The Design of Software (CLOSED)

A public forum for discussing the design of software, from the user interface to the code architecture. Now closed.

The "Design of Software" discussion group has been merged with the main Joel on Software discussion group.

The archives will remain online indefinitely.

Paper on Dynamic Optimization/Compiling


I'm looking for this paper about dynamic optimization/dynamic compiling by Aggarwal or Aggarval. I'm not to sure about the exact name of the paper, or even the correct spelling of the researcher so I'm hoping someone perhaps more familiar with this topic would recognize which paper I am referring to.
I do know with better certainty however that Transmeta based their code-morphing technology on this paper.

If anyone knows which paper I am referring to, or has any better ideas on how to find this paper than Google I'd be happy to hear from you.

CompSci Grad Student
CompSci Grad Student
Friday, February 18, 2005
I know of one prof named "agarwal".

Bankstrong Send private email
Friday, February 18, 2005
What the heck is code morphing? Am I falling behind?
Friday, February 18, 2005
Probably not. Code Morphing is marketspeakbullshit for binary translation. This is turning one type of machine code into another, also known as "recompilation", and is used in place of the interpretive approach when simulating foreign CPUs. COmpared to interpreting the foreign instructions, you get a significantly lower per-simulated-instruction cost at the drawback of a potentially significant translation cost.

(The per-simulated-instruction cost is almost always better than the interpretive equivalent due to the avoidance of the almost-always-mispredicted instruction dispatch branch you must have in the interpretive situation, but this may fall down if you must cater for self-modifying code and keeping hardware in sync with the CPU clock and so on.)

The Transmeta CPU does this because it doesn't run x86 code, but instead translates it into its own format before running it. I believe the Pentium 4 does the same thing.
Saturday, February 19, 2005
Have you tried talking to somebody in the CS faculty? Your universitys' library?

That should give you at least the name of the prof - quite possibly, somebody there might even know what paper you are talking about.
Robert 'Groby' Blum Send private email
Saturday, February 19, 2005
There are too many Aggarwals but no Aggarval. You can search for them in .

According to their publications, Aneesh Aggarwal seems to correspond to what you're looking for. He has a lot of papers about software optimization:

Hope this helps.
Developer next door
Monday, February 21, 2005

This topic is archived. No further replies will be accepted.

Other recent topics Other recent topics
Powered by FogBugz