K
KR Williams
Indeed. Some still like these "mistakes" though.
"Slimmed down" would be my description, but either works.
Indeed. Some still like these "mistakes" though.
"Slimmed down" would be my description, but either works.
K
Kai Harrekilde-Petersen
Well, "rest of the world" doesn't include Denmark, for one.
What Terje describes is standard practice here in Denmark. Well, was,
I suppose, as they power companies have now spun so many fibers across
the country that most of the fibers are dark, and it'll take a loooong
while before all of it ever sees light [sic].
The railway company, however, devices a special car that could plow
down fiber besides the tracks. For obvious reasons the railway
company use some of the fiber pairs for their on signalling (just
think about a heavy diesel-electric train braking. It can induce
several hundred volts in a (shielded) cable running parallel to the
tracks).
Not to mention the granite.
/Kai
What Terje describes is standard practice here in Denmark. Well, was,
I suppose, as they power companies have now spun so many fibers across
the country that most of the fibers are dark, and it'll take a loooong
while before all of it ever sees light [sic].
The railway company, however, devices a special car that could plow
down fiber besides the tracks. For obvious reasons the railway
company use some of the fiber pairs for their on signalling (just
think about a heavy diesel-electric train braking. It can induce
several hundred volts in a (shielded) cable running parallel to the
tracks).
Not to mention the granite.
/Kai
Z
Zak
Terje said:
Here's one that uses the 'ground' wire to hide the fibers. Not what we
were looking for but an interesting read anyway:
http://users.castel.nl/~berka01/hvinfo.html
Or see: http://www.esi.co.za/last/ESI_3_2002/ESI32002_066_1.htm
which at least mentions some methods:
* Embedded: OPGW – composite earth-wire and embedded fibre optic cable
* Externally attached: wrap technique
* Externally attached: lash technique (ADLash)
* Externally attached: ADSS (all dielectric self supporting)
* Externally attached: MASS (metal armoured self supporting).
"whereas the Eskom Enterprises technique of using helicopters and
live-line work teams to augment conventional techniques of attaching
fibre to powerlines has resulted in rollout speeds of 26 - 30 km per
day, without introducing outage time on the power network"
I guess they work on live lines from the helicopter and indeed install
cable ties every meter - eeks.
Thomas
G
Grumble
Black said:
The one-dollar question is:
Will Intel put the Mobile Pentium 4 and Dothan in the same category?
Or will they pretend that the Mobile Pentium 4 is just a *cough*
"power-efficient" desktop Pentium 4?
BTW, the Pentium M is rated "higher" than Prescott (7xx vs 5xx).
S
Sander Vesik
In comp.arch Nick Maclaren said:
I think you get very serious problems in totaly unpopulated areas with it
too. Consider for example that you need a way to get the people there, have
them be housed somewhere and get food / fuel to them. Totaly unpopulated
also tends to mean 'totaly roadfree'.
R
RusH
Grumble said:
300 -
Celeron M: 340: 1.5GHz, 330: 1.4GHz, 320: 1.3GHz
Celeron D (Prescott-V): 340: 2.93GHz, 335: 2.8GHz, 330: 2.66GHz, 325:
2.53GHz
500 -
Mobile Pentium 4
Pentium 4 (Prescott) 570: 3.8GHz, 560: 3.6GHz, 550: 3.4GHz, 540:
3.2GHz, 530: 3GHz, 520: 2.8GHz
700 -
Pentium M (Dothan): 755: 2GHz, 745: 1.8GHz, 735: 1.7GHz, 725: 1.6GHz,
715: 1.5GHz
Pentium 4 Extreme Edition
Pozdrawiam.
K
KR Williams
ChamBanna gets cold (and *hot*), but it's nothing like Rochester
MN. Friends on a trek to Rochester got literally blown out of
their hotel rooms by the glass shattering due to the cold.
Rochester is *cold*, even more so than Burlington (that's saying
something). The old IBM located in some god-forsaken places.
Oh yeah, it's cheap alright. That's why our "cheap" Hydro-Quebec
power costs me $.12/kWh.
R
Rob Stow
Where I am (Saskatchewan), the urban residential rate is currently
$0.0795/kWh and Toronto is currently $0.081/kWh according to someone
I correspond with there. Urban commercial rates in Sask are
currently $0.0745/kWh. I know rural rates are significantly
higher but I don't have a dollar figure for you.
BTW, those are Canadian dollar numbers I cited, not YankeeBucks.
Discount by about 25% or 30% to get a rough estimate in YankeeBucks.
Hydro-Q (and other Canadian power utilities) sell their excess
juice south of the border in what amounts to a spot market.
Long term deals for cheap power to most parts of the US market
seem to have become a thing of the past - they went out the
window when deregulation came in down south. Blame *your*
government for preventing your utilities from negotiating
long term deals for cheaper power. As well, there are
tariffs levied by *your* government on most power imported
from Canada, and, as always, that tariff gets passed onto the
consumer - *you* in other words. Time to write your
congresscritter perhaps ?
$0.0795/kWh and Toronto is currently $0.081/kWh according to someone
I correspond with there. Urban commercial rates in Sask are
currently $0.0745/kWh. I know rural rates are significantly
higher but I don't have a dollar figure for you.
BTW, those are Canadian dollar numbers I cited, not YankeeBucks.
Discount by about 25% or 30% to get a rough estimate in YankeeBucks.
Hydro-Q (and other Canadian power utilities) sell their excess
juice south of the border in what amounts to a spot market.
Long term deals for cheap power to most parts of the US market
seem to have become a thing of the past - they went out the
window when deregulation came in down south. Blame *your*
government for preventing your utilities from negotiating
long term deals for cheaper power. As well, there are
tariffs levied by *your* government on most power imported
from Canada, and, as always, that tariff gets passed onto the
consumer - *you* in other words. Time to write your
congresscritter perhaps ?
T
Terje Mathisen
Sander said:
I think he meant access to dig the trench, not to live there, but the
problem of nearly no topsoil on 97% of Norway still holds, i.e. blasting
a trench vs running a robot along the wire is really no contest.
Terje
B
Black Jack
Grumble said:
The Intel model numbers seem to be much more marketing driven than the
Opteron model numbers. That is, the Opteron 1-, 2-, and 8-series
models represent their ultimate multiprocessing capabilities, whereas
the Intel series numbers seem to be totally randomly derived. And the
processors that get put into the various series, also seem to be
randomly placed. Bound to cause more confusion among Intel customers.
Yousuf Khan
K
KR Williams
Gee, priced according to cost. What an interesting concept.
Here we subsidize the rural customers (flattened rates no matter
what the cost of delivery) just to be "fair".
Nope. Long-term contract. That's the problem.
Nope. Our contracts are long-term (20-30 years, IIRC).
Blame *my* government (regulators, in fact) for negotiating a
crappy *LONG*-term contract with HQ. We even tried suing HQ to
get out of the contract when they couldn't supply (ice damage).
Not that I thought the suit was justified, rather an excuse.
Nope. Not a tariff issue. A crappy bit of negotiation a decade
or two ago (more than a decade because it was before I moved
here).
D
Dorothy Bradbury
Looks like someone hired the mobile phone monthly package marketeers,
ie, bury people in so many layers they simply just buy what they're given
The P-M broke the clock-speed-as-criteria marketing, so they do need
something to change that - perhaps beyond using just the Centrino Brand.
P-M is in the nicest architecture they have.
ie, bury people in so many layers they simply just buy what they're given
The P-M broke the clock-speed-as-criteria marketing, so they do need
something to change that - perhaps beyond using just the Centrino Brand.
P-M is in the nicest architecture they have.
R
Robert Myers
Looks like someone hired the mobile phone monthly package marketeers,
ie, bury people in so many layers they simply just buy what they're given
The P-M broke the clock-speed-as-criteria marketing, so they do need
something to change that - perhaps beyond using just the Centrino Brand.
P-M is in the nicest architecture they have.
And, if you believe Nick Maclaren, who is a regular poster to
comp.arch, the future of Intel x86. He may mean to imply that it is
the future of Intel CPU's, period.
If either interpretation is correct, Intel has a bigger problem than
merely obfuscating the fact that megahertz was never meaningful as a
raw measure of performance. They will also have to obfuscate the fact
that they are abandoning the NetBurst architecture. The public is
mostly unaware of the Itanium saga, and I'm much more ready to believe
that they are ready to dump NetBurst than I am that the decision point
for IA-64 has been reached.
RM
D
Dorothy Bradbury
I think it is inevitable...
o Heat output is supposed to stall to 8-5%/yr from "here" (~103-120W)
o Yet P-M is a miserly 22W & Opteron is now considerably lower
P-M proves small drop in performance = disproportionate drop in watts.
Probably a portfolio of chips, re thermal output.
That said, higher thermal outputs are an important driver for change:
o Hot 1U = new 50x40x38mm counter-rotating dual-motor fans
o Hot 1U = new 10,000rpm SCSI 2.5" disks re 4-in-1x3.5"
o Hot = new form-factors re BTX & probably 1U varient of it
o Hot = oh, you'll have to junk your blade/blade-cage & buy new
o Hot = new socket & m/b to cope with the current needed
Will be amusing if Intel forces the world to change due to heat, only to
then deliver a set of lower power options of substitutable performance.
Consultancy is hours billed, taxi-drivers are A-B-C-D v A-B, and
IT industry is closer to the taxi-driver re business strategy attain & cost.
o Heat output is supposed to stall to 8-5%/yr from "here" (~103-120W)
o Yet P-M is a miserly 22W & Opteron is now considerably lower
P-M proves small drop in performance = disproportionate drop in watts.
Probably a portfolio of chips, re thermal output.
That said, higher thermal outputs are an important driver for change:
o Hot 1U = new 50x40x38mm counter-rotating dual-motor fans
o Hot 1U = new 10,000rpm SCSI 2.5" disks re 4-in-1x3.5"
o Hot = new form-factors re BTX & probably 1U varient of it
o Hot = oh, you'll have to junk your blade/blade-cage & buy new
o Hot = new socket & m/b to cope with the current needed
Will be amusing if Intel forces the world to change due to heat, only to
then deliver a set of lower power options of substitutable performance.
Consultancy is hours billed, taxi-drivers are A-B-C-D v A-B, and
IT industry is closer to the taxi-driver re business strategy attain & cost.
J
Jan Panteltje
Ahum, an interesting remark, now, in the Netherlands, with force 12Robert Myers said:
strom, my sat dish blown out of alignment, and so much noise from teh storm
it almost scares one.
Never before have I seen storms like this in this time of the year.
120km/h winds, but decreasing now.
The fact is: You CANNOT hear the PC with this wind
Open a window and you need no fans...
JP
G
greg mushial
Have to agree - around here with the exception of one P4, all the
desktop machines have evolved to 1ghz C3's - which has cut down on the
office background noise level and likewise the A/C load... and when
the P3s were replaced, no mention was made that the C3 were going
to be 1.5x slower... and after 8 months no-one seems to have noticed
(but the quieter office has been clearly noticed and appreciated).
greg