Intel: The chipset is the product

[Robert Myers]

What do you do when you're having competitive problems with your main
product and there is no guaranteed relief in sight?

You change the subject. At least, that's what I infer Intel to be doing
with Grantsdale:

http://www.chron.com/cs/CDA/ssistory.mpl/tech/news/2591485

<quote>

Intel making big push for chip set
Reuters News Service

SAN FRANCISCO - Intel Corp. plans to focus unprecedented attention on
one of the more obscure components of personal computers when it
launches its newest chip set next month.

< Not to be obscure for long, though. >

<snip>

Chip sets -- groups of integrated circuits that work together as humble
gatekeepers for data coming into and out of the core of the PC -- have
tended to miss out on the media and marketing attention paid to the
brains of the operation, the microprocessor.

< I'm enjoying this way too much. It's almost as if the Reuters
correspondent were on the Intel payroll. >

<snip>

Intel designed Grantsdale to lead a new generation of "entertainment
PCs" to be shipped later this year, part of a plan to bring PCs into the
living room and displace consumer electronics equipment.

William Leszinske, the director of marketing for Intel's "digital home"
initiative, said Intel is making a special effort to train retail
salespeople to explain the benefits of the Grantsdale chip set, even if
consumers aren't introduced to the intricacies of how a chip set works.

< No training needed here. Intel chipsets are already its not-so-secret
weapon. Let's see... what do we do better than AMD? Chipsets! When
did anybody see a chipset for an AMD processor they really trusted?
Chipsets it is then. >

< Much snippage. You really have to read the original story. >

The features in Grantsdale also could help persuade shoppers to seek out
Intel-based computers, said Intel spokeswoman Laura Anderson. That may
also steer shoppers away from PCs built with chips from rival Advanced
Micro Devices Inc..

< And finally, a nod to what is really going on. >

<snip>

Grantsdale, he said, will improve memory speed, communications,
multimedia, and wireless communications.

<end quote>

And fend off AMD products while Intel's architects regroup (that is, I
suspect, when they are not being reminded of just how much is on the
line at the moment, like their jobs).

Now then, I hope you hadn't invested too much time in trying to make
sense of all those Intel product numbers. Who wants a, er, whatever the
processor is called, when you can have the chipset of your dreams and an
"entertainment PC" to blow your neighbors away?

The _other_ company sells performance. We sell dreams. .


RM

 

[Grumble]

The _other_ company sells performance. We sell dreams. .

My, oh my. Are we in a trolling mood or what!

 

[Robert Myers]

My, oh my. Are we in a trolling mood or what!

The implication being that selling dreams over performance is a bad
thing? I don't happen to think it is.

Intel's Chief Architect of the P4 era says that he warned Intel
management that the NetBurst architecture had bought them some breathing
room--a few years at most.

The players in the Intel architecture mini-drama seem to an outsider as
if they might be stubbornly goal-oriented, but they don't, any of them,
seem inept.

Intel sometimes seems to act as if it could rearrange virtually
anything, including physics, to suit its market objectives. Anybody
knows that, in a showdown between physics and marketing, physics wins.

Except if you're Intel, apparently. .

Over the long haul, if people really need performance, and Intel can't
deliver it, Intel is in trouble. On the short haul, it would be
difficult and dangerous for Intel to try to persuade people that it is
silly for them to focus on performance differences they will never
notice. In a year or two, Intel might be right back where it was with
P4, trying to convince them that a 2.4GHz CPU was faster than a 1.7Ghz
CPU in a way that justified an upgrade. Right now, though, Intel needs
to get its customers to think about something other than single-threaded
performance without engaging in tedious and dangerous explanations.

Intel has big architectural changes in mind: offloading significant
pieces of work, like network processing, from the main CPU. If Intel
were giving advice in "The Graduate," it would be whispering
"Multi-threading." Intel has the clout to make a major change in
programming style like that stick, but even Intel can't make it happen
overnight. In the short haul, Intel has to concede single-threaded
performance to AMD and to get its customers to think about something
else. It looks to me as if they understood exactly what they need to
do, and they are doing it.

Eventually, Intel will be back to selling performance, but the kind of
performance it will have to sell is going to require significant
customer education.

The consumer CPU business may repeat the mistake of HPC and try to force
a vector quantity (usable performance) into a nearly meaningless scalar
(linpack, top 500, etc.), but I can hope not. With any luck, Intel will
be able to move the focus off single-threaded performance and things
that look like stand-ins for single-threaded performance onto more
sophisticated measures of value.

Sophistication in measuring value is great conversation for hardware
groups on Usenet, but when you're trying to make a sale, you don't want
the party reaching for his or her credit card to be thinking about
complexity. In such a circumstance, dreams are the preferred commodity
to be offering for sale.

RM

 

[George Macdonald]

What do you do when you're having competitive problems with your main
product and there is no guaranteed relief in sight?

You change the subject. At least, that's what I infer Intel to be doing
with Grantsdale:

http://www.chron.com/cs/CDA/ssistory.mpl/tech/news/2591485

Hmm, Centrino for the desktop? All they need is a name for the "package"
now. Any guesses... anybody?

Where do they get this stuff: "Because the chip set incorporates features
like Dolby audio and advanced 3-D video previously found only in add-on
cards..."? Wot a loada BULLSHIT! Intel plays catch-up and a buncha
anal...ysts drop their drawers in public!!! How embarrassing.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[Yousuf Khan]

What do you do when you're having competitive problems with your main
product and there is no guaranteed relief in sight?

You change the subject. At least, that's what I infer Intel to be
doing with Grantsdale:

http://www.chron.com/cs/CDA/ssistory.mpl/tech/news/2591485

They've already had experience hyping a chipset, otherwise known as
Centrino. They'll have to come up with a slick moniker for Grantsdale too. I
wonder if they're going to do like in the Centrino campaign, start
emphasizing the whole processor/chipset combo? They certainly can't hype the
Pentium 4 Prescott as it now stands -- it's a big embarrassment for them.
Perhaps with a re-emphasis on a chipset combo, rather than a processor, they
can easily take people's focus off of the actual processor and slip the old
processor away (Prescott), and replace it with a new processor (Dothan)
without people noticing? Prestidigitation.

Yousuf Khan

 

[Robert Myers]

Hmm, Centrino for the desktop? All they need is a name for the "package"
now. Any guesses... anybody?

Since picking a brand name is a big deal these days (there are IP issues
with practically any name you can think of), and since there is so
little time, one suspects that Intel is going to have to do without. If
this weren't a hurry-up job, they'd have a brand name ready and we'd
have been exposed to it dozens of times by now. Think of the marketing
barrage that preceded Centrino.

Where do they get this stuff: "Because the chip set incorporates features
like Dolby audio and advanced 3-D video previously found only in add-on
cards..."?

They get it from the press release, one gathers. I haven't yet found
the culpable press release on Intel's site, though.

Wot a loada BULLSHIT! Intel plays catch-up and a buncha
anal...ysts drop their drawers in public!!! How embarrassing.

You worry me sometimes. Can't you just relax and enjoy the show? .

RM

 

[chrisv]

The features in Grantsdale also could help persuade shoppers to seek out
Intel-based computers, said Intel spokeswoman Laura Anderson. That may
also steer shoppers away from PCs built with chips from rival Advanced
Micro Devices Inc..

Not to appear to defend a bunch of marketing BS, but some of us do
shop the chipset, obviously. Intel chipsets are the reason I've
always been an Intel customer. In fact, lately I've been buying
Intel-branded motherboards, confident in their quality, and confident
that if/when I install Linux on them, they'll be completely supported.
(I'm responsible for quite a number of computers at home and work.)

I just haven't seen a need to look elsewhere, recently. It's not like
I'd actually notice the 10% improvement in performance that I'd get by
going with AMD and a brand-X chipset (given equal dollars). Looking
forward, it appears that Intel may be in trouble with the lame
Prescott going against the superb Athlon 64, and I could live with an
Nvidia chipset... I'll cross that bridge when I'm next in the market
for a home machine.

No flames, please, we're all entitled to our opinions.

 

[Robert Myers]

Not to appear to defend a bunch of marketing BS, but some of us do
shop the chipset, obviously. Intel chipsets are the reason I've
always been an Intel customer. In fact, lately I've been buying
Intel-branded motherboards, confident in their quality, and confident
that if/when I install Linux on them, they'll be completely supported.
(I'm responsible for quite a number of computers at home and work.)

I just haven't seen a need to look elsewhere, recently. It's not like
I'd actually notice the 10% improvement in performance that I'd get by
going with AMD and a brand-X chipset (given equal dollars). Looking
forward, it appears that Intel may be in trouble with the lame
Prescott going against the superb Athlon 64, and I could live with an
Nvidia chipset... I'll cross that bridge when I'm next in the market
for a home machine.

No flames, please, we're all entitled to our opinions.

Why flame everyday good sense? Who could argue that Prescott is not a
disappointment or that the chipset is not an important part of a
purchase decision for a knowledgeable buyer?

If you look at what Intel is doing as purely a marketing ploy, it's
fairly transparent and easy to make fun of. If the market for PC-type
processors were mature (and some people do think it is), we should be
preparing for a future of Coke vs. Pepsi and Hertz vs. Avis marketing
campaigns.

As I read the signs and portents coming from both Intel and IBM, though,
the technology should not be taken to be mature and the market could be
headed for chaos.

In particular, the von Neumann architecture is a technological dead end
as far as increasing the power of microprocessors is concerned. It's a
dead end because, while there may be a few more doublings left in
Moore's law, the von Neumann architecture is already having trouble
making good use of the available transistors for a number of reasons:
heat, leakage, and the inevitable triumph of wire delay come to mind first.

If Intel wants buyers to think about upgrading for more power, it is
going to have to get buyers to think about something more complex than a
more powerful single-threaded x86 processor. It's not a problem Intel
is facing by itself. IBM, on whom AMD is reliant at the moment for
process technology, is having problems with scaling, too. It is hard to
imagine that AMD is not going to run into the same brick wall as Intel,
albeit more slowly because AMD did not make the self-destructive choice
that Intel did: to get to a faster clock at all costs.

If the von Neumann architecture is running out of steam, and if
entertainment is the future, then the future of x86 for "personal"
computing has to be threatened, too. In a universe of wild imagining,
maybe Sony/Toshiba/IBM, not AMD, is the real threat to Intel's dominance.

Not likely, but Intel would have its work cut out for it in the
"personal" computing business even without AMD: either people are going
to lose interest, the business will become a true commodity dominated by
the likes of VIA and Red Dragon, or Intel has to come up with something
to match the hype of a Cell because that's how much razzle dazzle it
will require to keep Intel in the style to which it has become accustomed.

No two ways about it, the disappointing performance of Prescott matched
against the technological success of AMD/IBM have forced Intel to
scramble, but it's a reasonable guess that Intel has actually known for
a while in what direction it was going to scramble when forced. The
fact that Intel didn't take the future they've been hinting about
seriously enough to come up with a more mature marketing campaign does
suggest a certain level of denial at the highest levels of Intel
management, though. ;-).

RM

 

[Neil Maxwell]

They've already had experience hyping a chipset, otherwise known as
Centrino. They'll have to come up with a slick moniker for Grantsdale too. I
wonder if they're going to do like in the Centrino campaign, start
emphasizing the whole processor/chipset combo?

There's been a tendency to capitalize on existing successful branding
campaigns - look at the life of the Pentium name, which was originally
just a replacement for '586. That would argue for the possibility of
names such as Centrino-D (for desktop), Centrino II, and such. One
thing the marketeers are aware of is the risk of confusing your
non-technical customers with too many name brands. You want them to
go "Oh, sure. Centrino. That's good, right? And Centrino II must be
better!"


Neil Maxwell - I don't speak for my employer

 

[Yousuf Khan]

I just haven't seen a need to look elsewhere, recently. It's not like
I'd actually notice the 10% improvement in performance that I'd get by
going with AMD and a brand-X chipset (given equal dollars). Looking
forward, it appears that Intel may be in trouble with the lame
Prescott going against the superb Athlon 64, and I could live with an
Nvidia chipset... I'll cross that bridge when I'm next in the market
for a home machine.

As they say, if you don't try any other products other than the ones you're
comfortable with, then how are you ever going to know the quality of the
competing products?

Yousuf Khan

 

[Yousuf Khan]

As I read the signs and portents coming from both Intel and IBM,
though, the technology should not be taken to be mature and the
market could be headed for chaos.

Why because they're having trouble getting good power consumption out of
their 90nm processes? What relevance has that got with whether the
technology is mature or not? The technology itself is mature because it's
been around forever. Nobody cares how it's made though.

In particular, the von Neumann architecture is a technological dead
end

Wow, I'm not even touching that. Where'd that come from?

No two ways about it, the disappointing performance of Prescott
matched against the technological success of AMD/IBM have forced
Intel to scramble, but it's a reasonable guess that Intel has
actually known for
a while in what direction it was going to scramble when forced. The
fact that Intel didn't take the future they've been hinting about
seriously enough to come up with a more mature marketing campaign does
suggest a certain level of denial at the highest levels of Intel
management, though. ;-).

I'm sure the scenarios have been played out within Intel for some time.
Knowing about all of the possible scenarios is one thing, but knowing
exactly which scenario is going to play out is to become god.

Yousuf Khan

 

[Yousuf Khan]

There's been a tendency to capitalize on existing successful branding
campaigns - look at the life of the Pentium name, which was originally
just a replacement for '586. That would argue for the possibility of
names such as Centrino-D (for desktop), Centrino II, and such. One
thing the marketeers are aware of is the risk of confusing your
non-technical customers with too many name brands. You want them to
go "Oh, sure. Centrino. That's good, right? And Centrino II must be
better!"

No doubt about it. There is nothing within the name "Centrino" that suggests
that it's only about mobile products, so it could easily be adopted by the
desktop guys.

Yousuf Khan

 

[Robert Myers]

Why because they're having trouble getting good power consumption out of
their 90nm processes? What relevance has that got with whether the
technology is mature or not? The technology itself is mature because it's
been around forever. Nobody cares how it's made though.

The question is most interesting if you take the technology to be the
PC. Your "nobody cares how it's made" would require a technlogical
miracle that no one in the history of parallel programming has so far
been able to deliver; viz, that you can add even a second thread without
rewriting everything.

To take an ulikely but still plausible script, suppose the Cell
processor is anything like the claims that have been made for it and
suppose that software developers come up with a way to harness all that
power in a way that makes a significant difference to the end user
experience. End of x86, unless x86 has something comparable to offer.

Wow, I'm not even touching that. Where'd that come from?

There are only so many ways you can get more throughput: wider, deeper,
faster. The performance of Prescott and IBM's declaration that scaling
is dead should give you a clue that faster and deeper are ceasing to be
options. That leaves only wider. Itanium carries the flag for
single-threaded wide issue in mainstream processors. It works, sort of,
but I know of no one who argues that it's a scalable solution. That
leaves multiple threads.

You can put more transitors on a die, but you can't use them for much of
anything because if you run them at a high clock to do useful work they
leak and they produce heat. One solution is to use those transistors
for huge cache: what a waste of beautiful silicon, and not a scalable
solution, either.

Slower, wider, cooler. That's the future. I can't believe you're
reading it here for the first time.

I'm sure the scenarios have been played out within Intel for some time.
Knowing about all of the possible scenarios is one thing, but knowing
exactly which scenario is going to play out is to become god.

Yes. I'm sure that Intel management has been told so many times that
the sky was falling that they could reasonably be excused for acting as
if the sky would never fall, even if they knew that it would. Denial,
though, is denial. The sky is going to fall.

RM

 

[KR Williams]

Why flame everyday good sense? Who could argue that Prescott is not a
disappointment or that the chipset is not an important part of a
purchase decision for a knowledgeable buyer?

If you look at what Intel is doing as purely a marketing ploy, it's
fairly transparent and easy to make fun of. If the market for PC-type
processors were mature (and some people do think it is), we should be
preparing for a future of Coke vs. Pepsi and Hertz vs. Avis marketing
campaigns.

As I read the signs and portents coming from both Intel and IBM, though,
the technology should not be taken to be mature and the market could be
headed for chaos.

In particular, the von Neumann architecture is a technological dead end
as far as increasing the power of microprocessors is concerned. It's a
dead end because, while there may be a few more doublings left in
Moore's law, the von Neumann architecture is already having trouble
making good use of the available transistors for a number of reasons:
heat, leakage, and the inevitable triumph of wire delay come to mind first.

If Intel wants buyers to think about upgrading for more power, it is
going to have to get buyers to think about something more complex than a
more powerful single-threaded x86 processor. It's not a problem Intel
is facing by itself. IBM, on whom AMD is reliant at the moment for
process technology, is having problems with scaling, too. It is hard to
imagine that AMD is not going to run into the same brick wall as Intel,
albeit more slowly because AMD did not make the self-destructive choice
that Intel did: to get to a faster clock at all costs.

If the von Neumann architecture is running out of steam, and if
entertainment is the future, then the future of x86 for "personal"
computing has to be threatened, too. In a universe of wild imagining,
maybe Sony/Toshiba/IBM, not AMD, is the real threat to Intel's dominance.

Not likely, but Intel would have its work cut out for it in the
"personal" computing business even without AMD: either people are going
to lose interest, the business will become a true commodity dominated by
the likes of VIA and Red Dragon, or Intel has to come up with something
to match the hype of a Cell because that's how much razzle dazzle it
will require to keep Intel in the style to which it has become accustomed.

No two ways about it, the disappointing performance of Prescott matched
against the technological success of AMD/IBM have forced Intel to
scramble, but it's a reasonable guess that Intel has actually known for
a while in what direction it was going to scramble when forced. The
fact that Intel didn't take the future they've been hinting about
seriously enough to come up with a more mature marketing campaign does
suggest a certain level of denial at the highest levels of Intel
management, though. ;-).

Good grief, Robert! We agree on something substantial! Write
this date down! ;-)

 

[George Macdonald]

Since picking a brand name is a big deal these days (there are IP issues
with practically any name you can think of), and since there is so
little time, one suspects that Intel is going to have to do without. If
this weren't a hurry-up job, they'd have a brand name ready and we'd
have been exposed to it dozens of times by now. Think of the marketing
barrage that preceded Centrino.

And yet most people still don't know what Centrino is or what it is
supposed to cover. How many times do you hear someone refer to a "Centrino
CPU"? I guess Centrium is too obvious as a name for the desktop package
and possibly, as you say, already taken... but a quick search doesn't turn
up anything in the computer industry which clashes.

I'm not so sure about "hurry-up" here... the way Intel has been sponsoring
startups in the home oriented multimedia sphere and filling in with
in-house nuts-n-bolts stuff. A big fanfare at rollout with dog 'n' pony
shows all over could have more impact than a ramped info-trickle.

They get it from the press release, one gathers. I haven't yet found
the culpable press release on Intel's site, though.


You worry me sometimes. Can't you just relax and enjoy the show? .

I can stand clueless but clueless pretending to be expert advice/opinion
prickles with me - bald-faced lies pisses me off.

Rgds, George Macdonald

"Just because they're paranoid doesn't mean you're not psychotic" - Who, me??

 

[Yousuf Khan]

And yet most people still don't know what Centrino is or what it is
supposed to cover. How many times do you hear someone refer to a
"Centrino CPU"? I guess Centrium is too obvious as a name for the
desktop package
and possibly, as you say, already taken... but a quick search doesn't
turn up anything in the computer industry which clashes.

Hell, even most of us that know better have to often use the Centrino term
instead of Pentium-M, just so that others will understand what we're talking
about.

Yousuf Khan

 

[chrisv]

No two ways about it, the disappointing performance of Prescott matched
against the technological success of AMD/IBM have forced Intel to
scramble, but it's a reasonable guess that Intel has actually known for
a while in what direction it was going to scramble when forced. The
fact that Intel didn't take the future they've been hinting about
seriously enough to come up with a more mature marketing campaign does
suggest a certain level of denial at the highest levels of Intel
management, though. ;-).

Well, they're probably in a state of shock over what appears to be the
impending death of IA64. They had high hopes for that unclonable
architecture, and the profitability of a de facto monopoly at the high
end. Now it seems that they'll have to slug it out in AMD and the
others for a couple more decades... 8)

 

[Robert Myers]

KR Williams wrote:

<snip

Good grief, Robert! We agree on something substantial! Write
this date down! ;-)

As with all things Usenet, the event is safely recorded for posterity.
Think of the effort required for constantly successful disagreement. It
just couldn't be worth it. .

RM

 

[Robert Myers]

I'm not so sure about "hurry-up" here... the way Intel has been sponsoring
startups in the home oriented multimedia sphere and filling in with
in-house nuts-n-bolts stuff. A big fanfare at rollout with dog 'n' pony
shows all over could have more impact than a ramped info-trickle.

They've definitely been clearing ground for this current agenda for a
while, but I think it's fair to infer that the exact timing and delivery
of much of this stuff is being forced upon them. If Prescott had turned
out the way Intel wanted to, we'd be hearing about Megahertz, not chipsets.

I can stand clueless but clueless pretending to be expert advice/opinion
prickles with me - bald-faced lies pisses me off.

Intel is using the relative technological unsophistication of those who
write for the press to get its advertising message across as hard news.
They didn't invent the game, of course. Every technology company
draws from the same pool of PR types, and it would be amazing if the PR
style of Intel differed significantly from industry norms in terms of
what comes natural.

Intel does seem to me to be much more calculating about its message than
most, and they seem to make it work for them. I am probably more
inclined than the average technologist to pay attention to these sorts
of things, but it really does seem to me that you can't understand what
Intel is up to without understanding the messages it is trying to
convey. That's why I take up bandwidth in hardware groups calling
attention on it. .

As to genuine cluelessness/misinformation, it seems to me like you would
need some kind of logarithmic scale. Consumers aren't very well
informed about the actual properties of the laundry detergents they buy.
The difference, you might fairly object, is that technical-sounding
press releases from Proctor and Gamble don't frequently show up in the
press as hard news. Don't know what to say about that.

I give Intel considerable credit for having successfully cultivated a
market by persuading so many people that they needed all that muscle to
begin with. I don't think things like that just happen. I have to be
careful with this line of thinking, though, because it would eventually
lead to my expression very grudging admiration for Microsoft, and we
wouldn't want that.

RM

 

[chrisv]

As they say, if you don't try any other products other than the ones you're
comfortable with, then how are you ever going to know the quality of the
competing products?

Well, I have eyes and ears, and yes, still some prejudices. 8)

One example: A friend of mine, who had purchased a PC pre-configured
with Lindows 3. Cheap, brand-X motherboard and chipset, of course.
When he installed Lindows 4 on the same machine, the sound wouldn't
work. We tried some other Linux distributions and had issues with the
on-board video. Yuck. He's now ordering, on my recommendation, an
Intel D865GBFL, and I'd be shocked it if wasn't properly supported
(we're putting on the new Fedora Core 2, again on my recommendation).