02/14/90)
In article <1446@key.COM> russ@xanadu.key.COM (Russell Donnan) writes:
-
-First of all, let me say that I hate parity memory as much as the next
-guy... BUT, I understand its usefulness. A flipped bit where it doesn't
-count will not make you lose work in another area. Parity is checked on
-a READ of memory. So if you get a parity error, it is either your
-current application and associated data, or the Mac OS which had troubles.
I thought parity is checked whenever the memory is refreshed.
After all, a refresh is pretty much a read-and-write cycle. But then,
maybe it is much more efficient just to parity check memory that is being
read...
--
Yuan Chang "What can go wrong, did"
UUCP: {uunet,ucbvax,dcdwest}!ucsd!nosc!uhccux!yuan
ARPA: uhccux!yuan@nosc.MIL "Wouldn't you like to
INTERNET: yuan@uhccux.uhcc.Hawaii.Edu be an _A_m_i_g_o_i_d too?!?"