Y Combinator is restoring one of Alan Kay's Xerox Alto machines

[Eric Smith]

On Fri, Sep 9, 2016 at 2:35 AM, Adrian Graham <binarydinosaurs at gmail.com> wrote:
> Apropos of 4116 RAM chips over the last 8 or so months I've restored 4 CBM
> PETs, 3 Apple ][s and a few Sinclair ZX Spectrums. All these use either
> 4116-2 or 4116-3 (or can use either as long as they're matched in 8s) and
> they've all had one or more RAM failures.

There's no reason why the speed grades of the RAMs have to be matched,
as long as they are all AT LEAST as fast as the machine requires.

Note that the speed grade of a DRAM only guarantees that it is at
least as fast as the grade; it may be faster.  For instance, an
MK4116-2 is rated for maximum 150 ns access, while an MK4116-3 is
rated for maximum 200 ns access.  If a machine is designed to work
with the MK4116-3 (maximum 200 ns access), either the MK4116-2 or
MK4116-3 will work, and the machine cannot distinguish them, as it's
entirely possible and likely that many 4116-3 chips are under typical
conditions actually just as fast as the 4116-2.  This is true whether
the mix of -2 and -3 parts are in the same bank or different banks.

Of course, the reverse is not true.  If the machine is designed to
require the 150 ns access DRAM, there's no way to be sure that a 200
ns rated DRAM will work.  It's possible that it may work most of the
time, but fail to deliver the correct data very rarely. Unfortunately
in systems without error checking, this can silently corrupt your
data, which may not be apparent until much later, e.g., if data in
memory is written to a disk file.