Berkeley researcher describes parallel path

Anne & Lynn Wheeler

unread,

Feb 22, 2008, 12:28:00 AM2/22/08

to

Berkeley researcher describes parallel path
http://www.eetimes.com/news/latest/showArticle.jhtml?articleID=206801229
Wintel will fund parallel software lab at Berkeley
http://www.eetimes.com/showArticle.jhtml;?articleID=206503988
Multicore puts screws to parallel-programming models
http://www.eetimes.com/news/latest/showArticle.jhtml?articleID=206504466

from above:

Intel Corp. and Microsoft Corp. have awarded an estimated $10 million,
five-year grant to help fund a new Parallel Computing Lab at the
University of California at Berkeley, with 14 faculty members initially
involved. As many as 20 universities, including MIT, Stanford and the
University of Illinois, competed for funding.

... snip ...

misc. past:

Anne & Lynn Wheeler

unread,

Feb 22, 2008, 2:40:37 PM2/22/08

to

re:
http://www.garlic.com/~lynn/2008d.html#81 Berkeley researcher describes parallel path

another of the articles on the same theme:

Opinion: Time to plow multiple paths to parallel computing
http://www.eetimes.com/news/latest/showArticle.jhtml?articleID=206801376

from above:

The electronics industry is not investing nearly enough time, energy and
money to address what has become a pressing need for a genuine
technology breakthrough in parallel computing. It is time to fund
multiple large-scale projects in this area and multiple stakeholders
need to step up to the plate quickly.

... snip ...

we had been doing some of this sort of stuff when we were doing
ha/cmp
http://www.garlic.com/~lynn/subtopic.html#hacmp
scaleup
http://www.garlic.com/~lynn/95.html#13
and
http://www.garilc.com/~lynn/lhwemail.html#medusa

some old reference to genre with this old reference
http://www.garlic.com/~lynn/2007d.html#email870604
in this post
http://www.garlic.com/~lynn/2007d.html#39 old tapes

of course lots of this also showed up in the "killer micros" from the
90s ... that were going to replace all the mainframes. business process
reengineering was going to move all the business applications off the
backroom mainframes onto large arrays of smaller processors.

in the financial industry, large amounts were spent in conjunction with
the parallelising efforts which would also address the increasing
"overnight batch window" problem ... with (significantly parallelized)
straight through processing. some of the disastrous failures of the
period was attempts to use COTS and standardized parallelizing solutions
... which turned out to increase the processing by factors of one
hundred times (vis-a-vis the batch implementations) .... totally
obliterating any hopes of increased workload thruput.

Al Kossow

unread,

Feb 22, 2008, 3:05:02 PM2/22/08

to

Anne & Lynn Wheeler wrote:
> It is time to fund
> multiple large-scale projects in this area and multiple stakeholders
> need to step up to the plate quickly.

Microsoft's effort is being lead by Chuck Thacker.
Should be interesting...

--
Posted via a free Usenet account from http://www.teranews.com

Anne & Lynn Wheeler

unread,

Feb 22, 2008, 3:27:17 PM2/22/08

to

re:
http://www.garlic.com/~lynn/2008d.html#81 Berkeley researcher describes parallel path

http://www.garlic.com/~lynn/2008d.html#87 Berkeley researcher describes parallel path

somewhat related (including parallelized straight through processing and
eliminating overnight batch window):

Accelerating Wall Street
http://wallstreetandtech.com/accelerate/

part of the agenda:
http://wallstreetandtech.com/accelerate/agenda.jhtml;jsessionid=NCF5XKQUSJJBEQSNDLPSKHSCJUNN2JVN

from above:

Data Tsunami: Using Multi-Core Processing to Stay Ahead

Message volume is going through the roof. NYSE Euronext is continually
setting new records for message volume. The Options Price Reporting
Authority (OPRA) is projecting required message capacity of 907,000
messages per second (mps) by this summer (80 percent higher than a year
ago mps) with a 20 percent growth to more than 6 billion messages per
day. What can multi-core processors and advanced architectures do to
help firms stay ahead of the data flow?

... snip ...

Anne & Lynn Wheeler

unread,

Feb 22, 2008, 3:33:36 PM2/22/08

to

Al Kossow <a...@spies.com> writes:
> Microsoft's effort is being lead by Chuck Thacker.
> Should be interesting...

re:
http://www.garlic.com/~lynn/2008d.html#87 Berkeley researcher describes parallel path

the referenced article
http://www.eetimes.com/news/latest/showArticle.jhtml?articleID=206801376

also mentions:

I know Intel has already funded some efforts to help train the next
generation of programmers in the latest parallel techniques. And
Microsoft has hired a handful of heavy hitters in parallel programming
including Burton Smith and Dan Reed.

... snip ...

the above has pointer to this article from last summer:

M'soft: Parallel programming model 10 years off
http://www.eetimes.com/showArticle.jhtml;?articleID=201200019

from above:

"There is a fundamental change in computer architecture coming," Craig
Mundie, chief research and strategy officer at Microsoft, said in an
interview with EE Times. "I personally think this is one of the most
disruptive things the industry will have to go through."

... snip ...

and from last fall:

Supercomputer vet joins Microsoft Research
http://www.eetimes.com/showArticle.jhtml;?articleID=202804148

from above:

Reed will collaborate with Burton Smith, another parallel computing guru
who joined Microsoft in 2005 to help spearhead work on multicore
issues. In an interview earlier this year, Smith discussed work on at
least two functional languages at Microsoft.

... snip ...

Quadibloc

unread,

Feb 22, 2008, 5:24:33 PM2/22/08

to

On Feb 22, 12:40 pm, Anne & Lynn Wheeler <l...@garlic.com> quoted, in
part:

> The electronics industry is not investing nearly enough time, energy and
> money to address what has become a pressing need for a genuine
> technology breakthrough in parallel computing.

This is not necessarily true.

The existence of an energy crisis does not necessarily imply that the
government is not spending enough time and money on researching
perpetual motion machines.

Similarly, the fact that the electronics industry is not researching
ways for nine women to give birth to one baby in one month may simply
indicate good common sense on the part of the electronics industry, no
matter how badly we need faster computers, and how limited we are to
parallelism as a way to make them.

But training programmers - or writing compilers - to make good use of
the opportunities for exploiting parallelism that are available, where
steps are not genuinely dependent - of course, is a good idea. Looking
for the magic bullet, though, may not be, and the quote could be taken
as advocating the search for exactly that.

John Savard

Scott Lurndal

unread,

Feb 22, 2008, 9:11:26 PM2/22/08

to

Al Kossow <a...@spies.com> writes:
>Anne & Lynn Wheeler wrote:
>> It is time to fund
>> multiple large-scale projects in this area and multiple stakeholders
>> need to step up to the plate quickly.
>
>Microsoft's effort is being lead by Chuck Thacker.
>Should be interesting...
>

I think the paradigm needs to change. Perhaps back to data-flow?

scott

CBFalconer

unread,

Feb 22, 2008, 10:40:57 PM2/22/08

to

Quadibloc wrote:
>
... snip ...

>
> Similarly, the fact that the electronics industry is not researching
> ways for nine women to give birth to one baby in one month may simply
> indicate good common sense on the part of the electronics industry, no
> matter how badly we need faster computers, and how limited we are to
> parallelism as a way to make them.

I think it more likely indicates that the men involved are not
especially interested in heavy baby production. They find the
preliminaries far more amusing.

--
[mail]: Chuck F (cbfalconer at maineline dot net)
[page]: <http://cbfalconer.home.att.net>
Try the download section.

Walter Bushell

unread,

Feb 23, 2008, 12:38:40 AM2/23/08

to

In article <47BF95C9...@yahoo.com>,
CBFalconer <cbfal...@yahoo.com> wrote:

> > Similarly, the fact that the electronics industry is not researching
> > ways for nine women to give birth to one baby in one month may simply
> > indicate good common sense on the part of the electronics industry, no
> > matter how badly we need faster computers, and how limited we are to
> > parallelism as a way to make them.
>
> I think it more likely indicates that the men involved are not
> especially interested in heavy baby production. They find the
> preliminaries far more amusing.

But has anyone tried the opposite approach, one woman sleeping with 9
men to get a baby in a month?

Quadibloc

unread,

Feb 23, 2008, 11:39:21 PM2/23/08

to

On Feb 22, 7:11 pm, sc...@slp53.sl.home (Scott Lurndal) wrote:

> I think the paradigm needs to change. Perhaps back to data-flow?

But you can't fit patch cords and plugboards on a chip, so data flow
is badly suited to today's hardware!

John Savard

Andrew Swallow

unread,

Feb 24, 2008, 9:44:10 AM2/24/08

to

But you can fit High-Low gates and digital cross-switches.
See FPGAs.

Andrew Swallow

Larry Elmore

unread,

Feb 24, 2008, 2:42:20 PM2/24/08

to

I knew a gal in the Air Force who tried it. She decided she really
didn't like being in the service and while in tech school did her
absolute damndest to get pregnant (that being an easy way for women to
"opt out"). She succeeded. No baby in one month, though at the rate
she was going through men, she might have been trying for the "baby in a
week" record. I heard she got an abortion shortly after separating.
And no, I wasn't part of the experiment.

Lon

unread,

Feb 24, 2008, 5:04:54 PM2/24/08

to

Those wouldn't be as fitting as nanomachines. It would be interesting
to see whether or not you could duplicate some of the old IBM Relay
logic and how elegantly with micro actuators, etc.

greymaus

unread,

Feb 24, 2008, 5:25:28 PM2/24/08

to

One woman being married to many men was common in the Tibet area.. at
one time, anyway.. polyandry.?

--
greymaus
Just Another Grumpy Old Man

Eugene Miya

unread,

Feb 25, 2008, 2:11:26 PM2/25/08

to

In article <c283e8e6-b1a4-4251...@o77g2000hsf.googlegroups.com>,

Well, what do you think cross bar switches are?

Today's hardware is still largely the same kinds of von Neumann
microprocessors used in the 80s and 90s data flow machines.
Hardware isn't your problem if you want to do data flow.

I just hope that the Manchester has pieces stored some where.
I suspect the DEC M31 Andromeda was torn apart.

You have to decide first why you are assemlbing a data flow machine
because the drop in apps aren't there. Not a problem for research.

--

Eugene Miya

unread,

Feb 25, 2008, 2:18:05 PM2/25/08

to

In article <47BF95C9...@yahoo.com>,
CBFalconer <cbfal...@maineline.net> wrote:

>Quadibloc wrote:
>> Similarly, the fact that the electronics industry is not researching
>> ways for nine women to give birth to one baby in one month may simply
>> indicate good common sense on the part of the electronics industry, no
>> matter how badly we need faster computers, and how limited we are to
>> parallelism as a way to make them.

The software engineering community knows about Brooks' law, but the
architecture community always attempts to reject Amdahl's similar effort.
Wonder who will win?

>I think it more likely indicates that the men involved are not
>especially interested in heavy baby production. They find the
>preliminaries far more amusing.

A good point.
This is also one of the criticism informally leveled at AI.

--

Eugene Miya

unread,

Feb 25, 2008, 2:23:06 PM2/25/08

to

In article <265f1cc5-a17a-448f...@d21g2000prf.googlegroups.com>,
Quadibloc <jsa...@ecn.ab.ca> wrote:
>On Feb 22, 12:40=A0pm, Anne & Lynn Wheeler <l...@garlic.com> quoted, in

>part:
>> The electronics industry is not investing nearly enough time, energy and
>> money to address what has become a pressing need for a genuine
>> technology breakthrough in parallel computing.
>
>This is not necessarily true.

electronics != computer

>The existence of an energy crisis does not necessarily imply that the
>government is not spending enough time and money on researching
>perpetual motion machines.
>
>Similarly, the fact that the electronics industry is not researching
>ways for nine women to give birth to one baby in one month may simply
>indicate good common sense on the part of the electronics industry, no
>matter how badly we need faster computers, and how limited we are to
>parallelism as a way to make them.

We don't need faster computers badly enough.

>But training programmers - or writing compilers - to make good use of
>the opportunities for exploiting parallelism that are available, where
>steps are not genuinely dependent - of course, is a good idea. Looking
>for the magic bullet, though, may not be, and the quote could be taken
>as advocating the search for exactly that.

The number of universities training programmers how to write decent
parallel compilers in the USA can be counted on 1 hand. I think the
number in Canada is 0 or maybe 1. Depends how one counts.

--

Eugene Miya

unread,

Feb 25, 2008, 2:26:18 PM2/25/08

to

In article <47bf1f5a$0$26049$8826...@free.teranews.com>,

Al Kossow <a...@spies.com> wrote:
>Anne & Lynn Wheeler wrote:
>> It is time to fund
>> multiple large-scale projects in this area and multiple stakeholders
>> need to step up to the plate quickly.
>
>Microsoft's effort is being lead by Chuck Thacker.
>Should be interesting...

Wasted money. God didn't have an established base to quote an old
Datamation or Computer World.

Lynn sounds like Mr. Watson.

--

Al Kossow

unread,

Feb 25, 2008, 3:01:16 PM2/25/08

to

To be fair to Lynn, that was Patterson's quote.

MS/Intel are lining up some parallel architecture hw heavy hitters. I've not
heard of anyone I recognize on the software side.

This is all about semi vendors trying to figure out what to do with all the
transistors they can fab. I knew someone at LSI Logic working on tools methodology
for dealing with the complexity of billion transistor logic design. Haven't heard
much about that in the last couple of years.

Message has been deleted

Anne & Lynn Wheeler

unread,

Feb 25, 2008, 3:55:36 PM2/25/08

to

Al Kossow <a...@spies.com> writes:
> This is all about semi vendors trying to figure out what to do with
> all the transistors they can fab. I knew someone at LSI Logic working
> on tools methodology for dealing with the complexity of billion
> transistor logic design. Haven't heard much about that in the last
> couple of years.

some of the old timers at lsi logic had come from varian where cp67 was
used for running some of the design tools ... and then they had vm370
infrastructure deployed at lsi logic for quite some time ... well into
the late 80s ... before starting to transition to other platforms.

one of the people there had done lots of work on c compiler optimizing
for (cms) 370 ... as part of porting some of the design tools from
berkeley.

we would frequently visit with them up thru late 90s. there were some
x-over between some lsi logic chip people and sun people ... both on
sparc chip ... and for some complicated reasons ... the sun
object-oriented spring operating system (in same timeframe and similar
effort to what apple was doing with pink). at some point we were
approached about heading up effort turning spring into commercial
offering ... in conjunction with large collections of sparc processors
with a form of SCI interconnect.

LSI logic also was doing some work in the area of chip design
verification with some people at stanford and had also hired the math
institute in a different country.

old email discussing older SCI meetings:
http://www.garlic.com/~lynn/2001f.html#email920630
in this post:
http://www.garlic.com/~lynn/2001f.html#11 Climate, US, Japan & supers query

misc. past post in thread:

http://www.garlic.com/~lynn/2008d.html#81 Berkeley researcher describes parallel path

http://www.garlic.com/~lynn/2008d.html#87 Berkeley researcher describes parallel path
http://www.garlic.com/~lynn/2008d.html#89 Berkeley researcher describes parallel path
http://www.garlic.com/~lynn/2008d.html#90 Berkeley researcher describes parallel path

Anne & Lynn Wheeler

unread,

Feb 25, 2008, 4:26:06 PM2/25/08

to

for more multi-core:

Sun Leaks 6-core Intel Xeon, Nehalem Details
http://www.dailytech.com/Sun%20Leaks%206core%20Intel%20Xeon%20Nehalem%20Details/article10834.htm

from above:

Dunnington includes 16MB of L3 cache shared by all six processors. Each
pair of cores can also access 3MB of local L2 cache. The end result is
a design very similar to the AMD Barcelona quad-core processor; however,
each Barcelona core contains 512KB L2 cache, whereas Dunnington cores
share L2 cache in pairs.

... snip ...

Andrew Swallow

unread,

Feb 27, 2008, 11:01:38 PM2/27/08

to

Al Kossow wrote:
[snip]

> This is all about semi vendors trying to figure out what to do with all the
> transistors they can fab.

Programmers can still use more ram than the computer can hold. On chip
ram can clock at a faster speed than the bus. So on chip ram permits
programs to be both bigger and faster.

Andrew Swallow

krw

unread,

Feb 28, 2008, 7:35:37 PM2/28/08

to

In article <xZudneaFe9-7r1va...@bt.com>,
am.sw...@btinternet.com says...

Really? On chip RAM can't be bigger than off, so you statement is
horse-hockey.

--
Keith

Andrew Swallow

unread,

Feb 28, 2008, 7:53:34 PM2/28/08

to

Now you are being silly.

You can have slow off CPU chip ram as well.

Andrew Swallow

Lon

unread,

Feb 28, 2008, 9:26:54 PM2/28/08

to

... "This, is a nanosecond"

Pay attention. He said it can clock at higher speeds, which indicates
that he might actually know something about hardware. On the other
hand....

krw

unread,

Feb 28, 2008, 10:11:23 PM2/28/08

to

In article <78udnRsYs-cQylra...@bt.com>,
am.sw...@btinternet.com says...

> krw wrote:
> > In article <xZudneaFe9-7r1va...@bt.com>,
> > am.sw...@btinternet.com says...
> >> Al Kossow wrote:
> >> [snip]
> >>
> >>> This is all about semi vendors trying to figure out what to do with all the
> >>> transistors they can fab.
> >> Programmers can still use more ram than the computer can hold. On chip
> >> ram can clock at a faster speed than the bus. So on chip ram permits
> >> programs to be both bigger and faster.
> >
> > Really? On chip RAM can't be bigger than off, so you statement is
> > horse-hockey.
> >
> Now you are being silly.

No, you're being stupid, again.

> You can have slow off CPU chip ram as well.

Then it's not the on-chip RAM that's permitting bigger programs,
huh?

--
Keith

krw

unread,

Feb 28, 2008, 10:13:14 PM2/28/08

to

In article <J5ydneiFm-Lo8Fra...@comcast.com>,
lon.s...@comcast.net says...

> krw wrote:
> > In article <xZudneaFe9-7r1va...@bt.com>,
> > am.sw...@btinternet.com says...
> >> Al Kossow wrote:
> >> [snip]
> >>
> >>> This is all about semi vendors trying to figure out what to do with all the
> >>> transistors they can fab.
> >> Programmers can still use more ram than the computer can hold. On chip
> >> ram can clock at a faster speed than the bus. So on chip ram permits
> >> programs to be both bigger and faster.
> >
> > Really? On chip RAM can't be bigger than off, so you statement is
> > horse-hockey.
> >
> ... "This, is a nanosecond"

It's bigger than that. It's about a 8" long on a circuit board.

>
> Pay attention. He said it can clock at higher speeds, which indicates
> that he might actually know something about hardware. On the other
> hand....

Hardly. He said "So on chip ram permits programs to be both bigger
and faster".

Faster, maybe. Bigger? Bullshit.

--
Keith

Rostyslaw J. Lewyckyj

unread,

Mar 6, 2008, 12:20:52 AM3/6/08

to

Quadibloc wrote:
> On Feb 22, 12:40 pm, Anne & Lynn Wheeler <l...@garlic.com> quoted, in
> part:
>
>> The electronics industry is not investing nearly enough time, energy and
>> money to address what has become a pressing need for a genuine
>> technology breakthrough in parallel computing.
>
> This is not necessarily true.
>
> The existence of an energy crisis does not necessarily imply that the
> government is not spending enough time and money on researching
> perpetual motion machines.
>
> Similarly, the fact that the electronics industry is not researching
> ways for nine women to give birth to one baby in one month
>

Well no, you rally need to reformulate that problem a bit.
And nine women is too few.
You should have about eighteen women. Then after a lead time of nine
months you should comfortably be able to produce one baby per month.
And you could probably get away with having only one male although
he might not last for long.

Rostyslaw J. Lewyckyj

unread,

Mar 6, 2008, 1:08:07 AM3/6/08

to

There is enough historical experience to show that that doesn't work.

krw

unread,

Mar 6, 2008, 6:16:13 PM3/6/08

to

In article <69Lzj.11766$dT....@bignews1.bellsouth.net>,
urj...@bellsouth.net says...

> Quadibloc wrote:
> > On Feb 22, 12:40 pm, Anne & Lynn Wheeler <l...@garlic.com> quoted, in
> > part:
> >
> >> The electronics industry is not investing nearly enough time, energy and
> >> money to address what has become a pressing need for a genuine
> >> technology breakthrough in parallel computing.
> >
> > This is not necessarily true.
> >
> > The existence of an energy crisis does not necessarily imply that the
> > government is not spending enough time and money on researching
> > perpetual motion machines.
> >
> > Similarly, the fact that the electronics industry is not researching
> > ways for nine women to give birth to one baby in one month
> >
> Well no, you rally need to reformulate that problem a bit.
> And nine women is too few.

If nine American women can't do the job then out source it to China.

> You should have about eighteen women. Then after a lead time of nine
> months you should comfortably be able to produce one baby per month.
> And you could probably get away with having only one male although
> he might not last for long.

With eighteen women and a new baby a month? There is no such thing
as "long", at least to an outsider (relativity applies).

--
Keith

Eugene Miya

unread,

Mar 6, 2008, 7:58:01 PM3/6/08

to

In article <MPG.2239cd55e...@news.individual.net>,

krw <k...@att.bizzzzzzzzzz> wrote:
>In article <69Lzj.11766$dT....@bignews1.bellsouth.net>,
>urj...@bellsouth.net says...
>> Quadibloc wrote:
>> > On Feb 22, 12:40 pm, Anne & Lynn Wheeler <l...@garlic.com> quoted, in
>> > part:
>> >> The electronics industry is not investing nearly enough time, energy and
>> >> money to address what has become a pressing need for a genuine
>> >> technology breakthrough in parallel computing.
>> >
>> > This is not necessarily true.
>> >
>> > The existence of an energy crisis does not necessarily imply that the
>> > government is not spending enough time and money on researching
>> > perpetual motion machines.
>> >
>> > Similarly, the fact that the electronics industry is not researching
>> > ways for nine women to give birth to one baby in one month
>> >
>> Well no, you rally need to reformulate that problem a bit.
>> And nine women is too few.
>
>If nine American women can't do the job then out source it to China.

Nice try, nice thought, but if the spec is American and calls for
blond haired, blue eyed babies, the Chinese would have a hard time
fulfilling a spec. They would have to sub-contract back to the USA.

It's those hidden small print clauses which cause such problems.

>> You should have about eighteen women. Then after a lead time of nine
>> months you should comfortably be able to produce one baby per month.

Pipelining? Naw, we want immediate parallelism for future Crusaders.

>> And you could probably get away with having only one male although
>> he might not last for long.
>
>With eighteen women and a new baby a month? There is no such thing
>as "long", at least to an outsider (relativity applies).

Cloning.

--

CBFalconer

unread,

Mar 6, 2008, 11:26:47 PM3/6/08

to

krw wrote:
> urj...@bellsouth.net says...
>>
... snip ...

>>
>> You should have about eighteen women. Then after a lead time of
>> nine months you should comfortably be able to produce one baby
>> per month. And you could probably get away with having only one
>> male although he might not last for long.
>
> With eighteen women and a new baby a month? There is no such
> thing as "long", at least to an outsider (relativity applies).

Shouldn't the subject become "Pipelineing with a 50% duty cycle
limit"?

--
[mail]: Chuck F (cbfalconer at maineline dot net)
[page]: <http://cbfalconer.home.att.net>
Try the download section.

--

krw

unread,

Mar 7, 2008, 9:01:48 AM3/7/08

to

In article <47d09319$1@darkstar>, eug...@cse.ucsc.edu says...

> In article <MPG.2239cd55e...@news.individual.net>,
> krw <k...@att.bizzzzzzzzzz> wrote:
> >In article <69Lzj.11766$dT....@bignews1.bellsouth.net>,
> >urj...@bellsouth.net says...
> >> Quadibloc wrote:
> >> > On Feb 22, 12:40 pm, Anne & Lynn Wheeler <l...@garlic.com> quoted, in
> >> > part:
> >> >> The electronics industry is not investing nearly enough time, energy and
> >> >> money to address what has become a pressing need for a genuine
> >> >> technology breakthrough in parallel computing.
> >> >
> >> > This is not necessarily true.
> >> >
> >> > The existence of an energy crisis does not necessarily imply that the
> >> > government is not spending enough time and money on researching
> >> > perpetual motion machines.
> >> >
> >> > Similarly, the fact that the electronics industry is not researching
> >> > ways for nine women to give birth to one baby in one month
> >> >
> >> Well no, you rally need to reformulate that problem a bit.
> >> And nine women is too few.
> >
> >If nine American women can't do the job then out source it to China.
>
> Nice try, nice thought, but if the spec is American and calls for
> blond haired, blue eyed babies, the Chinese would have a hard time
> fulfilling a spec. They would have to sub-contract back to the USA.

Ok, Russia.

> It's those hidden small print clauses which cause such problems.

Shoot the lawyers.

> >> You should have about eighteen women. Then after a lead time of nine
> >> months you should comfortably be able to produce one baby per month.
> Pipelining? Naw, we want immediate parallelism for future Crusaders.
> >> And you could probably get away with having only one male although
> >> he might not last for long.
> >
> >With eighteen women and a new baby a month? There is no such thing
> >as "long", at least to an outsider (relativity applies).
>
> Cloning.
>

Don't cloned babies wake up for 2:00AM diapering and feeding too?

--
Keith
--
Keith

Eugene Miya

unread,

Mar 7, 2008, 4:06:05 PM3/7/08

to

>> >> >> The electronics industry is not investing nearly enough time, energy and
>> >> >> money to address what has become a pressing need for a genuine
>> >> >> technology breakthrough in parallel computing.

>> >> > The existence of an energy crisis does not necessarily imply that the
>> >> > government is not spending enough time and money on researching
>> >> > perpetual motion machines.
>> >> > Similarly, the fact that the electronics industry is not researching
>> >> > ways for nine women to give birth to one baby in one month
>> >> Well no, you rally need to reformulate that problem a bit.
>> >> And nine women is too few.
>> >If nine American women can't do the job then out source it to China.
>>
>> Nice try, nice thought, but if the spec is American and calls for
>> blond haired, blue eyed babies, the Chinese would have a hard time
>> fulfilling a spec. They would have to sub-contract back to the USA.

In article <MPG.223a7e51f...@news.individual.net>,
krw <k...@att.bizzzzzzzzzz> wrote:
>Ok, Russia.

Russians might work. Subcontract to them, take a cut.

>> It's those hidden small print clauses which cause such problems.
>
>Shoot the lawyers.

Whereas this has been cited, and occasionally tried, it seems that the
cases where this has been caught on video the guys doing the shooting
appear to have been poor shots <or poor blows> and used less than lethal ammo.

I suspect that the shear numbers of lawyers no longer make this practical.
I would prefer to suggest perhaps sterilization (something like
Cobalt-60 and sterile screw worm flies), but that might take too
long and their education system might produce many more.

>> >> You should have about eighteen women. Then after a lead time of nine
>> >> months you should comfortably be able to produce one baby per month.
>> Pipelining? Naw, we want immediate parallelism for future Crusaders.
>> >> And you could probably get away with having only one male although
>> >> he might not last for long.
>> >
>> >With eighteen women and a new baby a month? There is no such thing
>> >as "long", at least to an outsider (relativity applies).
>>
>> Cloning.
>>
>Don't cloned babies wake up for 2:00AM diapering and feeding too?

You breed the clones by human selection to require fewer 2AM wake ups
and feeding. Since you are dealing with many embryos, you can select
your clones faster than normal reproductive cycles.

--

Rostyslaw J. Lewyckyj

unread,

Mar 7, 2008, 5:04:00 PM3/7/08

to

CBFalconer wrote:
> krw wrote:
>> urj...@bellsouth.net says...
> ... snip ...
>>> You should have about eighteen women. Then after a lead time of
>>> nine months you should comfortably be able to produce one baby
>>> per month. And you could probably get away with having only one
>>> male although he might not last for long.
>> With eighteen women and a new baby a month? There is no such
>> thing as "long", at least to an outsider (relativity applies).
>
> Shouldn't the subject become "Pipelineing with a 50% duty cycle
> limit"?
>

But is it pipelining, since the operation is not broken up into
stages. I think it's rather overlapped execution of parallel
execution units invoked in a round robin fashion.
So definitely parallel cpus.
Anyway, don't change the subject. That would surely get /BAH
all annoyed.
--
Rostyk

jmfb...@aol.com

unread,

Apr 4, 2008, 6:34:46 AM4/4/08

to

In article <rRLzj.343$DT3...@bignews3.bellsouth.net>,

What are you talking about? 8 of them could be shooting blanks.

/BAH

Rostyslaw J. Lewyckyj

unread,

Apr 4, 2008, 1:20:10 PM4/4/08

to

But the process in question is to get not to beget a baby.
You could also have quibbled about the interpretation of sleeping.
> /BAH