10 Oct
2006
10 Oct
'06
3:05 p.m.
John Holden wrote:
On 05/10/2006, at 10:15 PM, Dave Horsfall wrote:
I got them here. :-)
Hmm, but yeah. Since the bus might be idle even though the CPU is
running, you might not get a very good reading by looking at the unibus.
But I think there are other signals that would be better to integrate
over. Such as the MASTER lamp (or is it the CPU lamp perhaps?) of the
front panel.
Johnny
On Wed, 4 Oct 2006, Brian Knittel wrote:
Um, depends what you mean by thrust. It was originally designed for a
11/45
where bus activity would have been a fare indication of machine load
since
all memory and i/o used the unibus (I'm excluding the effects of fastbus
memory on 11/50 and 55's).
The 11/70 unibus was actually slower than the 11/45, and generally
didn't
have memory or disk/tape i/o on it (separate memory bus and massbus). So
'thrust' was probably mainly character i/o (dh and dz's) and older
disk/tapes (RK05's etc)
Now, where's 11/70 maintenance printset? The thrust meter project -- was that an analog
meter that displayed %
CPU utilization? I remember that Tom Ferrin had one mounted in the
middle of a DEC panel filler on the 11/70 at the Computer Graphics
Lab
at UCSF. It was really delightful having this analog meter
bouncing up
and down as people worked away.
It integrated the BUS BUSY signal over a suitable time constant
(and my
electronics knowledge is a bit too rusty to figure out Tc, but with a
10uF tantalum capacitor I imagine it would be a few seconds).