27 Oct
2004
27 Oct
'04
9:22 a.m.
---------- Forwarded message ----------
From: David Forbes
I was reading that 1961 BRL survey, and found the user reports
of the IBM 701 fascinating. They apparently had an average
uptime of about 2-4 hours. That doesn't seem like much by
today's standards, but then how often does Windows crash on you?
The amazing thing is how much they paid per calculation. $50,000
rental per month for a typical installation at 500 multiplies per
second is about 10,000 multiplies per dollar! We seem to get a
better price these days with a typical PC. It's about a trillion
multiplies per dollar, which is a factor of 100 million better.
--
--David Forbes, Tucson, AZ
http://www.cathodecorner.com/
28 Oct
28 Oct
1:12 p.m.
I was reading that 1961 BRL survey, and found the user
reports
of the IBM 701 fascinating. They apparently had an average
uptime of about 2-4 hours. That doesn't seem like much by
today's standards, but then how often does Windows crash on you?
One of the often told stories of the early days of computers is that of
the completely unreasonable downtime, due to tube failure. There is even a
myth (pretty much disproved) about AN/FSQ-7 SAGE techs wearing roller
skates to replace the constantly dying tubes.
Basically, tube computers were not that bad, all of the time. The use of
"average" is being distorted here. When a 701 or 650 or whatever was
placed into service, you could expect tubes to die out (rarely from
"burnout" - loss of emission was the main culprit) at an alarming
rate. One every two hours is probably not unreasonable, as the marginal
tubes in their infancy are weeded out. Once these tubes were gone - I have
heard this could be six months - tube computers were quite reliable. Often
whn an old IBM or Burroughs module is unearthed, a quick look at the
datecodes on the tubes will reveal that most are probably originals to the
machine.
William Donzelli
aw288(a)osfn.org
3:08 p.m.
I believe tht the reliability problem with the 701
was not hard tube failures, but transient errors
in the Williams tube memories. This problem was
corrected when IBM transitioned to core memory.
--Bill
--- William Donzelli <aw288(a)osfn.org> wrote:
I was reading
that 1961 BRL survey, and found the
user reports
of the IBM 701 fascinating. They apparently had
an
average
uptime of about 2-4 hours. That doesn't seem
like
much by
today's standards, but then how often does
Windows
crash on you?
One of the often told stories of the early days of
computers is that of
the completely unreasonable downtime, due to tube
failure. There is even a
myth (pretty much disproved) about AN/FSQ-7 SAGE
techs wearing roller
skates to replace the constantly dying tubes.
Basically, tube computers were not that bad, all of
the time. The use of
"average" is being distorted here. When a 701 or 650
or whatever was
placed into service, you could expect tubes to die
out (rarely from
"burnout" - loss of emission was the main culprit)
at an alarming
rate. One every two hours is probably not
unreasonable, as the marginal
tubes in their infancy are weeded out. Once these
tubes were gone - I have
heard this could be six months - tube computers were
quite reliable. Often
whn an old IBM or Burroughs module is unearthed, a
quick look at the
datecodes on the tubes will reveal that most are
probably originals to the
machine.
William Donzelli
aw288(a)osfn.org
4:47 p.m.
I believe tht the reliability problem with the 701
was not hard tube failures, but transient errors
in the Williams tube memories. This problem was
corrected when IBM transitioned to core memory.
Yes, you are correct - Williams Tube sucked.
Still, the myth of non-reliability of tube machines is still around.
William Donzelli
aw288(a)osfn.org
29 Oct
29 Oct
10:01 a.m.
On Thu, 28 Oct 2004, William Donzelli wrote:
Oh neither David nor I take it that way; "downtime" likely has
systemic implications. Peripheral equipment, problems (hardware
or otherwise) with newfangled tape drives, who knows.
More, it was a comment on user perception of reliability,
and an obvious and gratuitous jab at M$ :-)
I was reading
that 1961 BRL survey, and found the user reports
of the IBM 701 fascinating. They apparently had an average
uptime of about 2-4 hours. That doesn't seem like much by
today's standards, but then how often does Windows crash on you?
One of the often told stories of the early days of computers is that of
the completely unreasonable downtime, due to tube failure. There is even a
myth (pretty much disproved) about AN/FSQ-7 SAGE techs wearing roller
skates to replace the constantly dying tubes.
Basically, tube computers were not that bad, all of the time. o
7168
days inactive
7170
days old
4 comments
3 participants
participants (3)
-
aw288＠osfn.org -
tomj＠wps.com -
wmaddox＠pacbell.net