Hi,
Just a quick message to tell you that the CPU speed computation seems to be
limited
to 50000, with a 450Mhz you get like 45000 and with a 1Ghz you only get
50000.
(from “sin in” or “sin ne”)
Regards.
I believe that the speed index that is reported by sin is an average of a
number of different things. The CPU speed is just one of them. Having said
that, it does sound like you’re not seeing enough of a speed increase. You
might want to check things like cache options and make sure that everything
is configured optimally.
On the other hand, it may be that there is a theoretical limit of 50000 on
the speed index calculation. Someone from QSSL will have to respond.
Bill Caroselli
“jhr” <nospam@nospam.com> wrote in message news:9lr05c$727$1@inn.qnx.com…
Hi,
Just a quick message to tell you that the CPU speed computation seems to
be
limited
to 50000, with a 450Mhz you get like 45000 and with a 1Ghz you only get
50000.(from “sin in” or “sin ne”)
Regards.
jhr (nospam@nospam.com) wrote:
: Hi,
: Just a quick message to tell you that the CPU speed computation seems to be
: limited
: to 50000, with a 450Mhz you get like 45000 and with a 1Ghz you only get
: 50000.
: (from “sin in” or “sin ne”)
Closer to the truth to say that it is modulo 65536 
From my home network:
Nid Machine Cpu Fpu Speed Memory Hard Flop Other Display Flags
1 PCI 686 687 19117 133.7M 99191M 1.4M TCP,cd,tp VGA Color -3+±—8P
2 PCI 686 687 64101 536.3M 35673M 1.4M cd VGA Color -3+±—8P
3 PCI 686 687 40597 268.0M 8257M 1.4M cd VGA Color -3+±—8P
6 PCI 686 687 42699 536.4M 8257M 1.4M cd VGA Color -3+±—8P
Node 1 is a 750 MHz; node 2 is a 1.3 GHz, node 3 is a 350 MHz, and node 6 is a 1 GHz.
Cheers,
-RK
–
Robert Krten, PARSE Software Devices; email my initials at parse dot com
Consulting, Systems Architecture / Design, Drivers, Training, QNX 4 & Neutrino
Check out our new QNX 4 and Neutrino (QRTP) books at http://www.parse.com/
Wanted PDP-8/9/10/11/12 Systems/documentation/spare parts! Will trade books!
Quite amazing to think that they made the same mistake again.
What would it have taken to say:
If calculation overflow
speed = 64K;
Previously, Robert Krten wrote in qdn.public.qnx4:
jhr (> nospam@nospam.com> ) wrote:
: Hi,: Just a quick message to tell you that the CPU speed computation seems to be
: limited
: to 50000, with a 450Mhz you get like 45000 and with a 1Ghz you only get
: 50000.: (from “sin in” or “sin ne”)
Closer to the truth to say that it is modulo 65536 >
From my home network:
Nid Machine Cpu Fpu Speed Memory Hard Flop Other Display Flags
1 PCI 686 687 19117 133.7M 99191M 1.4M TCP,cd,tp VGA Color -3+±—8P
2 PCI 686 687 64101 536.3M 35673M 1.4M cd VGA Color -3+±—8P
3 PCI 686 687 40597 268.0M 8257M 1.4M cd VGA Color -3+±—8P
6 PCI 686 687 42699 536.4M 8257M 1.4M cd VGA Color -3+±—8PNode 1 is a 750 MHz; node 2 is a 1.3 GHz, node 3 is a 350 MHz, and node 6 is a 1 GHz.
Cheers,
-RK–
Robert Krten, PARSE Software Devices; email my initials at parse dot com
Consulting, Systems Architecture / Design, Drivers, Training, QNX 4 & Neutrino
Check out our new QNX 4 and Neutrino (QRTP) books at > http://www.parse.com/
Wanted PDP-8/9/10/11/12 Systems/documentation/spare parts! Will trade books!
–
Mitchell Schoenbrun --------- maschoen@pobox.com
Mitchell Schoenbrun <maschoen@pobox.com> wrote:
Quite amazing to think that they made the same mistake again.
What would it have taken to say:If calculation overflow
speed = 64K;
Well it is the same problem… but we didn’t make the same mistake.
The mistake was to base some (critical to proper operation of certain
hardware) timing loops on that speed value. In QNX4, nothing in OS
code makes any decisions based on that speed value – it is purely for
display purposes.
So, while it doesn’t look pretty, it doesn’t affect how the system
runs. (Which is, to my eyes at least, the important part.)
And, if I remember my history properly, that bit of QNX4 was designed
BEFORE we ran into any machines where there was a problem with QNX2.
(So, we hadn’t (yet) had a chance to completely learn from that mistake.)
-David
QNX Training Services
dagibbs@qnx.com
Previously, David Gibbs wrote in qdn.public.qnx4:
Mitchell Schoenbrun <> maschoen@pobox.com> > wrote:
Quite amazing to think that they made the same mistake again.
What would it have taken to say:If calculation overflow
speed = 64K;Well it is the same problem… but we didn’t make the same mistake.
The mistake was to base some (critical to proper operation of certain
hardware) timing loops on that speed value. In QNX4, nothing in OS
code makes any decisions based on that speed value – it is purely for
display purposes.
Well that was a second problem. The problem I refer to is that it
rolls over. The QNX2 problem would not have shown itself for quite
a few years, and would not have needed a patch.
So, while it doesn’t look pretty, it doesn’t affect how the system
runs. (Which is, to my eyes at least, the important part.)
Well I’ll give you that.
And, if I remember my history properly, that bit of QNX4 was designed
BEFORE we ran into any machines where there was a problem with QNX2.
(So, we hadn’t (yet) had a chance to completely learn from that mistake.)
Yes, but you did have many releases subsequent to finding out about
the problem.
Of course one of the interesting things about this is that it shows
how difficult it is to not be short sighted with computers. I recall
that when the upped the QNX4 counter by a factor of 10, the attitude
was something like, there’ll be flying cars before this runs out.
Mitchell Schoenbrun --------- maschoen@pobox.com
Mitchell Schoenbrun <maschoen@pobox.com> wrote:
Previously, David Gibbs wrote in qdn.public.qnx4:
Mitchell Schoenbrun <> maschoen@pobox.com> > wrote:
Quite amazing to think that they made the same mistake again.
What would it have taken to say:If calculation overflow
speed = 64K;Well it is the same problem… but we didn’t make the same mistake.
The mistake was to base some (critical to proper operation of certain
hardware) timing loops on that speed value. In QNX4, nothing in OS
code makes any decisions based on that speed value – it is purely for
display purposes.Well that was a second problem. The problem I refer to is that it
rolls over. The QNX2 problem would not have shown itself for quite
a few years, and would not have needed a patch.So, while it doesn’t look pretty, it doesn’t affect how the system
runs. (Which is, to my eyes at least, the important part.)Well I’ll give you that.
And, if I remember my history properly, that bit of QNX4 was designed
BEFORE we ran into any machines where there was a problem with QNX2.
(So, we hadn’t (yet) had a chance to completely learn from that mistake.)Yes, but you did have many releases subsequent to finding out about
the problem.Of course one of the interesting things about this is that it shows
how difficult it is to not be short sighted with computers. I recall
that when the upped the QNX4 counter by a factor of 10, the attitude
was something like, there’ll be flying cars before this runs out.
Whew, then this covers our butt then:
sorry, just passing though and read the flying car comment and I couldn’t
resist. 
\
Mitchell Schoenbrun --------- maschoen@pobox.com