Interesting read about upcoming K9 processors

  • Thread starter Thread starter Yousuf Khan
  • Start date Start date
|>
|> Definitely! At least publicly Intel is STRONGLY saying that IA64 is
|> the one true path for the future, customers be damned!

If Intel even let hint that they were having second thoughts, IA64
would be dead overnight except in HP and SGI "big iron". Somewhat
surprisingly, I think that SGI would weather that pretty well, but
HP would be in dead trouble.

My belief is that Intel are trying to deny the failure of IA64 until
they get their new range of designs onstream (planned for 2007).

|> >What kind of stranglehold could Intel have over Microsoft anyway? Did
|> >they promise to stop supporting Linux or something? I can't really
|> >find a reasonable conspiracy theory here.
|>
|> I don't think any conspiracy theory is needed...

It may not be needed, but there is enough evidence of conspiracies
to safely assume that there are several. Whether any of them have
anything to do with the matter in hand is another matter. Quite
likely not, as you say.

|> >PS: Any recent figures for "datacenter" IA64 vs Opteron servers, and the
|> >percentage running Windows 64?
|>
|> I suspect that the figures are probably measured in single-digit units
|> for both, so they probably aren't all that meaningful. IA64 seems to
|> be either small (1-4P servers) or big iron from HP (running HP-UX or
|> Linux) or SGI (running Linux). Opterons, on the other hand, are all
|> 1-4P servers and therefore wouldn't really fall into the "datacenter"
|> category.. at least assuming you define "datacenter" in a similar way
|> to how MS defines it.

With that definition, agreed. I believe that you can buy some in
Japan, but they are probably only being used for development and
testing at present.


Regards,
Nick Maclaren.
 
Nick said:
|>
|> Definitely! At least publicly Intel is STRONGLY saying that IA64 is
|> the one true path for the future, customers be damned!

If Intel even let hint that they were having second thoughts, IA64
would be dead overnight except in HP and SGI "big iron". Somewhat
surprisingly, I think that SGI would weather that pretty well, but
HP would be in dead trouble.
Click to expand...

I think Intel could cut off the flow of Itanics right now and
HP would hardly notice. HP has had a couple of years now to
recognize that Itanic is never going to be anything more than
a niche market. They are not all idiots over there - they
began dealing with the fact that they will never come remotely
close to recovering their Itanic investment a long time ago.
 
I think Intel could cut off the flow of Itanics right now and
HP would hardly notice. HP has had a couple of years now to
recognize that Itanic is never going to be anything more than
a niche market. They are not all idiots over there - they
began dealing with the fact that they will never come remotely
close to recovering their Itanic investment a long time ago.
Click to expand...

I agree that the technical staff are not idiots, but I have heard
reports about the reaction of executives to proposals of backup
strategies. Take a look at their accounts, and remember that VMS
and NonStop (or whatever it is now) are secondary only to HP-UX
in their service provision.

And then think about how customers would react if told that the
ONLY future platforms for HP-UX, VMS and NonStop had just been
cancelled. Overjoyed isn't the word I would use ....


Regards,
Nick Maclaren.
 
|>
|> Definitely! At least publicly Intel is STRONGLY saying that IA64 is
|> the one true path for the future, customers be damned!

If Intel even let hint that they were having second thoughts, IA64
would be dead overnight except in HP and SGI "big iron".
Click to expand...

I'm not sure that would be very different from where they stand now.
From what I understand HP and SGI make up well over 90% of all Itanium
server revenues. Everyone else (IBM, Dell, Unisys, Bull and whoever
else) are mostly fighting for a few scraps.
Somewhat
surprisingly, I think that SGI would weather that pretty well, but
HP would be in dead trouble.
Click to expand...

I'd say that it's completely the opposite. SGI has absolutely nothing
else but Itanium going for it. Their old MIPS sales are all but dried
up and the revenue their getting from service contracts is *rapidly*
vanishing.

HP, on the other hand, could easily afford to just ditch their
Integrity group altogether and keep going with the rest of their
businesses. It might hurt a bit, but all their Itanium servers
combined only make up a relatively small portion of the companies
total market. Their Xeon-based Proliant servers still bring in a LOT
more revenue.
My belief is that Intel are trying to deny the failure of IA64 until
they get their new range of designs onstream (planned for 2007).
Click to expand...

I doubt that they'll ever admit failure... did they ever admit that
the i860 never came remotely close to the lofty expectations for it?
What was supposed to revolutionize the entire processor industry ended
up as little more than a co-processor for storage and networking
devices?

Nahh, the Itanium will quietly be swept under the rug as a mainstream
chip and it will remain a niche product to power SGI and HP's
big-iron.
 
That's the whole point: that backward compatibility wasn't promised by
Intel, but Microsoft's market position and the market's de facto demand
for backward compatibility at all costs was so strong that a minor
implementation detail of a new chip needed Microsoft's go-ahead.
Click to expand...

Then Intel doesn't truely care about backwards compatability (which was
my point).
If Microsoft is good at anything, it's marketing. And that's why they
now care about security.
Click to expand...

Are you really buying that crap?
 
Sure enough.


Nonsense. That would only be the consequence if the market cared about
"designed for Windows". Nobody - at least hereabouts - does in the consumer
market, at least.
Click to expand...

Then you really don't have a clue about this market. These logos and
the ability to claim compatability mean *everything* to the market. When
was the last time you saw a HPaQ/Dell/IBM/whatever (non-whitebox) that
didn't have a "designed for Win" logo.
So that's why ISA devices are still somewhat supported, and all that
other legacy crap? How big is the success of the legacy-free PC
Microsoft and Intel tried to push?
Click to expand...

ISA isn't supported anymore. Try buying a recent motherboard with any ISA
slots. It's about time it died and it took M$ to put the stake through
its heart (the only good thing they've done, IMO).
Let's just agree that we live in different universes.
Click to expand...

You *certainly* do! Yikes!
 
Tony Hill said:
On Tue, 03 Aug 2004 08:56:53 +0200, Ketil Malde <[email protected]>

Definitely! At least publicly Intel is STRONGLY saying that IA64 is
the one true path for the future, customers be damned!
Click to expand...

They have publicly stated that x86 is here for a long time. Show me one
public statement anywhere from Intel, anywhere, that says IA64 is the only
future processor, whether strong or weak.

OTOH, virtually all OEMs but one are offering IA64 systems, and that one is
'looking at it'. As far as dollars, I believe that IA64 systems accounted
for a significantly larger amount of system revenues than Opteron. That
may change, but to imply that customers don't want IA64 is disingenuous at
best...
Now you've got it!

Actually it all seems to tie back in to the fact that MS decided to
push all their future OSes back until they get WinXP SP2 out, and that
seems to be taking forever! Each time SP2 gets pushed back everything
else gets pushed back behind it.
Click to expand...

The explanation is most likely that the Windows OS is likely a huge pile of
spaghetti code that is a nightmare to maintain - including full 64-bit
operation. For those who have never worked on a commercial product of any
size, all it takes is a few customers complaining about a bug that 95% will
never encounter to extend a beta - and that 95% will scratch their heads and
claim that the product is *so* stable. Sure, you can get the 'core'
features to work fine, but the corner cases can be a major bitch... :-).

Regards,
Dean
 
Rob Stow said:
I think Intel could cut off the flow of Itanics right now and
HP would hardly notice. HP has had a couple of years now to
recognize that Itanic is never going to be anything more than
a niche market. They are not all idiots over there - they
began dealing with the fact that they will never come remotely
close to recovering their Itanic investment a long time ago.
Click to expand...

For those in .chips, since the reference was already made - this sounds an
awful lot like a certain individual claiming "DDR is dead, Dead, DEAD".

Some people should look in the mirror more often. :-).

Regards,
Dean
 
Tony Hill said:
I'm not sure that would be very different from where they stand now.
From what I understand HP and SGI make up well over 90% of all Itanium
server revenues. Everyone else (IBM, Dell, Unisys, Bull and whoever
else) are mostly fighting for a few scraps.
Click to expand...

That, by itself, is not an indication of the success or failure of Itanium.
First you have to provide the revenue numbers, and market percentage.
Otherwise you could state that because Intel gets almost 85% of all x86
revenues, that x86 is a failure...

Regards,
Dean
 
Dean Kent said:
OTOH, virtually all OEMs but one are offering IA64 systems, and that one is
'looking at it'. As far as dollars, I believe that IA64 systems accounted
for a significantly larger amount of system revenues than Opteron.
Click to expand...

Now that I think about it, I believe Itanium may have outshipped Operton in
units for all of 2003 as well. Intel claimed over 100,000 units shipped
(most at the end of the year), but I don't think AMD disclosed the Opteron
numbers. If they had outshipped Intel, I would think that they would have
made a comment to that effect, since JSIII had all but guaranteed it at the
beginning of the year...

Regards,
Dean
 
Dean said:
For those in .chips, since the reference was already made - this sounds an
awful lot like a certain individual claiming "DDR is dead, Dead, DEAD".
Click to expand...

I think the similarity is a *lot* stronger to the
"Rambus is dead, Dead, DEAD." that far more people
subscribed to.
 
The first Google hit on "opteron shipments" is xbitlabs:

| A report over InformationWeek web-site cites analyst Dean McCarron for
| Mercury Research who claims that AMD supplied about 70 000 of AMD
| Opteron microprocessors in the first quarter this year. By contrast,
| the Sunnyvale, California-based microprocessor maker supplied about 65
| 000 of its server microprocessors in 2003. According to some other
| estimates, AMD only sold 40 000 of AMD Opteron products last year.

So if the 100000 units figure is correct for Itanium, you seem to be
right. What was the "end of year" thing, though? It was hardly
Christmas shoppers, was it?

While (of course) Intel never committed to terminating x86, it
is clear that they wanted IA64 to be vastly more mainstream than it
seems to be. That, or the press were victim of the greatest
mass-misunderstanding I've seen. Randomly Googling around brings me
to e.g. http://www.dqindia.com/content/top_stories/102041601.asp, with a
nice chart showing an estimated 1.5-2 million IPF server shipments in
2003 -- and the article is dated early 2002. FWIW, I don't think it's
likely that they will exceed IA32 servers in 2005 either, but we'll
see.

-kzm
 
Then Intel doesn't truely care about backwards compatability (which was
my point).
Click to expand...

Humbug. They don't want to be held back by _accidental_ backward
compatibility, which is a big difference.
Are you really buying that crap?
Click to expand...

MS's marketing? No. But I can read about the effects of their recent
patches, not only to the base operating system.

Jan
 
HP, on the other hand, could easily afford to just ditch their
Integrity group altogether and keep going with the rest of their
businesses. It might hurt a bit, but all their Itanium servers
combined only make up a relatively small portion of the companies
total market. Their Xeon-based Proliant servers still bring in a LOT
more revenue.
Click to expand...

Who cares about revenue? It's EBIT or one of its variations that's relevant,
and EBIT/revenue is perhaps even more important. I'd rather have a company
with a tenth of the revenue and solid earnings than a company with high
revenue and bleeding red ink, wouldn't you?

Jan
 
Rob Stow said:
I think the similarity is a *lot* stronger to the
"Rambus is dead, Dead, DEAD." that far more people
subscribed to.
Click to expand...

Of course you do. Just as Corse did. Makes perfect sense.

I'll say no more on the subject, because it is just as emotional and just as
stupid as that argument was. I can see that the .chips crowd hasn't
changed a bit - how sad.

Regards,
Dean
 
Ketil Malde said:
The first Google hit on "opteron shipments" is xbitlabs:

| A report over InformationWeek web-site cites analyst Dean McCarron for
| Mercury Research who claims that AMD supplied about 70 000 of AMD
| Opteron microprocessors in the first quarter this year. By contrast,
| the Sunnyvale, California-based microprocessor maker supplied about 65
| 000 of its server microprocessors in 2003. According to some other
| estimates, AMD only sold 40 000 of AMD Opteron products last year.

So if the 100000 units figure is correct for Itanium, you seem to be
right. What was the "end of year" thing, though? It was hardly
Christmas shoppers, was it?
Click to expand...

No, just that Intel stated that the shipments were 'back end loaded',
meaning that most of the shipments had occurred at the end of the year (I
believe that came from the earnings conference call). I have absolutely no
idea what it means, but there had been much noise made earlier in the year
about how Opteron would outsell Itanium in Opteron's first year of sales
because the first couple of quarters were pretty dismal for Itanium.
While (of course) Intel never committed to terminating x86, it
is clear that they wanted IA64 to be vastly more mainstream than it
seems to be. That, or the press were victim of the greatest
mass-misunderstanding I've seen. Randomly Googling around brings me
to e.g. http://www.dqindia.com/content/top_stories/102041601.asp, with a
nice chart showing an estimated 1.5-2 million IPF server shipments in
2003 -- and the article is dated early 2002. FWIW, I don't think it's
likely that they will exceed IA32 servers in 2005 either, but we'll
see.
Click to expand...

I believe that Itanium is far less ubiquitous than Intel desired, but the
evidence seems to indicate it is far less than the total disaster that some
wish to spin. In fact, as the technologies mature (both hardware and
software), it could be argued that the momentum is starting to build.

Now, I know that there are some who claim that Intel wanted IA64 to replace
x86 very early on, but...

1) The NY Times quoted Andy Grove in 1998: "I don't see Merced appearing on
a mainstream desktop inside of a decade." -ANDY GROVE, Ex-CEO, Intel (New
York Times, 5 April 98)

2) In 1997, the now infamous Bob Colwell was scheduled to give a talk at the
Microprocessor Forum about the IA32 enhancements "beyond the end of the
decade", so it was obviously not the intent at that time to replace IA32
anytime really soon:
http://www.intel.com/pressroom/archive/releases/sp100997.HTM

3) Gordon Moore, in late 1996/early 1997, in an interview with PC Magazine
stated: "GORDON MOORE: Oh yeah, sure, 64 bits means new instructions. But it
will still run the older software compatibly. You know, that's one thing we
have, is the idea of carrying a compatible family along--even if we have to
put two processors on the chip, one 32-bit and one 64-bit, it's going to run
that old software effectively."
http://www.pcmag.com/article2/0,1759,35750,00.asp

So, it would appear that even in 1996 the concept was not to eliminate x86
entirely, in 1997 it was publicly stated that IA32 would be around for some
time after Y2K, and in 1998 it was publicly stated that IA64 would not be on
the desktop for at least another 3.5 years from *today*. This despite the
recollections of a few who are certain that Intel had more nefarious plans
early on...

Regards,
Dean
 
|> >In article <[email protected]>,
|> >|>
|> >|> Definitely! At least publicly Intel is STRONGLY saying that IA64 is
|> >|> the one true path for the future, customers be damned!
|> >
|> >If Intel even let hint that they were having second thoughts, IA64
|> >would be dead overnight except in HP and SGI "big iron".
|>
|> I'm not sure that would be very different from where they stand now.
|> From what I understand HP and SGI make up well over 90% of all Itanium
|> server revenues. Everyone else (IBM, Dell, Unisys, Bull and whoever
|> else) are mostly fighting for a few scraps.

No. "Dead" as in Norwegian Blue parrots, not merely half-dead as
at present.

|> > Somewhat
|> >surprisingly, I think that SGI would weather that pretty well, but
|> >HP would be in dead trouble.
|>
|> I'd say that it's completely the opposite. SGI has absolutely nothing
|> else but Itanium going for it. Their old MIPS sales are all but dried
|> up and the revenue their getting from service contracts is *rapidly*
|> vanishing.

Er, no. Look at their financials and customers. The former are
now balanced, and the latter have retrenched to the 'technical'
marketplace. And we are a damn sight less worried about which
CPUs we use than the 'commercial' market, as we are accustomed to
rebuilding codes on a few year timescale. Provided that Intel
still produces IA64 systems and delivers boring enhancements
(shrinks, clock rate, caches etc.), SGI is OK for now.

|> HP, on the other hand, could easily afford to just ditch their
|> Integrity group altogether and keep going with the rest of their
|> businesses. It might hurt a bit, but all their Itanium servers
|> combined only make up a relatively small portion of the companies
|> total market. Their Xeon-based Proliant servers still bring in a LOT
|> more revenue.

Look at their margins. They are zilch in that division, and there
is little chance of a change. They are in printing and services,
and the latter will disappear if all of the SuperDome, VMS and
Integrity products do so. And all of those are currently dependent
on IA64.

|> >My belief is that Intel are trying to deny the failure of IA64 until
|> >they get their new range of designs onstream (planned for 2007).
|>
|> I doubt that they'll ever admit failure... did they ever admit that
|> the i860 never came remotely close to the lofty expectations for it?
|> What was supposed to revolutionize the entire processor industry ended
|> up as little more than a co-processor for storage and networking
|> devices?

Internally. I agree that they will never admit failure in public.


Regards,
Nick Maclaren.
 
Keith said:
Then you really don't have a clue about this market.
These logos and the ability to claim compatability mean
*everything* to the market.
Click to expand...

Am I part of that market?

Speaking as a customer, I don't care whether it has some silly logo
on it. I *know*, from bitter experience, that I can buy graphics cards
from the likes of ATI or nVidea, put them on a Win2K or XP system,
and be unable to boot up because the driver faults.

OK, I'm a developer so I'm using the debug kernel, but so should they
be -- where "they" are both "whoever wrote the driver" and "whoever
gave it the shiny logo". The fact that *clearly* *neither* have tried
it on a sufficiently wide range of PCs (I expect it is motherboard-
dependent, or something) tells me all I need to know about the Device
Driver Tax.

Off the top of my head, I know of annoying but less serious problems
with Creative sound cards and just about anything from HP. (The
latter appear to sub-contract their driver writing, which may be
relevant.) Presumably this "certified" hardware does work on the
vast majority of systems. However, so does non-certified hardware.
The logo isn't giving us anything new.

I'm also using the latest service pack, which may not have been around
at the time of certification, but whilst that may excuse the vendor
and WHQL, it hardly praises the team at MS who produce service packs
and it does point to a fundamental flaw in the certification process.
When was the last time you saw a HPaQ/Dell/IBM/whatever
(non-whitebox) that didn't have a "designed for Win" logo.
Click to expand...

The logo costs a few thousand (small beer for a major manufacturer)
so yes, it's a while since I saw a vendor of mainstream hardware who
wasn't displaying the logo. It probably comes out of the marketing
budget and is a relatively good return on investment in that sense.

For more obscure hardware, the logo compliance would only cover things
like "doesn't prevent booting up or hibernation" and "superficially
appears to let the demo app access the device without blue screening".
There's little point in getting the logo and for the (usually small)
companies concerned, the few thousand quid just isn't worth it. The
customers' purchasing decisions are based on very different criteria.
 
I wouldn't really call it fighting; more like having something
available, just in case anybody would want it. The numbers seem to
indicate few people do. If you want to compare market share, you
should compare HP IA64s to IBM Powers and Sun Sparcs.
Er, no. Look at their financials and customers. The former are
now balanced, and the latter have retrenched to the 'technical'
marketplace. And we are a damn sight less worried about which
CPUs
Click to expand...

I think most of HP's IA64 sales also are going to the scientific
computing segment. And while the commercial customers may be more
worried, it depends on what HP replaces IA64 with; I'm not sure people
would object strongly to Superdomes with Opterons or 64-bit Xeons, for
instance.

-k
 
|>
|> I think most of HP's IA64 sales also are going to the scientific
|> computing segment. And while the commercial customers may be more
|> worried, it depends on what HP replaces IA64 with; I'm not sure people
|> would object strongly to Superdomes with Opterons or 64-bit Xeons, for
|> instance.

If your first sentence is true, HP's IA64 lines are in worse trouble
than even I thought they were.

No, I don't think people would object, but that's not my point.
It is the time to revalidate and the loss of confidence - both are
more of an issue in the commercial arena.


Regards,
Nick Maclaren.
 
Back
Top