AMD [IBM] press release

[Bob Drzyzgula]

On Wed, Nov 20, 2002 at 04:07:11PM +0100, Alan Scheinine wrote:
> 
>    This is one point in which we clearly have a difference of
> opinion concerning the computer market in general.  During several
> years of asking for price quotes, it has been my impression that a
> high-performance "server" from Compaq or HP (when they were separate
> companies) had a price that was about twice what I would need to pay
> to get an equivalent computer assembled locally, at the same level
> of performance and sometimes with the same motherboard.

Certainly brand names do add significantly to the price of
a unit. But also, top-quality enclosures and power supplies
add real value, and usually these systems include several
years of on-site service. The value of some of these things
will vary from customer to customer, and for many people, the
locally-assembled, or self-assembled option will be a better
choice. Of course, with non-commodity systems, such as those
based on the Itanium, the Power4, UltraSPARC, etc., these
latter choices are often not an option.

>    Dr. Hahn mentioned a 4-way Xeon, around $20k from Dell.  My
> impression is that Dell has rather low prices, whereas, my intention
> was to say that in general, including HP and IBM, PC labelled "server"
> have relatively high prices.  With regard to Dell, we have some
> commercial software that does not run efficiently on a cluster, so I
> was wondering if a 4-CPU shared memory PC would be a cost-effective
> choice for the group that ran this commercial software.  The Dell
> Poweredge 6600 has a cost of $10,288 for a two-processor version with
> a 2-GHz Xeon CPU, one GigaByte of memory, and one 18 GB Ultra 160
> SCSI harddisk.  But wouldn't I be able to get a two-processor PC with
> this level of performance for about half the price?  (Perhaps even
> with two-way interleaved memory.)  The price of a Dell Poweredge 6600
> with four (4) processors (2-GHz Xeon) and 2 GigaBytes of memory
> is $25,886.  My impression is that I could assemble a similar computer
> for half the price.

Make sure that you are comparing apples to apples. AFAIK,
Quad- processor (Pentium 4) Xeon systems all use the
Xeon MP processors, which have larger caches and are
quite expensive -- from $1,000 to $4,000 or more each,
and they need high-end motherboards which cost thousands
of dollars. Just a motherboard with four 2GHz/2MB Xeon
MPs and 2GB of memory could easily cost $20,000 or more.
Single- and Dual-processor systems, by contrast, can be
built with standard Xeons and Intel 7500 chipset-based
motherboards each costing only hundreds of dollars;
a Tyan Tiger i7500 with two 2.8GHz Xeons and 1GB of
memory can probably be had for well less than $2,000,
to which, of course, one has to add all the rest of the
system.

>    My impression is that the price of a Dell server is at the low end
> of the Xeon-based servers from various companies.

Which is one of the reasons Dell sells so many systems;
they are very, very competitive.

>    Overall, the question that I intended to put forward in my previous
> email message was primarily based on the Itanium / Hammer level of CPU
> performance.  It was my impression that by this date Itanium was supposed
> to arrive at the desktop.  In fact it has, since one group here has
> a slick-looking Itanium workstation on loan from HP.  But at what cost?

Probably high. The real desktop, or at least "workstation" Itaniums,
"Deerfield", are not likely to show up until sometime next year.

You might find interesting some of the recent articles at Ace's
Hardware:

  * Chris Rijk' guide to Volume Multi-Processor Systems,
    starting with part 1 at:

    http://www.aceshardware.com/read.jsp?id=45000338

  * Opteron and Itanium: Two Roads to 64-bit Computing,
    by Johan De Gelas, at:

    http://www.aceshardware.com/read.jsp?id=50000240

  * More Details on AMD's Opteron, also by Johan De Gelas,
    at:

    http://www.aceshardware.com/read.jsp?id=50000313


>    My point, though not stated clearly, was the price of the Itanium 2.
> The article from Supercomputingonline gave me the impression that
> Itanium 2 was equivalent in price and performance.  Though I do not
> have the URL, I remember reading a month ago that AMD said that their
> first 64-bit CPUs would be for the server market.  Comments concerning
> that announcement were that nobody knows what that means, so I asked
> the Beowulf group if they knew with the AMD CPUs would have a price
> and availability similar to Itanium 2 initially, for example, in the
> first half of 2003.  Would AMD be thinking, if Intel can sell their
> Itanium 2 at the stratospheric price of a Power4, maybe we should try
> starting at that price level? (???)  

They've probably thought it, and they might try, but it would most
likely be at the quad-processor level that this would happen, possibly
by playing a "Athlon XP/Athlon MP" trick, where only a small number of
high-test samples of the Opteron would be rated for use in Quad-processor
systems. At the dual level, they really need to attack the Xeon, not
the Itanium.

>    This is very useful information.  I have the impression that there
> is substantial uncertainty with regard to the timing of the availability
> of the AMD processors and motherboards.  Would anyone else on the
> beowulf mailing list have similar information about how concrete
> are the announcements?

Anybody with hard information here is probably under NDA.

--Bob