SI
SI
discoversearch

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor.  We ask that you disable ad blocking while on Silicon Investor in the best interests of our community.  If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.
Technology Stocks : Rambus (RMBS) - Eagle or Penguin
RMBS 124.49+7.6%Jan 28 3:59 PM EST
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  
To: jim kelley who wrote (61655)11/21/2000 3:56:43 PM
From: SBHX  Read Replies (1) of 93625
 
Jim,

When these applications are recompiled the "sluggishness" in the legacy software will evaporate. This is going to be a big year for the software folks.

I won't be so sure. Many of the spreadsheet and database and word processing s/w that runs in the business world are non-streaming and could potentially suffer the most with granularity loss.

If the intention is to ignore all business s/w for benchmarking, then fine, there are video streaming and 3D applications [rewritten to change the ordering of the pre-geometry vertices (something called structure of arrays (proposed by intel) over array of structures (today in openGL and D3D)] that will benefit enormously from anything that uses more of the chunks of data being read in. Any app that addresses data sparsely will not like this.

I doubt if many of the business world apps will benefit from a 'recompile' without rethinking their underlying data structures to solve the granularity loss problem.

But I agree that what these machines need is a whole new set of benchmarks that run well on streaming data to make them look good.

Or so my graphics and computer architecture guru friend tells me. (G)

edit: actually, I take that back. There used to be certain mixes of instructions generated by a P54/P55C compiler to take advantages of the U-V pipes that caused horrible partial stalls on PentiumPros and PIIs. Perhaps something similar is there. Though I can't think anything off the top of my head about P6 style code generated by a C compiler that are order dependent --- the PII/PIII thing is superscalar out of order, so it shouldn't matter. But I could be wrong.

SbH
Report TOU ViolationShare This Post
 Public ReplyPrvt ReplyMark as Last ReadFilePrevious 10Next 10PreviousNext  

HomeMailHotSubjectMarksPeopleMarksKeepersSettings
Terms Of UseContact UsCopyright/IP PolicyPrivacy PolicyAbout UsFAQAdvertise on SI
© 2026 Knight Sac Media.  Data provided by Twelve Data, Alpha Vantage, and CityFALCON News