Anyone have a Toshiba A65-S1762 from Costco?

Archived from groups: comp.sys.laptops (More info?)

"Andrew" <usenetMYSHOES@bizaveMYSHOES.com> wrote in message
news:laaclahj52943648209@bizaveMYSHOES.com...
> Does anyone have or know about the Toshiba A65-S1762 laptop sold at
> Costco? As of today, it's going for $1399.99.
>
> As far as I can tell, this is a much beefier version of the A65-S126,
> which has a 2.8GHZ Celeron and only 256MB of RAM (and shared Video
> RAM). The Costco version has a 3.2GHZ Pentium 4 with a 1GB Cache and
> Hyperthreading. I'm wondering if this is a Prescott P4 CPU. It also
> has
> 512MB of RAM. a 533MHZ FSB, a 60GB hard drive, and a built-in DVD
> burner.
>
> I'm considering recommending this machine to my brother. He basically
> wants a desktop replacement-type laptop, something he can move around
> once in a while, but he doesn't care much about weight or battery
> life. His laptop is a Dell which he likes (despite the fact that it
> has given him all kinds of problems). He likes the Dell warranty
> service option, but getting a laptop with Costco's 6 month return
> policy might make him forego a much costlier Dell.
>
> Concerns about the A65: it seems cheaply made. I'm guessing it's
> made by Compal (19V power supply). The keyboard seems nice, but I
> don't like how the touchpad buttons feel. Also, if the CPU is indeed
> a Prescott, I worry about how hot it gets. I have a Prescott in my
> desktop and it idles at 47C under no CPU load!!!
>
> Any comments about any version of the A65 would be appreciated.
>
> Andrew
> --
> ----> Portland, Oregon, USA <----
> *******************************************************************
> ----> http://www.bizave.com <---- Photo Albums and Portland Info
> ----> To Email me remove "MYSHOES" from email address
> *******************************************************************
>

Pretty sure it is the mobile P4. The desktop P4 would be a priori too
hot for a laptop unless there has been a substantial improvement in heat
pipe cooling in the last few days. I have a mobile P4 2.6Ghz 845M
chipset that requires attention to the cooling. I have made some
modifications to the fan ducting and the heap pipe cooling flow that
lets the computer run about 120F/49C during normal use. CPU throttling
keeps it under load to 150F/65C. I very much like the capability of the
laptop, but the cooling performance is less than desirable and is no
doubt affecting longevity.

Anyway, I would avoid any P4 laptop due to potential heat problems, even
at the very attractive price for the Toshiba. The new Mobile Pentium M
laptops (Centrino) are every bit as capable as the P4s. The price
premium for the MPMs is money well spent for cooler operation and
consequent reliability and longevity, not even mentioning longer battery
life.

19volt power supplies are, IMO, no definitive evidence of a Compal unit
any longer. Simply adding up the power consumption at load during a
battery charge will illustrate that for 100 watts (typical of a mobile
P4) at 6 amps, a 15 volt supply is inadequate. 20 volts is the next
logical increase to meet the high power requirement of the P4 system.

Q

Archived from groups: comp.sys.laptops (More info?)

"Andrew" <usenetMYSHOES@bizaveMYSHOES.com> wrote in message
news:zcasdaclahjhj34662248409@bizaveMYSHOES.com...
> Quaoar <quaoar@tenthplanet.net> wrote:
> : Anyway, I would avoid any P4 laptop due to potential heat problems,
> even
> : at the very attractive price for the Toshiba. The new Mobile
> Pentium M
> : laptops (Centrino) are every bit as capable as the P4s. The price
> : premium for the MPMs is money well spent for cooler operation and
> : consequent reliability and longevity, not even mentioning longer
> battery
> : life.
>
> As I have posted here in the past, I evaluated a Pentium M/Centrino
> laptop when it first came out in early 2003. I have one particular
> application that is CPU-intensive (converting RAW images from my Canon
> Camera), and it flopped miserably. On the 1.3GHZ Pentium M Toshiba I
> tested, image conversion was something like 1.5X slower than on the
> 2GHZ
> Celeron laptop I got instead. And we are talking almost a minute per
> image on my Celeron laptop today. That's one big reason I didn't keep
> my first Centrino laptop.
>
> My brother uses the same Canon stuff so if this is still an issue, I
> doubt he'll want a Centrino laptop. The latest Pentium M has a 2MB
> cache so I should do another evaluation to see how today's Pentium M
> does on RAW conversion before dismissing it. I once convinced someone
> at Fry's to let me install the RAW converter on two demo laptops and
> compare - perhaps I can try again on a current Centrino machine.
>
> As I said, battery life isn't an issue at all for my brother - it's
> basically a portable desktop for him. Performance is more important.
>
> Andrew
> --
> ----> Portland, Oregon, USA <----
> *******************************************************************
> ----> http://www.bizave.com <---- Photo Albums and Portland Info
> ----> To Email me remove "MYSHOES" from email address
> *******************************************************************
>

I also do Canon RAW processing. I have done these on a variety of
computers and even a mobile AMD 1500+ is only incrementally slower than
my 2.6Ghz P4M. I am curious about the video adapters in your trial
Centrino machines - if they were the more or less crippled Intel
adapters with shared video ram. The AMD 1500+ and 1800+ both have S3
video with shared ram. The 1800+ is not noticeably slower at RAW image
conversion. Of all the apps I use, the Canon raw process is the most
annoyingly slow, FWIW. I would seek out one of the new Centrinos, 1.7 -
2.0 Ghz for additional testing. The 1.7Ghz, according to reviews, is
every bit the equivalent of my 2.6Ghz P4M. The P4 heat issue is
significant enough IMO that the Centrino should be given a fair test.

Q

Archived from groups: comp.sys.laptops (More info?)

"Andrew" <usenetMYSHOES@bizaveMYSHOES.com> wrote in message
news:zcasdlazcasdrkrhj92863648409@bizaveMYSHOES.com...
> Quaoar <quaoar@tenthplanet.net> wrote:
> : I also do Canon RAW processing. I have done these on a variety of
> : computers and even a mobile AMD 1500+ is only incrementally slower
> than
> : my 2.6Ghz P4M.
>
> A P4M is a Pentium 4 variant, not a Pentium M, so I don't expect that
> has much to do with my comparisons.
>
> : I am curious about the video adapters in your trial
> : Centrino machines - if they were the more or less crippled Intel
> : adapters with shared video ram. The AMD 1500+ and 1800+ both have
> S3
> : video with shared ram. The 1800+ is not noticeably slower at RAW
> image
> : conversion.
>
> Video memory (shared or otherwise) should have zero to do with RAW
> image conversion. RAW conversion is purely a CPU-intensive task that
> could be done without any graphics card, in theory.
>
> : Of all the apps I use, the Canon raw process is the most
> : annoyingly slow, FWIW. I would seek out one of the new Centrinos,
> 1.7 -
> : 2.0 Ghz for additional testing. The 1.7Ghz, according to reviews,
> is
> : every bit the equivalent of my 2.6Ghz P4M. The P4 heat issue is
> : significant enough IMO that the Centrino should be given a fair
> test.
>
> I like the Centrino technology and truly hope my own next laptop will
> contain it. I will definitely give the latest Centrino machines a
> fair trial. But if Pentium M is still slow at RAW conversion, I doubt
> I'll be buying one soon. Actually I believe Canon should do something
> to improve RAW conversion speed - it is annoyingly slow. For now, I
> am stuck with what they offer.
>
> Andrew
> --
> ----> Portland, Oregon, USA <----
> *******************************************************************
> ----> http://www.bizave.com <---- Photo Albums and Portland Info
> ----> To Email me remove "MYSHOES" from email address
> *******************************************************************
>

Sorry, I was unclear: You were proposing the purchase of a P4 laptop,
probably a P4M. I own a 2.4Ghz P4M laptop. The AMD laptops of putative
far less capability than this P4M have no problem matching or nearly
matching the RAW conversion speed of this P4M. I am suggesting that the
Mobile Pentium M (Centrino, MPM) at 1.3GHz should have been only
slightly less capable than a P4(M) laptop since AMD laptops of lesser
capability are nearly as capable as a P4M. It is apples and oranges to
some extent, but illustrates why I suspect the RAW conversion was
unnaturally slow in your in-store Centrino testing.

The problem with in-store testing on the Centrinos might be that those
are straight out of the box installations with all of the whizbang
startup apps loaded, no defrag, etc., etc. These OEM startup suites
cripple the performance of any laptop as you are no doubt aware. The
user community seems to be verifying that the Mobile Pentium M
(Centrino) is generally the performance equivalent of the P4/P4M
(depending on the chip speed) and the substantially poor performance you
experienced is in my mind suspect.

The heating problems in Pentium 4 laptops is broadly based: it affects
all brands/models, some more than others, and has in some reports
damaged the CPU in a short period of time and in others the laptop shuts
itself down during any intensive function. Frankly, the problem is
pervasive enough that vendors should be considering recalls/refunds for
some models of P4(M) laptops since the solution is often disassembly of
the cooling system to remove materials blocking the heat pipe exchanger
if not mainboard/CPU replacement. This is in my opinion a design
failure, even if it affects only a small percentage of units. In any
other product, it would be an actionable fault (IANAL) in the intended
use of the product. Consumers need to be aware of at least the
possibility of the problem; the solution is to avoid the P4 laptops as
if they were diseased.

Q

Archived from groups: comp.sys.laptops (More info?)

Lewin A.R.W. Edwards <larwe@larwe.com> wrote:
: > Video memory (shared or otherwise) should have zero to do with RAW
: > image conversion. RAW conversion is purely a CPU-intensive task that

: Not entirely true - remember the RAM has finite bandwidth, and
: shared-memory architectures consume some of that bandwidth
: transferring data to the video subsystem.

I think RAW conversion is a CPU-intensive, not memory-intensive
application. In any case, a shared-video system uses that bandwidth
for video when the display is changing. Unless you sit there waiting for
conversion to finish without playing a DVD or a game (I wasn't), I
think having shared RAM would not affect the conversion results at all.

I am most definitely planning to test RAW conversion on a newer
Centrino model to see if the speed is still an issue.

Andrew
--
----> Portland, Oregon, USA <----
*******************************************************************
----> http://www.bizave.com <---- Photo Albums and Portland Info
----> To Email me remove "MYSHOES" from email address
*******************************************************************