Well, hell hath frozen over, Dell takes Opteron

  • Thread starter Thread starter Yousuf Khan
  • Start date Start date
I guess there wasn't much choice left for Dell. Several months ago,
people would've looked at this as a good thing for AMD. Now it's looked
upon as a good thing for Dell.

CRN | Servers, Opteron | Dell To Use AMD In New Servers
http://www.crn.com/sections/breakin...B4WQWSQSNDBECKH0CJUMEKJVN?articleId=188100223

Dell Finds a Friend
http://www.thestreet.com/_yahoo/tech/hardware/10286708.html?cm_ven=YAHOO&cm_cat=FREE&cm_ite=NA

Yousuf Khan

And to add insult to injury, Dell is going to use AMD not in $299
desktops or even pizzabox 1-2p servers, but in top of the line 4p+
servers. I bet Opterons 8xx have the widest margin of all AMD lineup,
whereas Celerons (the $299 box stuff) have the margin close to 0,
possibly even on negative side ;-)
And after this psychological barrier is broken, it will be even easier
for Dell to add AMD to its 2p server and XPS lineup later on.

NNN
 
And to add insult to injury, Dell is going to use AMD not in $299
desktops or even pizzabox 1-2p servers, but in top of the line 4p+
servers. I bet Opterons 8xx have the widest margin of all AMD lineup,
whereas Celerons (the $299 box stuff) have the margin close to 0,
possibly even on negative side ;-)
And after this psychological barrier is broken, it will be even easier
for Dell to add AMD to its 2p server and XPS lineup later on.

NNN

Dell is simply sowing what it reaped...when the ink dried on the Intel
exclusive deal, Dell committed itself to a long protracted death. Smart
money would jump on the AMD bandwagon, and I would love to buy a Dell. But
as long as Dell is pandering to last generation technology in its consumer
desktop, they'll not get my business.

HP and Gateway are getting the dollars I would have spent on Dell if they
had AMD available.

Can't see paying more and getting less (Intel).

Bobby
 
NoNoBadDog! said:
Dell is simply sowing what it reaped...when the ink dried on the Intel
exclusive deal, Dell committed itself to a long protracted death. Smart
money would jump on the AMD bandwagon, and I would love to buy a Dell. But
as long as Dell is pandering to last generation technology in its consumer
desktop, they'll not get my business.

HP and Gateway are getting the dollars I would have spent on Dell if they
had AMD available.

Can't see paying more and getting less (Intel).

It's been absolutely amazing how quickly, and how broadly AMD has been
able to pry open the doors to OEM's, since they launched their lawsuit
against Intel. This was just the latest door to be busted down.

Incredibly, analysts are saying that this is nothing to get excited
about, because this news is already "built into AMD's stock price"! I
wonder how this news could already be built into the stock price, when
Dell didn't even announce that it was becoming an AMD customer till
yesterday. Also prior to yesterday, all of the smart money was saying
that Dell wouldn't ever buy a thing from AMD. So I ask, how could this
news be *already built in*?

Also they're trying to pour cold water over the announcement by saying
that there isn't likely to be much volume on this product. I mean, when
Apple announced that they were going Intel, it was God talking to Moses
time. It was pointed out that Apple would add a miniscule amount of
revenue to Intel. But analysts said, "but think of the future potential
of this Apple-Intel deal?" Double standards? You bet.

Yousuf Khan
 
Dell is simply sowing what it reaped...when the ink dried on the Intel
exclusive deal, Dell committed itself to a long protracted death.

You know if 14B/quarter in revenue is a slow death...I'd like 3.
Smart
money would jump on the AMD bandwagon, and I would love to buy a Dell. But
as long as Dell is pandering to last generation technology in its consumer
desktop, they'll not get my business.

I'm sure their sales reps are crying themselves to sleep right now
about that.
HP and Gateway are getting the dollars I would have spent on Dell if they
had AMD available.

OK.

Dell is only using AMD for the segment where they are at a competitive
disadvantage, which is larger servers. That should tell you
something...

DK
 
It's been absolutely amazing how quickly, and how broadly AMD has been
able to pry open the doors to OEM's, since they launched their lawsuit
against Intel. This was just the latest door to be busted down.

What other OEMs recently adopted AMD?
Incredibly, analysts are saying that this is nothing to get excited
about, because this news is already "built into AMD's stock price"! I
wonder how this news could already be built into the stock price, when
Dell didn't even announce that it was becoming an AMD customer till
yesterday.

Easy, some folks knew without having to read NY Times...
Also prior to yesterday, all of the smart money was saying
that Dell wouldn't ever buy a thing from AMD. So I ask, how could this
news be *already built in*?

Actually Dell has sold AMD products for quite a while, but they only
sold components, not entire systems. There's a rather substantial
difference.

I think the analyst is saying that according to their FCF models (or
whatever models they use), that the stock price already reflects the
earnings when you account for what they are likely to sell to Dell.
Also they're trying to pour cold water over the announcement by saying
that there isn't likely to be much volume on this product.

That's not so much poo-pooing as being tuned into reality.
I mean, when
Apple announced that they were going Intel, it was God talking to Moses
time. It was pointed out that Apple would add a miniscule amount of
revenue to Intel. But analysts said, "but think of the future potential
of this Apple-Intel deal?" Double standards? You bet.

I suspect Apple sells more chips than Dell 4S systems do. However, the
real interesting part (which you didn't quite mention) is that this may
be a segway into Dell using AMD in other product lines.

I don't think that they will use AMD for 2S or 1S servers, or laptops,
but perhaps on the desktop. Of course, if things change and Intel
screws up...who knows.

DK
 
And to add insult to injury, Dell is going to use AMD not in $299
desktops or even pizzabox 1-2p servers, but in top of the line 4p+
servers. I bet Opterons 8xx have the widest margin of all AMD lineup,
whereas Celerons (the $299 box stuff) have the margin close to 0,
possibly even on negative side ;-)
And after this psychological barrier is broken, it will be even easier
for Dell to add AMD to its 2p server and XPS lineup later on.

In a sense though, it really only made sense to start with the
4-processor servers. That's where AMD is absolutely destroying Intel
in terms of performance and that is where customer's are demanding
AMD-based systems. According to the following article AMD actually is
taking over 40% of the 4p x86 server sales now:

http://news.com.com/Dell+opts+for+AMDs+Opteron/2100-1006_3-6074059.html?tag=nefd.pop

Considering that they only have a 26% share of the overall server chip
market they are doing disproportionately well with 4P servers.

Ironically this comes just a few weeks/months before Intel is set to
release their new "Woodcrest" server chips, their most competitive
offerings since the Opteron was released 3 years ago. I still suspect
that Opteron will hold the edge in 4p+ servers due to memory bandwidth
restrictions on Intel's side of things, but it should be a MUCH closer
battle then it is now.
 
Also they're trying to pour cold water over the announcement by saying
that there isn't likely to be much volume on this product.

They're correct that it isn't going to be much volume, but it's
definitely HIGH profit volume. They are only going to be buying AMD
800-series Opterons, many of which sell for well over $1000. Even
with the rather generous discounts that AMD is probably offering Dell
they are still likely to making several hundred (thousand?) percent
margins on these chips.
I mean, when
Apple announced that they were going Intel, it was God talking to Moses
time. It was pointed out that Apple would add a miniscule amount of
revenue to Intel. But analysts said, "but think of the future potential
of this Apple-Intel deal?" Double standards? You bet.

I tend to agree with you here. The Apple-Intel deal is going to
result in larger volume than the AMD-Dell deal, but it's lower profit
volume. Even though Apple is using some fairly high-profit chips,
they aren't the same sort of 4p server chips that AMD is selling to
Dell.

Overall the Intel-Apple deal will probably bring in more profit for
Intel, given that they'll probably sell about 5 million chips next
year vs. probably something like 25,000 to 50,000 Opteron chips that
AMD will sell to Dell. However if you look at it relative to company
size (ie $/share, what you would expect analysts to look at) the
figures are probably pretty comparable.

As for that whole deal about the price being built into the stock,
that's just the sort of nonsense that makes me more than a bit
skeptical of analysts. Either they're full of it or they're getting
some sort of insider trading info, because as you said, NO ONE
expected this deal until it was announced.
 
As for that whole deal about the price being built into the stock,
that's just the sort of nonsense that makes me more than a bit
skeptical of analysts. Either they're full of it or they're getting
some sort of insider trading info, because as you said, NO ONE
expected this deal until it was announced.

Hmm how about saying it so that prices won't jump up too much until
them analyst got their hands on a few lots of AMD shares first? :ppPp
 
What other OEMs recently adopted AMD?

Lenovo made announcements a short while back - there were also various
mobile/laptop announcements: just this week, Benq, NEC & Fujitsu
(Fujitsu-Siemens has had AMD systems in Europe but in U.S. it's been
on-and-off) as well as the already established Acer, Gateway & HP. As for
large OEMs, which of them is not already using AMD CPUs? It's only a
question of getting into more new product lines *and* getting into
brochures.. which just happened with HP this week; previous HP product
brochures targeted at businesses did not mention their AMD systems.

The landscape *is* changing... and the lawsuit could well be a motivator.
Easy, some folks knew without having to read NY Times...


Actually Dell has sold AMD products for quite a while, but they only
sold components, not entire systems. There's a rather substantial
difference.

Uhh, no - it was rather well known that Dell would sell even complete AMD64
server systems if pressed by established "valuable" customers... which
apparently must have happened.
I think the analyst is saying that according to their FCF models (or
whatever models they use), that the stock price already reflects the
earnings when you account for what they are likely to sell to Dell.

Come on this is like GW models - they give emm, "desired results".. and if
they don't, well... the data is malleable.;-) A wet finger is about as
useful. The fact is that AMD gets no respect from the anal...ysts...
because they know it's more difficult to provoke a feeding frenzy with the
"#2" player.
That's not so much poo-pooing as being tuned into reality.


I suspect Apple sells more chips than Dell 4S systems do. However, the
real interesting part (which you didn't quite mention) is that this may
be a segway into Dell using AMD in other product lines.

Hmm, from the guy who took someone to task for err, spelling recently?
Would it be piddling to point out the difference in ASP between CPUs Apple
sells and those in 4S systems?
I don't think that they will use AMD for 2S or 1S servers, or laptops,
but perhaps on the desktop. Of course, if things change and Intel
screws up...who knows.

"think"?:-) From my POV, I "think" it's more likely that what happened
here is that Dell has hi-jacked a planned system from Alienware's future
product line... to bolster its top-end server line-up... "running a flag
up" as it were. It'll be easy enough to do the same with 1S or 2S systems
as the "need" arises... or the wind blows.:-) There's also the fact that
Chartered is just ramping up right now so it'll take a few weeks/months to
see if the supply across the board is going to meet what would be Dell's
expectation on volume on low/mid-range prods.

As well as being a bit of a slap in the face for Intel: "AMD CPUs are
good/better(?) for the top-end of the market??":-P... it's also a kick in
the ass to Intel from Mikey himself, given his current Q's performance.
 
Overall the Intel-Apple deal will probably bring in more profit for
Intel, given that they'll probably sell about 5 million chips next
year vs. probably something like 25,000 to 50,000 Opteron chips that
AMD will sell to Dell. However if you look at it relative to company
size (ie $/share, what you would expect analysts to look at) the
figures are probably pretty comparable.

Multiply that by 4 - nobody in their right mind will order a 4s server
with just one socket filled, or even 2. Or you already accounted for
that?

NNN
 
Hmm how about saying it so that prices won't jump up too much until
them analyst got their hands on a few lots of AMD shares first? :ppPp

I dunno what the rules are on that but most reputable anal...yst articles
have a statement as to whether they own any of the stock in question.
 
Tony said:
They're correct that it isn't going to be much volume, but it's
definitely HIGH profit volume. They are only going to be buying AMD
800-series Opterons, many of which sell for well over $1000. Even
with the rather generous discounts that AMD is probably offering Dell
they are still likely to making several hundred (thousand?) percent
margins on these chips.

Yeah, this entire event has AMD entering from the position of power.
It's one thing if AMD had entered the halls of Dell with a bunch of
low-priced Semprons, but it's quite another thing when it's entering
Dell with their highest-end product.

Also, another reason why AMD is entering from the power position is
because AMD just saved Dell's stock price. If Dell hadn't announced that
it was going with AMD yesterday, then it would've been slaughtered on
the market. There was no way in hell that it's stock price could've
possibly gone up after that earnings report it posted. It was just being
made to look even worse because HP posted stellar earnings a couple of
days earlier, and people would've said, "why is HP thriving now, and
you're not?" So it turned out all of Dell's customers screaming in its
ears about going with Opteron couldn't sway it, but as soon as Dell's
stock price was jeopardized, along comes Opteron!

Dell takes care of Wall Street, not customers with AMD move | The Register
http://www.theregister.co.uk/2006/05/19/dell_opteron_wallstreet/
As for that whole deal about the price being built into the stock,
that's just the sort of nonsense that makes me more than a bit
skeptical of analysts. Either they're full of it or they're getting
some sort of insider trading info, because as you said, NO ONE
expected this deal until it was announced.

Quite obviously the price wasn't built into AMD's stock price yet at
that point yesterday because AMD was up over 10% in that one day. You
tend to think after awhile that analysts are just hired for their
running mouths.

Yousuf Khan
 
Quite obviously the price wasn't built into AMD's stock price yet at
that point yesterday because AMD was up over 10% in that one day. You
tend to think after awhile that analysts are just hired for their
running mouths.

Yousuf Khan

If only this news came just a few months back, AMD share price
would've shot above 50, probably even 60 would not be that much of a
stretch - it was 42 even without Dell. But these days it seems only
to slow down the slide... I still believe that AMD will be again
above 40 by this time next year, but chances are we'll see it below 30
before this year ends. It's not exactly consoling that INTC is at
52wk low and still sliding. Yet it seems to me that INTC might be a
good buy just before Core2 launch - I couldn't even think of buying
anything Intel for years, but this time around investing in it might
make sense.

NNN
 
I dunno what the rules are on that but most reputable anal...yst articles
have a statement as to whether they own any of the stock in question.

Hmm, would that matter in this case? Say I'm an analyst who didn't buy
into AMD, so I write an article pooing this new development. At the
time of writing I have no stock interest in AMD. Right after
submitting the article, I buy 10,000 shares (I think that's some
$310,000 around the release of the news). Prices say shot up around
12% because nobody believes the crap I would had been sprouting, I
sell off my AMD shares, make a quick $37,000 in just a couple of days.
THEN write some article saying how this price rise is just a market
kneejerk reaction by the ignorant blah blah and prices will drop, so
recommend a sell. At the bottom, I can still declare at the time of
writing I have no stock interest in AMD :P

And if the prices down after that, I can start crowing about how
accurate my predictions are :pPP

And probably do the same thing just before Intel's NDA on Conroe is
lifted :P
 
Multiply that by 4 - nobody in their right mind will order a 4s server
with just one socket filled, or even 2. Or you already accounted for
that?

I was already counting that, though it's obviously a VERY rough
approximation, given that no one really knows what percentage of
Dell's 4-processor servers will be these new AMD-based systems. It
could very easily be 4 times as much.
 
What other OEMs recently adopted AMD?
Lenovo made announcements a short while back - there were also various
mobile/laptop announcements: just this week, Benq, NEC & Fujitsu
(Fujitsu-Siemens has had AMD systems in Europe but in U.S. it's been
on-and-off) as well as the already established Acer, Gateway & HP.

IIRC, Fujitsu used AMD systems before, as did NEC. I don't know about
Benq...
As for
large OEMs, which of them is not already using AMD CPUs? It's only a
question of getting into more new product lines *and* getting into
brochures.. which just happened with HP this week; previous HP product
brochures targeted at businesses did not mention their AMD systems.

The landscape *is* changing... and the lawsuit could well be a motivator. [snip]

Uhh, no - it was rather well known that Dell would sell even complete AMD64
server systems if pressed by established "valuable" customers... which
apparently must have happened.

Interesting, can you prove this? I have never heard of this happening.
Come on this is like GW models - they give emm, "desired results".. and if
they don't, well... the data is malleable.;-)

Data isn't malleable, your assumptions are. You should try building
these models before you knock them. They are a lot more complex than
they seem. I think anyone here could build a simple model for a
company that only has USD denominated sales, with no debt, and is
otherwise relatively simple. Modelling AMD would be a bloody
nightmare.
A wet finger is about as
useful. The fact is that AMD gets no respect from the anal...ysts...
because they know it's more difficult to provoke a feeding frenzy with the
"#2" player.

I don't think so. There are plenty of analysts who cannot make a dime
(and their companies cannot make a dime) based on what they say, even
if they do move the market.
Hmm, from the guy who took someone to task for err, spelling recently?
Would it be piddling to point out the difference in ASP between CPUs Apple
sells and those in 4S systems?

I made a single spelling mistake. The gentlemen I chided made quite a
few more.

Of course it's not piddling...but we should probably wait to see how
much volume Dell gets on those systems.
"think"?:-) From my POV, I "think" it's more likely that what happened
here is that Dell has hi-jacked a planned system from Alienware's future
product line...

Are you joking? Alienware barely has a presence in the server market,
and they sure don't offer any 4P systems.
to bolster its top-end server line-up... "running a flag
up" as it were. It'll be easy enough to do the same with 1S or 2S systems
as the "need" arises... or the wind blows.:-)

I doubt it will be necessary, since Intel is much quicker to act with
the DP servers, since validation times aren't as long.
There's also the fact that
Chartered is just ramping up right now so it'll take a few weeks/months to
see if the supply across the board is going to meet what would be Dell's
expectation on volume on low/mid-range prods.

Why would dell bother? Woodcrest will beat K8 hands down on
performance and on power.
As well as being a bit of a slap in the face for Intel: "AMD CPUs are
good/better(?) for the top-end of the market??":-P... it's also a kick in
the ass to Intel from Mikey himself, given his current Q's performance.

Of course, the real question is whether Intel will kick back. I
suspect that if Intel were to stop giving rebates to Dell their next
quarter would look a hell of a lot worse than this one.

DK
 
Hmm, would that matter in this case? Say I'm an analyst who didn't buy
into AMD, so I write an article pooing this new development. At the
time of writing I have no stock interest in AMD. Right after
submitting the article, I buy 10,000 shares (I think that's some
$310,000 around the release of the news). Prices say shot up around
12% because nobody believes the crap I would had been sprouting, I
sell off my AMD shares, make a quick $37,000 in just a couple of days.
THEN write some article saying how this price rise is just a market
kneejerk reaction by the ignorant blah blah and prices will drop, so
recommend a sell. At the bottom, I can still declare at the time of
writing I have no stock interest in AMD :P

Like I said, there may be rules here to do with privileged position related
to insider trading.

Funny story which was claimed to be true, 1/2(?) years ago: The SEC
(Securities & Exchange Commission) collared some guy who had been making a
bundle of $$ on short-term trading... enough to make them suspicious...
buying just before some major stock up-swings and selling shortly after.
He'd done it several times and IIRC his total take was ~$2M!! He claimed
he was a time-traveller who was marooned in 2004(?) because his time
machine was broken. The reason he was able to predict the prices was
because he'd charted stock swing histories *before* coming back "here". He
was "sorry" about any apparent illegal behavior but he badly needed the
money to buy parts to fix his machine, so that he could go "home" again.
And if the prices down after that, I can start crowing about how
accurate my predictions are :pPP

And probably do the same thing just before Intel's NDA on Conroe is
lifted :P

<shrug>... how it works! Why do I call them anal...ysts?:-)
 
Hmmm, that took a while.:-)
IIRC, Fujitsu used AMD systems before, as did NEC. I don't know about
Benq...

I *said* what Fujitsu did before... NFM; I don't recall being able to buy
AMD64 systems from NEC and ATM I don't think they sell notebooks in the
U.S. at all.
As for
large OEMs, which of them is not already using AMD CPUs? It's only a
question of getting into more new product lines *and* getting into
brochures.. which just happened with HP this week; previous HP product
brochures targeted at businesses did not mention their AMD systems.

The landscape *is* changing... and the lawsuit could well be a motivator. [snip]

Uhh, no - it was rather well known that Dell would sell even complete AMD64
server systems if pressed by established "valuable" customers... which
apparently must have happened.

Interesting, can you prove this? I have never heard of this happening.

It was talked about here - it was convincing... I believed it; you can
choose to call people liars if you wish.
Data isn't malleable, your assumptions are. You should try building
these models before you knock them. They are a lot more complex than
they seem. I think anyone here could build a simple model for a
company that only has USD denominated sales, with no debt, and is
otherwise relatively simple. Modelling AMD would be a bloody
nightmare.

I *have* built "models" for a living - I'm well aware of the workings &
fungibility of numerical models. Whether you believe it or not, data *is*
malleable - to reiterate on GW, Remote Sensing Systems are experts at
"calibration" of data for those poxy models. Call it calibration,
normalisation or whatever, it happens.
I don't think so. There are plenty of analysts who cannot make a dime
(and their companies cannot make a dime) based on what they say, even
if they do move the market.


I made a single spelling mistake. The gentlemen I chided made quite a
few more.

Umm, a little more than a spelling mistake.
Of course it's not piddling...but we should probably wait to see how
much volume Dell gets on those systems.

Demand for AMD servers *is* on the rise; the demand for 4S systems will
rise as well... based on my 1st hand experiences of what is going on in
corporate computing just now: Sarbanes-Oxley is reshaping the whole
approach to storage & handling of data and its security in the U.S.; in
other countries there are similar driving forces. I fear we're about to
re-enter the "glass-house" - I don't like it but it *is* happening.
Are you joking? Alienware barely has a presence in the server market,
and they sure don't offer any 4P systems.

What do you not understand about "think" and "future"? Both SuperMicro and
Tyan already have Opteron 8xx 4S mbrds available off the shelf - there's no
big deal here.
I doubt it will be necessary, since Intel is much quicker to act with
the DP servers, since validation times aren't as long.


Why would dell bother? Woodcrest will beat K8 hands down on
performance and on power.

I expect it'll win by some amount on raw performance in 1P & 2P; power I'm
not so sure and price is going to be an issue with Intel's expensive
chipset solutions - it depends on what you want to measure... but all of
that that has yet to be revealed.
Of course, the real question is whether Intel will kick back. I
suspect that if Intel were to stop giving rebates to Dell their next
quarter would look a hell of a lot worse than this one.

It's a gamble I guess. Why would Intel whack their main outlet? Given
recent inventory history, I'd think they've issued kid gloves to the sales
teams.:-)
 
Back
Top