Pete Turnbull wrote:
> > On Dec 15, 22:42, Jerome H. Fine wrote:
> > Approximate speed increase from M8190-AB (KDJ11-BB)
> > with normal DEC memory (obviously below the CPU) vs
> > M8190-AE (KDJ11-BF) with DEC PMI memory above
> > the CPU was about 33%. An assembly which took about
> > 4.5 minutes was reduced to about 3 minutes. I can't recall
> > the exact numbers or exactly what I tried, but the test with
> > the M8190-AB and the PMI memory above the CPU took
> > about 23% less time than 4.5 minutes. I then tried the
> > M8190-AE and normal DEC memory below the CPU
> > and that took about 10% less time than 4.5 minutes. When
> > the PMI memory was below the M8190-AE, that also
> > took about 10% less than the 4.5 minutes.
> So, taking the M8190-AB as baseline, and doing some arithmetic:
> M8190-AB M8190-AE
> 15MHz 18MHz
> non-PMI 1.0 1.1
> PMI memory 1.3 1.5
> 18MHz is 1.2 times 15MHz, so it seems other operations on the bus
> are reducing the effect of the clock speed increase. Of course, that
> might also be due to using memory with a different cycle time. Still,
> it's clear that the PMI makes a bigger difference than the clock speed,
> going from 15MHz to 18MHz.
Jerome Fine replies:
YEP!!!
Of course, starting from a base of 1.0 with the M8190-AB, the
percentage is higher, but that is just the aspect of a reciprocal.
The other speed I found is that a PDP-11/93 is only about 10%
faster than the ( M8190-AE / PMI ) pair. So at the time that
the cost of PMI memory fell to its lowest, just switching memory
was the best solution.
But I also like the solution Megan Gentry suggested - change
the crystal on the M8190-AE to 20 MHz and see if it works.
But since that solution is 10 times more difficult for me than
using a 750 MHz Pentium III which has a speed factor of
about 15.0 (OK it is not that great for the CPU which might
be as low as 10.0 or just a 1000% factor, but overall when
the disk I/O is included is still probably 15.0 or 1500%),
I tend to stay with the Pentium III solution which does not
involve a soldering iron. In addition, using a 3.0 GHz P4
should allow for a factor above 50, maybe as high as 100.
I agree I don't hear those fans and disk drives of 250
GigaBytes can be standard (although under RT-11 I have
a hard time using an emulated drive of more than 2 GBytes),
but I guess I will have to grin and bear the difficulty.
Notice that I am not saying that you hardware addicts are
wrong and that me as a software addict is right - we each
have our own niche. In fact, I am beginning to really think
I must be crazy to want to make a Y10K version of RT-11
which I am almost sure no one will ever use - including ME!
But at least I will probably get to do things 100 times faster.
Sincerely yours,
Jerome Fine
--
If you attempted to send a reply and the original e-mail
address has been discontinued due a high volume of junk
e-mail, then the semi-permanent e-mail address can be
obtained by replacing the four characters preceding the
'at' with the four digits of the current year.
Received on Tue Dec 16 2003 - 23:02:03 GMT