What I was talking about, is the ASIC, not the DMM itself.  AND, in the
context of the fact that [today] you would just not sell enough 8.5-digit
meters to justify the development of a new ASIC.  I imagine that ASIC took
many "spins" through the fab to get it right, and today that would be in
the millions.  Combine that with the cost to re-design this meter [the
team, the time, the resources, etc.], and you would end up with a meter
that was so expensive that nobody would want to buy it.

I also suspect that even more than the military, the semiconductor
manufacturers will use many of the 3458A's [thousands of them for each
manufacturer], and I can't imagine them investing in equipment like that
unless HP/Agilent/Keysight would guarantee [in writing] long term support
for their investment.

The 3458A is not a DMM that you would use for casual bench use [for working
on other equipment] unless what you are working on requires that kind of
precision.  For repair work, most people prefer a hand-held DMM or a
6.5-digit bench DMM if more precision is required.  So, the user interface
of the 3458A is not likely to get a "new updated look" for a long time [if
ever], because the market will simply not support the development costs.
Even if Keysight were to come out with a "new" 8.5-digit DMM that was just
a 3458A analog board attached to new digital electronics and an upgraded
touch display, it would not have any better specs over the 3458A, so many
that already have a 3458A would not upgrade.

In short, there is just no compelling business reason to spend a lot of
money developing a new 8.5+ digit DMM, and I don't see that idea going
anywhere.  There may in fact BE an ongoing redevelopment effort within
Keysight, but it is my opinion that the new design will never be released
into production.  The business executives in Keysight [who are pretty smart
people], will want to see a return on investment, and there will be none of
that, so why bother?

I happen to own an older HP-branded 3458A, and I think it is "the bee's
knees", and would not upgrade [at probably US$12K+] if a new 8.5+ digit
meter came out.  How about others on this list?  What say you?  If Keysight
were to release a "new" DMM with a new beautiful interface, but with the
same analog electronics as are in the 3458A, and they priced it at [say]
US$12K, would you sell your 3458A on fleaBay, and then buy one of these new
meters, even though the specs are exactly the same as your old meter?

There is no point of change for change sake. What would I get that I do not
have now?

If it was faster and more stable (cut the temp coefficient and noise and
have a 1 ppm / year drift rate) it may be enough for some to justify but
not me.

Maybe if they could put a 3458A in a 3457A box with the multiplexer setup
the 3457A has there could be some interest but they would still have to
give me more stability and accuracy and make it sample faster. In short
make it more accurate and more channels.

I just do not see any mass market for such a device.

The other upgrade problem you have is a lot of these units are hard wired
into tests that have been running the same software for years and all that
would have to be certified. The test documents have 3458A listed as the
meter to use and it takes a lot of work to get someone to sigh off on a
different meter even if it is better.

There are cases where old 4.5 digit meters live because it is more cost
effective to replace the old meters with the same old model than to get
approval for  a new lower cost meter.
​