Searching \ for '[EE]:: Laptop underclocking' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/timers.htm?key=clock
Search entire site for: ': Laptop underclocking'.

Exact match. Not showing close matches.
PICList Thread
'[EE]:: Laptop underclocking'
2007\11\27@214617 by Apptech

face
flavicon
face
Summary:

How hard to UNDERclock an Intel laptop processor?

____________________

A new laptop is in the offing.
Minimum spec is "IBM compatible", dual core processor, 1 GB
RAM, 100 GB HDD.
PLUS - Longest possible battery life.
AND    Small and light is good

More or less as appropriate always good.
Target price not over about $US1200.

Uses will (of course) be varied but an  essential role is
the capability to load photos from a Flash card while
displaying slide shows or image manipulating. This is the
main reason that dual-core is essential - it's not the extra
speed (which helps) but the fact that, at least for the
software I use, partitioning the two tasks one per processor
makes a vast difference wrt running them on even the fastest
of single processor systems. And, I don't wish to change my
software.

Any number of 'notebooks' meet the spec until light and
long-battery-life are added.

The 12.1" VAIO family do this superbly but get very
expensive as the spec gets acceptable.
Down to about 1.2 kg and 6 to 7 hours battery life in "slow"
mode.

The 12.1" HP (possibly some being made for them by ?ACER?)
are excellent on weight but only OK on battery life.

It seems to me that if I could massively underclock and
undervoltage the CPU and run the backlight not too too
bright that I should be able to obtain similar to VAIO
results. Obviously speed will fall with underclocking but as
long as clock speed is easily altered this is not an issue,
and even if it took a reboot it would still probably be
acceptable.

Which leads to the question:

AMD processors are, I read, all capable of being
underclocked and under voltaged as they all use the same
control, IC with this capability and there are utilities
which perform this task well.

BUT Intel processors are reportedly inconsistent in their
ability to be underclocked.

Can people comment on the issues involved ?:
what it takes to be able to underclock and under voltage an
Intel CPU,
is it always possible,
are there gotchas,
have people done it ,
did it work OK,
and reliably,
should I buy an AMD based laptop if this is a core (no pun
quite intended) issue.
?????????????



           Russell



2007\11\27@223839 by Xiaofan Chen

face picon face
On 11/28/07, Apptech <spam_OUTapptechTakeThisOuTspamparadise.net.nz> wrote:
> Summary:
>
> How hard to UNDERclock an Intel laptop processor?
>
> ____________________
>
> A new laptop is in the offing.
> Minimum spec is "IBM compatible", dual core processor, 1 GB
> RAM, 100 GB HDD.
> PLUS - Longest possible battery life.
> AND    Small and light is good
>
> More or less as appropriate always good.
> Target price not over about $US1200.

Small and long battery life (>10hours) yet using a dual
core processor --> not cheap.

http://asia.cnet.com/reviews/notebooks/0,39050495,39133705,00.htm

Xiaofan

2007\11\27@225921 by M. Adam Davis

face picon face
Ever since the Pentium (maybe earlier?  But at least everything after
the Pentium..), Intel has used techniques that have a lower clock
speed limit, ie, they are decidedly non-static.

There's a range of usable frequencies, but they aren't published, and
they vary from processor to processor.

Further, Intel has used various means to prevent
overclocking/underclocking on its lower-end processors.  There are
some that don't have these limitations, but they are the high end
expensive models.

Here's one article on underclocking an older AMD processor for desktop:
http://pcstats.com/articleview.cfm?articleID=1798

For the most part everything is programmable on higher end
motherboards, but I expect that to save cost, weight, money, power
laptops aren't nearly as flexible.  You may have some difficulty
adjusting the voltage at all, and may only have a few steps for the
processor speed.  You may not be able to modify much else without
getting a higher end laptop (portable desktop, alienware, etc)

Although I'm curious why you have requirements suggesting high power
(dual core, etc) but still want the benefits of long battery life and
low cost.  Performance, long battery life, low cost - pick two.

Have you considered a laptop with two battery bays?  Or a separate battery?

-Adam

On Nov 27, 2007 9:46 PM, Apptech <.....apptechKILLspamspam@spam@paradise.net.nz> wrote:
{Quote hidden}

> -

2007\11\28@005143 by Robert Rolf

picon face
Apptech wrote:

> How hard to UNDERclock an Intel laptop processor?

VERY. They use various 'dynamic' cells in the CPU core to save on
transistor count. e.g. storing register values in gate capacitance.
But come flavours have good power management (shutdown unused parts
of the core).

For lowest intrinsic power, look to AMD. 60 Watts (AMD Athlon) vs 110 W for
SAME MIPS in Intel (desktop). Just look at the heat sinks to tell who is wasting
more power <G>.

This link says that AMD Athlons can clock down to 50Mhz. (page 42 in pdf. pg 30 in document).
http://www.amd.com/us-en/assets/content_type/white_papers_and_tech_docs/24685.pdf

Have you looked at setting a very short 'suspend' or 'hibernate' time,
and having the unit wake up as needed?

Set a short 'on time' for the backlight and hit the 'ctrl' key whenever you
are actually looking at the display?

Use one of the new 'hybrid' flash + hard drive hard drives (Seagate) so that you
don't need to spin the disk at all?

www.seagate.com/docs/pdf/marketing/ds_momentus_5400_psd.pdf
"Power efficiency—Laptop battery life is extended by the ability of the
Seagate® Momentus® 5400 PSD drive to power off the spin motor
and operate out of flash memory."


Most AMD based laptops come with a utility that lets you trade off power/speed/
brightness for battery life.

Robert

2007\11\28@024847 by Apptech

face
flavicon
face
> Although I'm curious why you have requirements suggesting
> high power
> (dual core, etc) but still want the benefits of long
> battery life and
> low cost.  Performance, long battery life, low cost - pick
> two.
>
> Have you considered a laptop with two battery bays?  Or a
> separate battery?

A camera (or two or three) is often enough attached
permanently to my right arm (9000+ images in 2 weeks in
China*) and a laptop that deals well with the demands of
image handling on the road is essential. I also photograph
occasional stage events, birthdays weddings etc "for fun"
more than for profit.

As noted, experience has shown that TWO cores handle flash
card reading into PC plus simultaneous image manipulation
far far far better than one core. Even a very slow dual core
is far more usable than a fast single core. I haven't
explored the "why" of this too far, but I also don't want to
have to craft my solutions around the limitations of a PC
more than I need to. The two core arrangement definitely
allows separation of otherwise system hogging tasks.

BUT the act of image display need not be overly processor
intensive. If a slide show changes images every say 4
seconds then I am happy for it to take most of that fetching
the image as long as the change to the next photo is near
instantaneous. Also, on a photo shoot, carrying a running or
hibernating laptop "has been known to happen" and the
ability to run for long periods with access on demand is
desired. Hibernation helps. The ability to edit, sort etc
while mobile whether on holiday over N days or eg twixt
church and reception is also valued.

And then, the ability to work on my "actual work", run
spreadsheet, compiler , word processor, all the usual boring
stuff... for extended periods requires often enough more
endurance than utter power.

As portability is a major plus then light weight is nice.
VAIO starts at around 1 kg and some even less. Alas that may
end up with no DVD or non standard sub 2.5" HDDs. The
IBM/Lenovo X61S 10+ hours operation) is very nice in most
respects but lacks DVD. I accept that that's a compromise
that suits some, but not me. Two batteries is fine, as long
as it's possible in the price range ;-).

An underclocked PC that can jump from sprint to crawl and
back nearly instantly seems like a good answer. AMD seems to
be the way to achieve it.



       Russell

* 9000 photos may seem to most to be an insane number of
photos to take in 2 weeks. I agree. But I wouldn't have
wanted to have taken fewer. 500 km train ride at up to 200
kph in parts Qingdao - Beijing. Russell stands at door at
carriage end for most of trip - fields, buildings, bridges,
people, vehicles, agriculture, animals, machinery,
landscapes, ... far more. Quality varies widely, but I'll
remember more and more of that 500 km ride over time as my
screen saver hauls back images after 1 minute of
non-keyboard time.

Beijing - 2 days on a bicycle. temples, parks, street, back
streets, houses, landscape, food, food, food, bikes, bikes,
motorbikes, food, bikes, people, power lines, roadworks,
hydrants, pumps, Olympic hoardings, traffic, churches,
soldiers, police, people ...  ... . Ask me about Beijing -
go on, just ask :-).

Hong Kong - light show where the whole city building
lighting is computer controlled by one source, lit up boats
of every imaginable type (except no warships)  ply the
harbour, ...  - how many photos can YOU take in 20 minutes
;-). Boats, bridges, people, food (we've had this list),
countrside tour - villages ... .

You get the idea.
Over time the trip is reinforced increasingly.

And then, it was also useful in the factory where I was
visiting :-).






2007\11\28@033026 by Artem Zezyulinskiy / SEDATELEC

flavicon
face
Hi, I use the utility RMClock from
http://cpu.rightmark.org/products/rmclock.shtml. For my Athlon Mobile is
work fine. But it is compatible with Intell processors too.

*RightMark CPU Clock Utility (RMClock)* is a small GUI application
designed for real-time CPU frequency, throttling and load level
monitoring and on-the-fly adjustment of the CPU performance level on
supported CPU models via processor's power management model-specific
registers (MSRs). In automatic management mode it continuously monitors
the CPU usage level and dynamically adjusts the CPU frequency, throttle
and/or voltage level as needed, realizing the "Performance on Demand"
concept.

Current release of RMClock utility implements the CPU frequency, CPU
load and throttling level determination on the following processor models:

   * AMD K7 (Athlon/XP/MP, Duron, Sempron) and K8 (Athlon 64/FX/X2,
     Opteron, Dual-Core Opteron, Sempron, Turion 64/X2) family processors.
   * Intel Pentium II/Celeron, Pentium III/Celeron, Pentium M/Celeron
     M, Pentium 4/Celeron (Northwood and Prescott cores), Pentium 4
     Extreme Edition (Gallatin and Prescott cores), Xeon (Prestonia,
     Nocona, Cranford, Irwindale, Potomac, Paxville and Dempsey cores),
     Pentium D and Pentium Extreme Edition (Smithfield, Presler and
     Cedar Mill cores), Core Solo/Celeron M/Core Duo (Yonah core) and
     Core 2 Duo/Core 2 Quad/Core 2 Extreme/Xeon (Conroe/Allendale,
     Merom, Woodcrest, Kentsfield/Clovertown cores).



Apptech a écrit :
{Quote hidden}

--
------------------------------------------------------------------------
Logo Sedatelec <http://www.sedatelec.com>
       
Artem ZEZYULINSKIY <artemzezspamKILLspamsedatelec.com>
SEDATELEC <http://www.sedatelec.fr/>, Chemin des Mûriers - Irigny 69540
- FRANCE
Tel : +33 [0]4 72 66 33 26






2007\11\28@042155 by Tamas Rudnai

face picon face
IBM and Dell provides tools for this on their standard installation already.
I am pretty sure all others do the same - even with Linux you can set a
different throttle settings for many power statuses (mains or battery
powered or the if the battery is flat etc). On my IBM T42p it makes around
1hr difference if I use the lowest throttle settings or the high performance
mode.

Tamas


On Nov 28, 2007 8:28 AM, Artem Zezyulinskiy / SEDATELEC <
.....artemzezKILLspamspam.....sedatelec.com> wrote:

{Quote hidden}

>

2007\11\28@044555 by Alan B. Pearce

face picon face
>Have you considered a laptop with two battery bays?  Or a separate battery?

This is probably the way to work. My Dell can replace the CD drive with a
removable battery or another hard drive.

The CD drive is USB, and I believe the hard drive caddy also does a USB
interface to a standard 2.5" drive, but there must be a separate set of
contacts to use a second battery in there. OK it won't have the capacity of
the primary battery, but anything that gets you extra life has to be good.

The second battery slot appears on the battery management panel that comes
up to show charge status, even with no battery in there.

2007\11\28@045944 by Alan B. Pearce

face picon face
>Hong Kong -
> ...  - how many photos can YOU take in 20 minutes
>;-). Boats, bridges, people, food (we've had this list),
>countrside tour - villages ... .

Trip up to the Peaks on the railway, fountain playing at the top ... ah the
memories of 10 years ago ...

2007\11\28@053820 by Apptech

face
flavicon
face
> >Hong Kong -
>> ...  - how many photos can YOU take in 20 minutes
>>;-). Boats, bridges, people, food (we've had this list),
>>countrside tour - villages ... .

> Trip up to the Peaks on the railway, fountain playing at
> the top ... ah the
> memories of 10 years ago ...
>
> --

2007\11\28@084205 by M. Adam Davis

face picon face
Short term gains with long term costs.  It's a well played tune all
over the world.

-Adam

On 11/28/07, Apptech <apptechspamspam_OUTparadise.net.nz> wrote:
> I liked China but I'm sad about what they have done and are
> continuing to do to it.


--
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
- - - - - - - - - - - - - - - - - - - - - - - - -
Moving in southeast Michigan? Buy my house: http://ubasics.com/house/

Interested in electronics? Check out the projects at http://ubasics.com

Building your own house? Check out http://ubasics.com/home/

2007\11\28@163428 by Dwayne Reid

flavicon
face
At 07:46 PM 11/27/2007, Apptech wrote:
>Summary:
>
>How hard to UNDERclock an Intel laptop processor?

Easy.

RM Clock works well for me (intel centrino)

<http://cpu.rightmark.org/products/rmclock.shtml>

dwayne



--
Dwayne Reid   <@spam@dwaynerKILLspamspamplanet.eon.net>
Trinity Electronics Systems Ltd    Edmonton, AB, CANADA
(780) 489-3199 voice          (780) 487-6397 fax
http://www.trinity-electronics.com
Custom Electronics Design and Manufacturing


'[EE]:: Laptop underclocking'
2007\12\02@090121 by Morgan Olsson
flavicon
face
Den 2007-11-28 08:07:54 skrev Apptech <KILLspamapptechKILLspamspamparadise.net.nz>:

> As noted, experience has shown that TWO cores handle flash
> card reading into PC plus simultaneous image manipulation
> far far far better than one core. Even a very slow dual core
> is far more usable than a fast single core.

Sounds like system is not operating correctly.
Have you tried changing OS and applicaiton?


--
Morgan Olsson

2007\12\02@163621 by Apptech

face
flavicon
face
>> As noted, experience has shown that TWO cores handle
>> flash
>> card reading into PC plus simultaneous image manipulation
>> far far far better than one core. Even a very slow dual
>> core
>> is far more usable than a fast single core.

> Sounds like system is not operating correctly.
> Have you tried changing OS and applicaiton?

Ah. True religion. And not even a smiley face :-)

I thought about carrying an IBM system 360, but they don't
make cords long enough, and the vacuum tap readers are a bit
heavy.

Seriously though, the dual core system I described works
well, and it makes sense that it does, and many products
nowadays have dual cores, so I see little sense in changing
the things that do work in order to pursue an idealistic
aim. The dual cores allow true multi-processing and it's
possible that some mutual attributes of the applications
that I use plus Windows task swapping functionality interact
in some undesired way. This may be fixed by changing to eg
Linux or System 360 or by using Photoshop Reader as a flash
card reader and Pro Paint 67 as an image manipulator, and it
may not, but why bother when a system which is cheaply
available and has definite technical merit can be had at
essentially no price penalty simply by choosing to have it?



       Russell

2007\12\02@172249 by William \Chops\ Westfield

face picon face

On Dec 2, 2007, at 1:36 PM, Apptech wrote:

> the dual core system I described works
> well, and it makes sense that it does

I dunno.  It doesn't make sense to me that a single core system would  
be that much slower than a dual core at importing pictures over USB.  
USB isn't really that fast in the first place (and neither is the  
flash?) and I thought that a lot of the overhead of USB was handled  
in chipset hardware anyway.

Have you tried to track down where the bottleneck is, by (for  
example) doing copies from the flash using traditional file copy  
utilities?  I'd be less surprised that a typical "photo" app decided  
to give priority to user interaction over photo import and suffered  
for it in your situation...

BillW

2007\12\02@175309 by Shawn Tan

flavicon
face
On Sunday 02 December 2007 22:22:42 William "Chops" Westfield wrote:
> On Dec 2, 2007, at 1:36 PM, Apptech wrote:
> > the dual core system I described works
> > well, and it makes sense that it does
>
> I dunno.  It doesn't make sense to me that a single core system would
> be that much slower than a dual core at importing pictures over USB.
> USB isn't really that fast in the first place (and neither is the
> flash?) and I thought that a lot of the overhead of USB was handled
> in chipset hardware anyway.

I'm not familiar with USB specifically, but it's usually true that multi-core
systems will have higher throughput than single core system. This is
particularly true when dealing with slow I/O. And I believe that I have read
somewhere than UHCI shifts the load of USB processing onto software, not the
chipset. So, it's quite possible that a single core system will be slower, if
there are any blocking I/O processes.

Cheers.

--
with metta,
Shawn Tan

Aeste Works (M) Sdn Bhd - Engineering Elegance
http://www.aeste.net

2007\12\02@185813 by Apptech

face
flavicon
face
>> the dual core system I described works
>> well, and it makes sense that it does

> I dunno.  It doesn't make sense to me that a single core
> system would
> be that much slower than a dual core at importing pictures
> over USB.

The thing that makes sense is that a dual core processor
offers the prospect of genuine multitasking as opposed to
Windows main frame time sharing dinosaur model. I have no
idea of how good a use they make of the true multitaking
that multiple cpus offer but, it seems, the answer is "good
enough in this case."

One can set Windows' "affinity" parameters on a multi core
system to direct given tasks to given processors, but it
does fine on 2 cores as is without doing this in my case.

> USB isn't really that fast in the first place (and neither
> is the
> flash?)

USB is good for about 400 mbps in theory. Much less in
practical reality.
The Flash that I usually use does an honest 150X+ MEASURED =
19.5 mbps =~ 2.5 mBps so no, it's not that fast, and not
overly taxing USB2 (or shouldn't be) at about 5% of its
theoretical maximum bit rate.

> and I thought that a lot of the overhead of USB was
> handled
> in chipset hardware anyway.

You can't argue with success :-) - especially so if it uses
what you have the way you like it and costs essentially no
more to do.

> Have you tried to track down where the bottleneck is, by
> (for
> example) doing copies from the flash using traditional
> file copy
> utilities?

I read the flash in an XP DOS box = a genuine Windows app
BUT I have tried it using various photo file loaders and got
similar results. Using a (pseudo) DOS box allows me to
manage various things on the fly that standard software may
not. For example, I copy and do not "move" files and modify
the attribute bits for the source as I do so the card
maintains a backup copy that is not redownloaded on
subsequent reads. This allows me, when venue and inclination
commend it,  to time multiplex a pocket full of cards so
photos are spread across cards and a card failure only loses
a time interleaved set of photos. Better to lose some of a
range of photos than all of eg a wedding service proper.
[Those who have never had a memory card failure and/or who
are confident that they never will have one may not feel the
need either for a backup or card interleaving.]. If I had a
camera that did writes to 2 cards at once (as some do) there
are occasions where I would use the feature. I don't have
many card failures but I have had some.

My experience is that when the camera gets noticeably hot in
use and you are taking so many photos that the flash cycle
time is a major annoyance and the batteries from the
5600HS(D) flash are far far too hot to handle when you dump
them from the flash as you run to the next shooting point
that card reliability starts to fall. Add higher ambient
temperatures and it's probably worse. Doesn't usually seem
to be a problem for the occasional scenic photo.

> I'd be less surprised that a typical "photo" app decided
> to give priority to user interaction over photo import and
> suffered
> for it in your situation...

Yes. But if the dual core fixes it and costs essentially no
more to buy these days, I'm happy if it lets me use the
software that does most of what I want in its specialist
area.



       Russell

2007\12\02@205138 by Xiaofan Chen

face picon face
On 12/3/07, Shawn Tan <RemoveMEshawn.tanTakeThisOuTspamaeste.net> wrote:

> I'm not familiar with USB specifically, but it's usually true that multi-core
> systems will have higher throughput than single core system. This is
> particularly true when dealing with slow I/O. And I believe that I have read
> somewhere than UHCI shifts the load of USB processing onto software, not the
> chipset. So, it's quite possible that a single core system will be slower, if
> there are any blocking I/O processes.
>

http://en.wikipedia.org/wiki/USB
UHCI/OHCI are more for the full-speed and low speed device. EHCI is
the one which supports High Speed. Hopefully the modern card reader
and flash disk uses high speed rather than full speed.

But I have a bit of problem understanding how dual core system
could be that faster when accessing a slow I/O like USB flash disk.

I do notice that copying many small files to a USB flash disk is slower.
Zip it and then copy the zip file and it is much faster. It is the
same for copying large amount of files over the Internet. Zip it
and then copy will be much faster.

Xiaofan

2007\12\02@232700 by William \Chops\ Westfield

face picon face

On Dec 2, 2007, at 5:51 PM, Xiaofan Chen wrote:

> I do notice that copying many small files to a USB flash disk is  
> slower.
> Zip it and then copy the zip file and it is much faster.

Not an option in this case.  Russell is copying the photos FROM the  
flash
card (where the camera put them), using a USB card reader.

BillW

2007\12\03@001137 by Apptech

face
flavicon
face
> But I have a bit of problem understanding how dual core
> system
> could be that faster when accessing a slow I/O like USB
> flash disk.

Regardless of why, the difference between dual core and
single core with 2 applications running is very substantial.
A graphics display task that is displaying a new image every
5 seconds or so, with files typically in the 2 to 3 MB range
slows to a crawl. Attempts to copy files using this program
similarly  slows. Also, and I'm not sure of this, AFAIR
inter PC LAN transfers also slow down. (I sometimes use 2
laptops for this with one for download and initial
manipulation and an ethernet cable link to another for
viewing of photos by guests or whoever.

> I do notice that copying many small files to a USB flash
> disk is slower.
> Zip it and then copy the zip file and it is much faster.
> It is the
> same for copying large amount of files over the Internet.
> Zip it
> and then copy will be much faster.

I'm sure there are many workarounds. But, as I said at the
start, I don't want to find out how to do things
differently - rather, I want to use a tool which allows me
to continue to use the solutions that work well for me. This
is a good model of life in general. One can get any number
of people telling you that you should or could do things
differently in order to fit in with the limitations of the
solutions they wish to offer, when what people would most
like are solutions that are solutions. In this particular
case, ANY dual core system that I have tried (and I've tried
a small number of both Intel and AMD offerings, included the
slowest that AMD had to offer at the time) and each has
proven to be a solution. That's what I'll be buying.
Anything that makes a dual core system work better for
little or no more $ is a bonus.


       Russell


2007\12\03@003708 by Jake Anderson

flavicon
face
Have you tried using something else to do the initial copy of the data?
Perhaps just a batch file or something?
still run it through your magic software but let it pull it from a hard
drive or something rather than flash.
(although in theory flash should be faster)

Core duo all the way in my opinion anyway. (Core quad on the desktop ;->)

Apptech wrote:
{Quote hidden}

2007\12\03@021725 by Apptech

face
flavicon
face
> Have you tried using something else to do the initial copy
> of the data?
> Perhaps just a batch file or something?
> still run it through your magic software but let it pull
> it from a hard
> drive or something rather than flash.
> (although in theory flash should be faster)

The magic software isn't magic at all - just low level and
useful.
But, whichever way I have tried to access the flash, has
produced the same slow down of the system.

The 'magic software' that I have settled on is in fact a
batchfile calling XXCOPY.
It's no slower than alternatives and does what I want.

I can run a batchfile "LOOK" which looks for a flash card in
any of the possible "drives" and tells me how many jpgs are
on the card and how many haven't been downloaded. usually
when 'on the run' I just run batchfile "SUCK" which uses
XXCOPY to download any jpgs which don't have the copy bit
set and then set it ( or is that clear and then clear it ...
? ;-) ).

Very workable and as fast as I can make it, as long as the
system has dual cores.

At the end of the session I run only slightly more magic
programs which correct Microsoft's data and time file
settings to the proper ones as per EXIF* and then add a 5
digit suffic to the file name which uniquely identifies them
to within 1 minute in 10 years. The 4 digit camera file name
then allows me to take up to 10000 photos per minute [ :-) ]
before I lose the uniqueness of the naming.

* For reasons known better to them than I, when you use most
systems to copy camera files from FLASH to disk, the file
time and maybe the date is usually corrupted to a value
which relates to the original but is wrong. Microsoft have
their own hand holding free program available to handle
this. The problem arises AFAIR due to date format issues in
the camera and XP O/S. I don't care why Microsoft do stupid
things, although I'd rather they didn't, if I can fix them,
although I'd prfere they got it right initially. I use the
brilliant JHEAD utility to set date and time to the EXIF
values

           Jhead -ft *.jpg

Doing this on files which you say are jpg but it says are
not can be fatal to their health, alas.
Date/Time prefix adding a simple (blush) BASIC program.

I'd be interested in other people's experiences of
displaying slide shows at about 0.5 MB second/ 5 seconds per
file and simultaneously loading a flash card to disk using
software of their choice.

My preferred slideshow and general photo playing system in
this context is Irfanview because it works well and is far
less handholding than most other software. I dislike
siftware which has its own idea of how things should be done
and stops you doing things quickly and directly. Anything
which doesn't allow you to mix mouse and keyboard commands
totally flexibly is a non starter where I have any choice.
Keyboard is often quicker.


           Russell

2007\12\03@030446 by Xiaofan Chen

face picon face
On 12/3/07, William Chops Westfield <spamBeGonewestfwspamBeGonespammac.com> wrote:
> > I do notice that copying many small files to a USB flash disk is
> > slower. Zip it and then copy the zip file and it is much faster.
>
> Not an option in this case.  Russell is copying the photos FROM the
> flash card (where the camera put them), using a USB card reader.

You are right. I think the better solution is to get a faster flash card
(the speed can vary quite a bit) and a faster USB card reader
(again the speed can vary a lot). Often the build-in camera
USB interface is quite slow (at least on my old Nikon 3100).

As for the host side, I actually feel that Windows is a bit
slower than Linux when copying from Flash Card. But I
have not really measured that so this may well
be a perception. Last time, I also read that hyperthreading
will make people *feel* the system is more responsive
but in reality do not offer real speed boost.


Xiaofan

2007\12\03@041732 by Apptech

face
flavicon
face
>> > I do notice that copying many small files to a USB
>> > flash disk is
>> > slower. Zip it and then copy the zip file and it is
>> > much faster.

>> Not an option in this case.  Russell is copying the
>> photos FROM the
>> flash card (where the camera put them), using a USB card
>> reader.

> You are right. I think the better solution is to get a
> faster flash card

No. Speed is OK by itself.

> ... and a faster USB card reader

No. Speed is OK by itself.

...
> Often the build-in camera USB interface is quite slow

Not relevant as I read the flash cards directly.

I stated the original requirement constraints initially.
I'll summarise and add a bit more.

I read Flash cards into a PC via a Cardbus CF reader.
I display photos on the same PC.
Either task works well in isolation.
With ANY dual core CPU the tasks work as well simultaneously
as when run in isolation.
With ANY single core cpu the display task slows very
substantially and the flash reading may also suffer.

The Flash cards I prefer are true measured 150X read capable
in a desktop PC and over 120X write in my 7D camera.
(Transcend 120X CF 2Gb or 4GB)
As opposed to many which claim faster but operate slower.
Camera write is what I most care about. PC read is nice but
secondary, within reason.

The Transcend cards are the best performance per $ cards I
have found and also better performers in absolute terms than
many others which claim more. They exceed 120X write in my
camera. I have a number of other cards brands (Sandisk,
Kingston, ...). Some have racing stripes and words like
Ultra or Ultra 2 written on them - but none work as fast as
my Transcend 120X cards.

I don't NEED to change anything.
I don't WANT to change anything.
A faster flash card would work faster but that is not the
issue. The issue is that the tasks slow down when run
simultaneously with a single core cpu but not with ANY dual
core cpu. I do not wish to change the O/S, the flash card,
the reader, or any other software. It works just fine as is
with dual cores and I am happy. The original query was, as
the subject line suggests, the ability to underclock a given
CPU when cpu speed is less important than power consumption.

> ... Last time, I also read that hyperthreading
> will make people *feel* the system is more responsive
> but in reality do not offer real speed boost.

But, dual coring, while it does not make the tasks run
fasterm stops them from running slower :-).




       Russell

2007\12\03@074324 by Jake Anderson

flavicon
face
Have you tried it on a Hyper threaded machine?
If its blocking IO thats the issue then HT would be a decent solution.

But still, from new core 2 is the way to travel.

Apptech wrote:
{Quote hidden}

2007\12\03@112007 by Morgan Olsson

flavicon
face
Den 2007-12-02 22:36:21 skrev Apptech <TakeThisOuTapptechEraseMEspamspam_OUTparadise.net.nz>:

>> Sounds like system is not operating correctly.
>> Have you tried changing OS and applicaiton?
>
> Ah. True religion. And not even a smiley face :-)

Sorry I am too tired at it for a smile...

My old 1,4GHz Thinkpad beats our new dual core 2GHz (?) AMD ASUS laptop in speed, plus the AMD is getting severely hot that we had to put a large 6mm alumina plate on the table so it could work all day.  You see we run severely heavy things like MPLAB and C18...  ;)
This is both when booting them with the WinXP thet they came with, plus updates.
The AMD is also having I/O problems with USB plugging/unplugging, it even blue-screened after having plugged in Real-ICE and it autosupidly tried to install a mouse driver... ...and PC-card and wireless we gave up on.
Very clearly AMD/ASUS/Microsoft failed miserably at together getting the ASUS system working.

You *might* find the Linux camp have solved it better.
A test would be to boot a Mandriva One 2008 or similar system-on-CD.

As for photo handling, retouch etc i found better tools on Linux than on Win, but i am no pro at that.

--
Morgan Olsson

2007\12\04@101234 by MatthewMucker

flavicon
face
This could be a hardware issue. If the processor gets too hot, the hardware
backs down the CPU clock.


{Quote hidden}

> --

More... (looser matching)
- Last day of these posts
- In 2007 , 2008 only
- Today
- New search...