Replies to post #30209 on Advanced Micro Devices Inc (AMD)

[Jules2]

Posted by: sgolds
In reply to: wbmw who wrote msg# 30205 Date:4/1/2004 11:37:15 AM
Post #of 30208

wbmw, I certainly agree that SGI is having success with Itanium. They are moving their customer base over to the new architecture. (I am less convinced that they are reaching out to new customers in large numbers. SGI is really small potatos.)

Now, what does that mean? We have two companies who have an Itanium strategy that will succeed - SGI and HPQ. A bunch of other OEMs offer Itanium as check list products but don't take it very seriously for their markets (such as IBM and Fujitsu).

SGI makes low volume, high performance systems. HP is using Itanium in low volume, high performance systems. When we step back to see the forest for the trees then the picture becomes that Itanium will have a chunk of this niche, and is not reaching out to the broader market.

Back when Intel posters were predicting that Itanium would dominate everything, and AMD posters were predicting the demise of Itanium, I was one of the few posters who said that Itanium would have success in HPC because HP would bring their customer base to it, but would not displace x86.

So far I see nothing to contradict my position. Itanium found its niche.

sgolds.

If you look back I allso stated that IMHO itanic would be relegated to a niche market.
Thats exactly where it's headed. In my opinion intel will never recoupe it's R & D.
I'm sure the intel investors will, at some point, wake up to that fact.

Regards

Jules

[borusa]

Sgolds, My knee jerk reaction to Itanium success:

While there may be a niche for such a product the costs make that an expensive niche and vulnerable to assult from below, a replay of what has already happened...

Maybe too simplistic, who knows.

[Tenchu]

Sgolds, Back when Intel posters were predicting that Itanium would dominate everything, and AMD posters were predicting the demise of Itanium

What do you mean "predicting that Itanium would dominate everything"? Even I let go of that belief years ago.

Tenchu

[chipguy]

Now, what does that mean? We have two companies who have an Itanium strategy that will succeed - SGI and HPQ. A bunch of other OEMs offer Itanium as check list products but don't take it very seriously for their markets (such as IBM and Fujitsu).

Take your blinkers off and look again. Unisys is pushing IPF
in its large commercial servers with remarkable success.
Last year 30% of its hardware sales were IPF based, about
twice what it had projected. Not bad considering Madison
was released mid year. NEC seems quite serious at giving
HP a run for its money in the 16 to 32 processor commercial
server field and Hitachi recently brought out a mid range
IPF server based on its own chipset.

In 4Q03 IPF system sales were AFAICT over $300m of
which HP had about half. Of the $150m+ that HP didn't grab
SGI accounted for about $50m. That leaves over $100m of
business divided up among the rest. That's not a huge deal
at the moment but if the 60% quarter over quarter IPF sales
growth rate that HP recently reported reflect the general
rate of market acceptance then it is obvious why Sun and
IBM are frantically throwing darts at it.