Transmeta Founder Talks Chips

gManZboy writes "Dave Ditzel, CTO and Founder of Transmeta (you remember Transmeta? weren't they supposed to kick some Intel booty?) sits down and speaks with Alpha and StrongARM chip designer Dan Dobberpuhl about the history of CPUs, where they're heading, and how the heck we'll keep up Moore's Law (if we can)."

Comments

What you're likely to see

[fnj]

I'd like to see ... 5W CPUs with the more processing muscle as today's 60W beasts

It would be a fine thing, but there's no sign of it happening. Instead, the next desktop CPUs are due to dissipate more like 103 watts. It's sad.

Re:Moore's Law

[Ignis Flatus]

Many people have made the observation that Moore's Law is probably a limited phenomenon, and while other increases may continue to fuel increased processing power, Moore's Law does not actually have anything directly to do with processing power.

Who needs Moore's Law when we've got Beowulf clusters?

And Beowulf clusters of Beowulf clusters.

And Beowulf clusters of Beowulf clusters of Beowulf clusters.

And...

Two Fabless Guys Talking Process Technology

[stevesliva]

Okay, I look at the impressive resumes belonging to both the interviewer and interviewee, and I cannot believe how little substance there is to their conversation. Why is that? They're almost powerless (no pun intended) to influence the development of process technologies. Transmeta is a fabless company that contracts with TSMC, I believe, to manufacture their processors, and the interviewee just started another fabless company. If you want to speculate on where process technology is going, ask someone with a fab!

They spend several paragraphs discussing NMOS capicitors in CMOS processes circa 1994, but apparently neither knew enough to speculate about MIM or Trench capacitor structures, two mature technologies used in DRAM. Yes, they were leading in to the gate leakage issue, but the substance of that boiled down to, "Leakage sure is a big problem." Their solution is low-voltage chips with fewer transistors. Revolutionary!

There's way more substance in press releases from Intel.

Moore's Law forever - NOT

[fnj]

Moore's Law is probably a limited phenomenon.

<pedantic>
Probably? Assuredly, I would say. If transistor count continues to double every 2 years, with 42M transistors per CPU in 2000, you would have 43 billion in 2010, 44 trillion in 2020, 47*10^21 in 2050, and 53*10^36 in 2100. If that hasn't reached the number of atoms in the known universe, then keep counting years and it will.
</pedantic>

Re:Moore's Law forever - NOT

[f97tosc]

Probably? Assuredly, I would say. If transistor count continues to double every 2 years, with 42M transistors per CPU in 2000, you would have 43 billion in 2010, 44 trillion in 2020, 47*10^21 in 2050, and 53*10^36 in 2100. If that hasn't reached the number of atoms in the known universe, then keep counting years and it will.

The number of atoms in the universe is not the limit for computation. The true limit is set by quantum states. It is actually possible to caluclate these limits, Professor Seth Lloyd at MIT calculated how much computation can possibly come out of 1kg of matter confined to a volume of 1l - "the ultimate laptop". And yes, if Moore's law keeps going then we will be there in a few hundered years. Of course, it could be argued that we will never get there in pracitce. On the other hand, current quantum computers are nearly optimal in this regard - unfortunately they only contain a few atoms or so...

Tor

Re:Translator code...

[msgmonkey]

The reason only x86 was done was because the other platforms already had low power chips, especially motorola and no one needs a low power alpha.