speculation: Apple wants to make its own media-hub game console with a CELL-style processor from INTEL

The more I hear about Cell, the more it sounds like snakeoil to me. The
history of the semiconductor industry has for the past decade at least
been one of hype and disappointment. The more shrill the hype, the more
disappointing the actual product is.

The two biggest examples so far are Itanium, and Transmeta.

Yousuf Khan

Yousuf Khan wrote:
The more I hear about Cell, the more it sounds like snakeoil to me. The
history of the semiconductor industry has for the past decade at least
been one of hype and disappointment. The more shrill the hype, the more
disappointing the actual product is.

The two biggest examples so far are Itanium, and Transmeta.

Both are shipping and fulfill their niche nicely. The Itanium should
never have had any x86 support in hardware, it is labeled as a poor
performer based on it's use as a very expensive slow x86. The actual
chip isn't bad, given a reasonable O/S (Linux) and compiler (Intel,
**NOT** gcc on this one). Or so the reports I read would say.

The real bust was the Intel 432 chip, and I have one on my wall next to
the Z80. It never had much performance or any significant market acceptance.

--
bill davidsen
SBC/Prodigy Yorktown Heights NY data center
http://newsgroups.news.prodigy.com

"Bill Davidsen" wrote in message
...
Yousuf Khan wrote:
The more I hear about Cell, the more it sounds like snakeoil to me. The
history of the semiconductor industry has for the past decade at least
been one of hype and disappointment. The more shrill the hype, the more
disappointing the actual product is.

The two biggest examples so far are Itanium, and Transmeta.

Both are shipping and fulfill their niche nicely. The Itanium should
never have had any x86 support in hardware, it is labeled as a poor
performer based on it's use as a very expensive slow x86. The actual
chip isn't bad, given a reasonable O/S (Linux) and compiler (Intel,
**NOT** gcc on this one). Or so the reports I read would say.

The real bust was the Intel 432 chip, and I have one on my wall next to
the Z80. It never had much performance or any significant market
acceptance.

Agree, Itanium is a strong performer when using Itanium optimized code which
of course no one wants to do. Transmeta
didn't do bad also considering they sold a lot of their tecnhology to others
who will make good use of it. Yousuf's point
is correct though that they came with a lot of hype and didn't deliver but
not because they were technologically inferior.
CELL is just plain inferior in many ways. I hope no one spent too many
billions on it. I'd hate to see more good engineers
laid off due to a bad product.