Cell Disclosures

IBM has released several Cell related docs:

• Cell docs "home" page (note that new docs are listed, with August 2005 dates)
• Discuss Cell arch (& docs) at developerworks forum.
  
• Biz center's Cell page
• Original Cell Research page

Sony's SCEI has released some too:

• Cell disclosures (English)
• Japanese disclosures
  

Toshiba's disclosures:

• Press release
  

Related pages:

• Cell overview paper (August 24, 2005)
  
• Cell basic paper
  
• Research on: Cell compiler
• Hints of Cell at rHype
  
• More Hints: systemsim and its forum
  
• DeveloperWorks: one example of SPU usage
• Interview with Dan Greenberg, Power Everywhere systems offerings program director (Aug 24, 2005)
  

East West perception diffs

"Asians, Americans Show Perceptual Divide" article points out the perceptual differences between Americans and Asians even those Asians raised in the U.S. Funny how culture affects even perceptual cognition. How that applies to usability and user interace are well worth thinking about....

Cell Publications

Here are some links to papers related to the project I'm currently involved in: IBM's Cell Blade info.

Cell version of the full system simulator is being prepared for release, as well!

Full System Simulator

I have been involved indirectly with the following product which was released to the public by IBM: Full System simulator IBM PowerPC 970.

See also discussion forum.

Next Programming Model

/. had pointer to: Sim Simeonov's "i-Technology Viewpoint: The Next Programming Models" which I thought might have something profound to say as AOP or OOP. But not so. More of the same by combining languages like AJAX and SOA or Service Oriented Architecture -- none of which are profound in my opinion. 60's and 70's gave birth to OOP. 90's gave birth to AOP. I doubt if we can expect anything profound for another 5-10 years (or more).

Unlike Simeonov, I've been programming since 1980 and have read many papers and books by visionaries like Ted Nelson and Alan Kay long before internet was widely available (I went to public universities like UCLA just to get my hands on IEEE and ACM papers as well order from Xerox PARC for copies of their reports). OOP made a splash in 1981 when Byte magazine published a whole issue on Smalltalk (I didn't grok it when it came out but it was about a year later when a college friend pointed it out and I got into it myself). I got to know AOP because I was in Silicon Valley and had the chance to attend a public talk at Xerox PARC in 1994 (back when it was called Open Implementation).

If the past is any indicator, since OOP took about 20 years to get "popular" (Smalltalk first real release was out in 1972 while C++ came out in late 80's and didn't gain traction until about early/mid-90's) I expect AOP to take 10 more years before it becomes common enough. And 5 more after that before anything profoundly new comes out, I suppose.....

Ebay for Education

So, what's needed to make a home for people to get together to learn? Ebay of education? Today, you can get together and shoot the breeze in any given blog or chat site or even USENET newsgroup. Or comments at shopping sites like amazon.com.

There's even MOO or MUD to allow creation of virtual worlds for education.

However, one cannot pool money and time to, say, explore the Titanic by merely chatting about it.

Some issues to work out: Money and Time can be harnessed only by building up trust (just as you trust an ebayer with high number of positive comments). And like any financial system, one can always game it to take advantage of it be it the (real) stock market or virtual worlds like EverQuest with real-world effects or pseudo-virtual worlds like HSX.

(I'm a believer in open source and I'm against software patents so I'm all for talking about my ideas openly in the public so that it's not hidden.)

Best Programmers make the difference

Reading Joel's "Hitting the High Notes" points out how stars make a difference in movies and they are paid accordingly. And Joel argues that best programmers are similar:

Here's why: duplication of software is free. That means that the cost of programmers is spread out over all the copies of the software you sell. With software, you can improve quality without adding to the incremental cost of each unit sold.

Essentially, design adds value faster than it adds cost.

Or, roughly speaking, if you try to skimp on programmers, you'll make crappy software, and you won't even save that much money.

I personally would argue that the best programmers are artists and most programmers are craftsmen. I don't know how people -- let alone companies -- can weed out artists from craftsmen. One can test for basic competency of a craftsman but then how to weed out the best from the common? And if you're one of the better programmers, how can you assert your abilities?

new commenting!

Haloscan commenting and trackback have been added to this blog.

Danny's new blog

This is my third blog I'm starting since I want to blog on tech without clutting my "main" page on home education and libertarian ideas. I'm currently a full time employee but I'm about to finish my current position at the end of the month and I hope to become a contractor/consultant, so I'll post more freely starting September sometime.