On BNET: Volunteer your way to a job
BNET Business Network:
BNET
TechRepublic
ZDNet
TalkBack 7 of 36:
Next »
« Previous
Very well put.
I would go so far as to say, on the Windows server platform, Itanium is a waste of time since Windows cannot use it to its fullest capacity and on the *nix server platform, more opterons in denser configurations with less cooling for price US$x would by far outperform less dense Itaniums with greater cooling for same price. Want proof?

Look at the top 500 list of super computers and compare Opteron-based systems with Itanium based systems. If you look at the power requirements, number of CPU's and costs, an Intel config with n number of CPU's runs comparable or only slightly faster than an AMD config with 2n number of CPU's while costing more and requiring more power. Comparing equally mached power needs or costs, AMD systems easily outperform their Intel counterparts. Try 4xAMD nodes x 512 compared with 2xI2 nodes x 512. Bare in mind that network topology, protocol, etc. makes a difference when comparing.

Eg., System no 2 and system no 6, 2 has twice as many I2's at 1.4GHz as 6 has Opterons at 2.2GHz I2 system cost approx ten times the cost, consumes about 4 times the power and runs about twice as fast.

System no 5 and system no 6, 5 has only about 4/5th the number of I2's at 1.5GHz as 6 has Opterons at 2.2Ghz. I2 system cost about three times as much, consumes about the same amount of power and runs at almost the same speed.
Posted by: The King's Servant   Posted on: 03/08/05 You are currently: a Guest | Members login | Terms of Use
Reply to Story Reply to Message

Alert moderator to an offensive message

Subscribe to this discussion via Email or RSS

Intel's uncertain processor road-map (or the lack of a road...)  michael-t | 03/08/05
IBM has the advantage here  nucrash | 03/08/05
Intel's problem is  michael-t | 03/08/05
Just Talk  mabricen | 03/09/05
He he he ...  michael-t | 03/09/05
I think Intel would even disagree with your post.  bjbrock | 03/08/05
Very well put.  The King's Servant | 03/08/05
Intel overlooked  michael-t | 03/08/05
Itanium vs. Alpha, not x86  Filker0_z | 03/08/05
Opinion noted  Roger Ramjet | 03/09/05
Back to be Basics?  nucrash | 03/09/05
Cell out  Roger Ramjet | 03/09/05
Itanium Headed for Museum...  Neutrodyne | 03/08/05
I doubt it.  Roger Ramjet | 03/09/05
Wishful thinking on Intel's part  balsover | 03/08/05
Misconception  Roger Ramjet | 03/09/05
Definite shift in the party line  John L. Ries | 03/08/05
Itanium for "Big Iron"?  mwagner@... | 03/08/05
Always room  Roger Ramjet | 03/09/05
Topic of interest: Should we stick with x86?  CobraA1 | 03/09/05
Sticking  Roger Ramjet | 03/09/05
Why Not Emulation  nucrash | 03/09/05
No emulation  Roger Ramjet | 03/09/05
excuse me?  doh123 | 03/09/05
No, really - according to Intel  fmcgowan | 03/09/05
it has some hardware to help, yes, but  doh123 | 03/09/05
We are stuck with X86 for another 20 years  Prognosticator | 03/10/05
Wishful Thinking  kelkins@... | 03/09/05
Itanic WILL win in the end  Roger Ramjet | 03/09/05
A couple of points  fmcgowan | 03/09/05
Points taken  Roger Ramjet | 03/09/05
Good to hear back so soon...  fmcgowan | 03/09/05
True, as there is nothing to really encroach on is there?  PaddyOmaddy | 03/09/05
The wild cats in my backyard won't disturb the unicorns..  Xunil_Sierutuf | 03/09/05
Its about market acceptance...  GeiselS@... | 03/09/05
Wait for Itanium? IBM PowerPC G5 is ready right now.  billmichael | 03/13/05

What do you think?

SponsoredWhite Papers, Webcasts, and Downloads

SmartPlanet

Click Here