On The Insider: Britney's Bikini-Clad Top 10
BNET Business Network:
BNET
TechRepublic
ZDNet
TalkBack 10 of 23:
Next »
« Previous
Is almost as bad as folk knowledge
Please cite a study that shows your contention. As a guy who wrote commercial compilers for almost all the major RISC chips when they came out I have seen no evidence of this claim.

In fact depending on the system and the software RISC versus CISC can vary all over the place. In particular, things like the size of the code versus the cache and memory can put RISC at a disadvantage.

Years ago, I did a study for a major PC manufactuerer on the potential performance on parallelism, from things like superpipeling, superscalar, and VLIW. One of the things that struck me was that research going back to the mid-60's was highly consistent on the performance benefits. Basically for common software (not special parallel cases) you find that there is improvement for up to 4 processing elements, after 4 improvement is rare.

After I did that study, I was talking to the head of Advanced R&D for the firm when Itanium first was being shown to the vendors. Intel was claiming great performance based on benchmarking Towers of Hanoi (the classic parallelism benchmark). I pointed out that they would never really get these numbers, and suggests they look at my paper again. I was later told I had predicted Itanium's performance curve very closely even though the paper was written years earlier.

Bottom line, is Intel based their design on flawed assumptions. There is no way to make this chip meet its hype.
Posted by: oldsysprog   Posted on: 08/31/05 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

The Problem With Itanium . . .  jjworleyeoe | 08/30/05
HP PA-RISC still shipping  dwjunix | 08/30/05
Yes, think Open Solaris on AMD x86-64  DonnieBoy | 08/31/05
Opteron beats Itanium by 250%  sharikou | 08/30/05
Truly unbeatable !  Prognosticator | 08/31/05
Unbiased link  Roger Ramjet | 08/31/05
Ok, show a different link  balsover | 08/31/05
What you don't know . . .  Roger Ramjet | 08/31/05
What you don't know about people  dwjunix | 08/31/05
Is almost as bad as folk knowledge  oldsysprog | 08/31/05
No flawed assumptions...  Prognosticator | 08/31/05
Yes compilers are hard, and the Itanium puts all the work on the compiler  oldsysprog | 08/31/05
Correct..as it should  Prognosticator | 08/31/05
You did not read my post  oldsysprog | 08/31/05
Thanks, I understand your post.  Prognosticator | 09/01/05
Yes, and x86-64 is still a whole lot cheaper, ask Google.  DonnieBoy | 08/31/05
Too Little, Too Late.  SimonSays | 08/31/05
The problem is, x86-64 is better and cheaper.  DonnieBoy | 08/31/05
Not necessarily for the reason stated  balsover | 08/31/05
Well, MIPS per dollar or MIPS per watt is what determines best.  DonnieBoy | 09/01/05
Another example of intel's...  wizzzer | 08/31/05
Itanium had failures on several fronts.  tyezek | 08/31/05
Some Misconceptions  Roger Ramjet | 09/01/05

What do you think?

SponsoredWhite Papers, Webcasts, and Downloads

SmartPlanet

  • Thought-provoking progressive ideas on diverse topics that intersect with technology, business, and life, and matter to the world at large. Visit SmartPlanet
  • More from IBM
  • Innovate your business' process model, play against the market, compete against others on our scoreboards and WIN! Try INNOV8 2.0: A BPM Simulator
  • Enabling Real-World Business Transformation through IBM Service Management Read the EMA Analyst Report
Click Here