PCs to change radically in 2004! (The Enquirer)

"David Maynard" wrote in message
...
John Lewis wrote:
On Fri, 30 Jan 2004 21:29:56 GMT, "Dashi" wrote:


That is a good timely report Wayne,

Dashi

"Wayne Youngman" wrote in message
...

"rstlne"wrote

You know
In Technolgoy terms
Ur about 1/3 of the lifecycle late on that report



Well, my *emphasise* was on the fact that it's all happening rather
quickly
for INTEL, allot of new changes all at once?. . .

Whats your thoughts then?
--
Wayne ][




The reference Intel BTX is a thoroughly flawed thermal design. For
details on BTX - see

http://www.anandtech.com/showdoc.html?i=1876

Why flawed ? It is CPU-thermal centric. The video card
is in the exhaust flow from the CPU. Much, much worse
than ATX.

You're looking at a micro-BTX in that article. And while one could note
the graphics card is 'in the exhaust flow' (to eliminate the need for a
separate fan) when in the vertical configuration that's not any worse than
the ATX design which dumps the CPU heat into the case and then places the
GPU chip on the opposite side of the card where there's practically no
airflow at all, necessitating a separate fan for it regardless.

There's nothing to prevent placing a fan on the BTX GPU card either.


Unfortunately, the next gen GPU chips are likely to have more transistors
than the CPUs

That's really unlikely.


Already do! The Nvidia FX5900 has 130 million transistors.

The P4 2.6C has only 55 million.

Dashi

On Sat, 31 Jan 2004 05:39:17 GMT, "Phil Weldon"
wrote:

Perhaps you should familiarize yourself with the thread before jumping in.
After all, the included text ought not to be long enoung to summarize the
thread, and top posting saves a lot of time for those who actually follow
the discussion.


Ah, but it is common for some regular contributors to this newsgroup
to occasionally wander far off-thread..... Some of these contributors
also top-post. With a short reply in a busy thread it is difficult to
be certain which message the reply is directed at without scrolling to
the bottom, if the reply is top-posted.

Top-posting isn't at all bad, if the person who is top-posting takes
the time to edit off all irrelevant parts of the original message, so
that there is no need to scroll madly to find out exactly what is
being replied to.

John Lewis

--
Phil Weldon, pweldonatmindjumpdotcom

"TomG" wrote
ah, but the sheep have been taken by surprise by a trick or two...



TomG (Data)
you are charged on three counts,

a) Mentioning the word *sheep* in a technical forum.

b) Attempting to *Hi-jack* a technical thread.

and a minor misdeminour of being a *habitual* top-poster

How do you plead?
--
Wayne ][

o IT selling is about A-B-C-D-E rather than A-E
o The more interfaces with the customer, the greater the profit
o The more you can incrementalise change, the more customer interfaces

Some prefer the word interactions over interfaces.

The biggest problem post-1995 was that control over incremental
change was lost - so products rarely stayed in cash-cow very long
before ending up in cash-dog. That forced up the R&D-cost per sale,
forced up the COGS per product-lifecycle and so on.


Intel is pushing BTX for some sound reasons:
o Chips are getting hotter = more cfm (m/sec over CPU) = more noise
---- 100W, 125W, 150W per CPU isn't impossible
---- cooligy technology will allow higher re 150W/cm^3
o More fans are needed = less profit for integrators
---- to make a fan spin faster, for more cfm, costs 0p
---- to replace 1x 36cfm noisy fan with 2x 18cfm quieter fans costs 2 fans
o One fan could be used to cool CPU/RAM/Graphics
---- an inline-blow-thro design allows quieter operation
o Higher wattage CPUs, mean more PSU load, mean more PSU heat
---- so migrate the PSU back to cooling itself mainly (& HD etc)

Personally, BTX isn't that great a LT solution:
o 1 fan to cool CPU/Graphics/RAM has limits
---- today is 89+55+15 = 159W for 1-CPU
---- 2yrs on 150+90+25 = 265W for 1-CPU
------- 300cfm cools 1500W, 265W needs 53cfm
------- even 92mm fans get noisy by 53cfm
o PSUs using oblongised-ATX is still 86mm outlet plate
---- that forces 80mm exhaust fan, ok 120mm inlet fan possible
---- but since BTX uses an adapter to fit ATX, chance was missed

Admittedly, the driver here is fan size = cost & fan qty = cost.
A big OEM selling point of BTX is less fans over ATX designs,
not that Dell haven't shown how it can be done - Apple too altho
that was perhaps with an excessive number of fans quite frankly.

One nice point is the downsizing of case size over ATX, which
in tower and even midi-tower form are somewhat overly big. It
also allows much bigger & heavier heatsink designs with a very
robust motherboard/case reinforcement - and of course gives
Intel another socket design to terminate non-motherboardal u/g.
--
Dorothy Bradbury

"Dorothy Bradbury" wrote
snip
Intel is pushing BTX for some sound reasons:
o Chips are getting hotter = more cfm (m/sec over CPU) = more noise
---- 100W, 125W, 150W per CPU isn't impossible
---- cooligy technology will allow higher re 150W/cm^3
o More fans are needed = less profit for integrators
---- to make a fan spin faster, for more cfm, costs 0p
---- to replace 1x 36cfm noisy fan with 2x 18cfm quieter fans costs 2 fans
o One fan could be used to cool CPU/RAM/Graphics
---- an inline-blow-thro design allows quieter operation
o Higher wattage CPUs, mean more PSU load, mean more PSU heat
---- so migrate the PSU back to cooling itself mainly (& HD etc)



Hi,
wow! do you work for NASA or what? That post you made just nearly *crashed*
me. Very well written and plenty of details for my old
sponge-like-brain-matter. Seeing you speak of these these things makes me
feel like I know nothing! (lol).

He-he I calmed down allot now since this thread started. I got a bit
*ruffled* that I may have to be without a *juicing* PC for a few months, but
I figure this stuff won't settle down for a few months yet, and when it does
I prefer not to beta-test it all (and pay off the R&D!).

Thanks again for taking the time to share your thoughts. . .
--
Wayne ][
new specs coming soon!

Actually BTX could run longer than ATX, looking at some later
thermal design specs - but the graphics card wattage is one issue.

Going to be interesting anyway, improvement on the ATX warehouse.
--
Dorothy Bradbury

Dorothy Bradbury wrote:
o IT selling is about A-B-C-D-E rather than A-E
o The more interfaces with the customer, the greater the profit
o The more you can incrementalise change, the more customer interfaces

Some prefer the word interactions over interfaces.

The biggest problem post-1995 was that control over incremental
change was lost - so products rarely stayed in cash-cow very long
before ending up in cash-dog. That forced up the R&D-cost per sale,
forced up the COGS per product-lifecycle and so on.


Intel is pushing BTX for some sound reasons:
o Chips are getting hotter = more cfm (m/sec over CPU) = more noise
---- 100W, 125W, 150W per CPU isn't impossible
---- cooligy technology will allow higher re 150W/cm^3
o More fans are needed = less profit for integrators
---- to make a fan spin faster, for more cfm, costs 0p
---- to replace 1x 36cfm noisy fan with 2x 18cfm quieter fans costs 2
fans
o One fan could be used to cool CPU/RAM/Graphics
---- an inline-blow-thro design allows quieter operation
o Higher wattage CPUs, mean more PSU load, mean more PSU heat
---- so migrate the PSU back to cooling itself mainly (& HD etc)

Plus the switch to PCI Express and the dump of more legacy I/O.


Personally, BTX isn't that great a LT solution:
o 1 fan to cool CPU/Graphics/RAM has limits
---- today is 89+55+15 = 159W for 1-CPU
---- 2yrs on 150+90+25 = 265W for 1-CPU
------- 300cfm cools 1500W, 265W needs 53cfm
------- even 92mm fans get noisy by 53cfm
o PSUs using oblongised-ATX is still 86mm outlet plate
---- that forces 80mm exhaust fan, ok 120mm inlet fan possible
---- but since BTX uses an adapter to fit ATX, chance was missed

Admittedly, the driver here is fan size = cost & fan qty = cost.
A big OEM selling point of BTX is less fans over ATX designs,
not that Dell haven't shown how it can be done - Apple too altho
that was perhaps with an excessive number of fans quite frankly.

One nice point is the downsizing of case size over ATX, which
in tower and even midi-tower form are somewhat overly big. It
also allows much bigger & heavier heatsink designs with a very
robust motherboard/case reinforcement - and of course gives
Intel another socket design to terminate non-motherboardal u/g.

guilty on all counts, your honor.

next to Jef and Qed, I am one of the best thread hijackers out there...

--

Thomas Geery
Network+ certified

ftp://geerynet.d2g.com
ftp://68.98.180.8 Abit Mirror ----- Cable modem IP
This IP is dynamic so it *could* change!...
over 120,000 FTP users served!
^^^^^^^




"Wayne Youngman" wrote in message
...
"TomG" wrote
ah, but the sheep have been taken by surprise by a trick or two...



TomG (Data)
you are charged on three counts,

a) Mentioning the word *sheep* in a technical forum.

b) Attempting to *Hi-jack* a technical thread.

and a minor misdeminour of being a *habitual* top-poster

How do you plead?
--
Wayne ][

Dorothy Bradbury wrote:
Actually BTX could run longer than ATX, looking at some later
thermal design specs - but the graphics card wattage is one issue.

The 75 watt max?


Going to be interesting anyway, improvement on the ATX warehouse.

Admittedly, the driver here is fan size = cost & fan qty = cost.
A big OEM selling point of BTX is less fans over ATX designs,
not that Dell haven't shown how it can be done - Apple too altho
that was perhaps with an excessive number of fans quite frankly.

One nice point is the downsizing of case size over ATX, which
in tower and even midi-tower form are somewhat overly big. It
also allows much bigger & heavier heatsink designs with a very
robust motherboard/case reinforcement - and of course gives
Intel another socket design to terminate non-motherboardal u/g.
--
Dorothy Bradbury



I dont think the problems are Fans/Heatsink(size)...

I think the problem is down more to overall airflow through the case and the
thermal property's of the material being used.. I say it's time to bring in
some of the newer cooling options, Using pulsating heat pipes, Like what
Tsheatronics used for their "Zen" coolers..
This (and other) technologys out there could move the heat over such a large
area that the whole cases could become the coolers, Instead of having air
flowing through the case you could simply have it flowing over the case, or
even through a cooler section of the case (allowing the guts to be seal'd)..

I have seen some new Heatsink materials out there on the market that would
make the SLK HSF's look like they dont even work..
When it's all said and done then I dont think we'll see thermal outputs of
150/200/250w coming from these chips.. My guess is that we'll start seeing
IBM/AMD bring in new processing methods that companys like Transmeta uses.
I mean why not.. Something as fast as a duron 500 that runs at 6w is what
we see out NOW...
If that same technology grows then in 2 or 3 years time we'll have something
similar that could still be passive cooled, and be more powerful than todays
64 3600+'s..

The idea of having everything being fed by a central fan will def work IF
the power consumption doesnt go verry high

My hopes is that the power consumptions will stay around where they are now,
100w for the CPU and 50-100w for the video/sound processing.. Anything more
is just overkill..