Cell computing explained for ninnies like me

As was pointed out on slashdot where I first read the article:

Nicholas Blachford is an idiot. Do not read any of his articles. Just to give you the best of Nicholas, read his antigravity article and visit his web site:

http://www.blachford.info/quantum/gravity.html

Also, look at the nose pictures of him ;)

http://www.blachford.info/other/me.html

Seriously, the guy has burned most of his sane braincells.

For serious laugh, read his article series 'building the next generation' from osnews. I really got good laughs from that 4 part series.

Also, it didn't take long to spot a totally idiotic statement from todays slashdotted article:

> Parallel programming is usually complex but in this case the OS will look at the
> resources it has and distribute tasks accordingly, this process does not
> involve re-programming.

Here Nicholas misses the core problem of parallel programming. The program algorithms _always_ have to made parallel. The OS can't do it.

Should this be called, "Cell computing explained by a ninny"?

nerds, attack!

The hype machine commeth...

And despite the fact that Sony promised the world and didn't deliver on PS2 we've already got a ton of people blindly believing they'll do it next time.

Sony company line (PS2):
Emotion engine is revolutionary, it'll do everything faster then it's ever done before. We'll take over PC's and all computing with our all powerful new processor. And despite the fact that the ps2 might have 4 processors developers will be able to use all of them in parallel with no difficulties. In fact if they don't figure out how to use our system it's not because of over hype, bad chip design, or poor developer support, it's because the developers are lazy.

Reality (PS2):
All quoted numbers on speed were for drawing unlit, untextured triangles. Real-world numbers were much lower. Dealing with multiple processors was a pain in the ass, so most launch titles didn't use any chips but the "emotion engine". Over time developers figured out how to get to the rest of the machine, but it was a pain in the ass.

Despite fast speeds at ship time within 3-6 month of the launch a high end PC's were faster.

Sony company line (PS3):
Cell is revolutionary, it'll do everything faster then it's ever done before. We'll take over PC's and all computing with our all powerful new processor. And despite the fact that the ps3 might have something around 32 processors developers will be able to use all of them in parallel with no difficulties. In fact if they don't figure out how to use our system it's not because of over hype, bad chip design, or poor developer support, it's because the developers are lazy.

Speculative Reality (PS3):
We'll get a lot of exciting benchmark numbers from Sony, which will be accurate under ideal circumstances and not have anything to do with the real world. The machine will still be damn fast, but launch titles won't look nearly as good as we'd expect because developers will have so much trouble trying to keep all the processors busy.

On the PC front with 3-6 months of launch both Nvidia and ATI will launch new cards based on upgrades of the tech they developed for the consoles. Intel and AMD will start launching CPUs with multiple cores on a single chip. The AMD 64x4 will blow away performance on any console. (Of course no games will be written to fully take advantage of it, because so few people actually own them)

Barrett,
Nicholas is not an Idiot - I disagree with what you said:

QUOTE
Also, it didn't take long to spot a totally idiotic statement from todays slashdotted article:

> Parallel programming is usually complex but in this case the OS will look at the
> resources it has and distribute tasks accordingly, this process does not
> involve re-programming.

Here Nicholas misses the core problem of parallel programming. The program algorithms _always_ have to made parallel. The OS can't do it.
UNQUOTE

FYI, OSes have been adding parallel processing to old apps for ages. Intel started adding hardware branch prediction units since the early Pentiums, which allow out-of-order execution and allow to achieve certain degree of parallelism.

Of course, these are not the ideal ways to explore full potential of parallel computing: for that all applications and algortihms need to be redesigned, but meanwhile these things offer a nice, cheaper alternative.

I worked on compiler constructions back in 1996, and we learned these tricks in school, so I don't think anyone with a formal computer science education will find Nicholas "Idiot".

And as for the nose pictures, it is a personal thing and you shouldn't judge his technical prowess by the size of his nose.

Enjoy!