Chips pass two billion milestone

The first chip to pack more than two billion transistors has been launched by silicon giant Intel. The quad-core chip, known as Tukwila, is designed for high-end servers rather than personal computers. It operates at speeds of up to 2Ghz, the equivalent of a standard PC chip. It marks the latest milestone in chip technology; Intel released the first processor to contain more than one billion transistors in 2006. "It's not revolutionary, it's another evolutionary step," said Malcolm Penn, an analyst at Future Horizons, of Tukwila.

View: Full Story @ BBC News

Report a problem with article
Previous Story

Windows Server 2008 & Windows Vista SP1 RTM Today

Next Story

Microsoft Plans to Roll Out Cheaper Xbox in Japan

15 Comments

Commenting is disabled on this article.

What happened to Intel? They were up to 3.6Ghz processors, and now all they produce is 2.0Ghz (Range) processors.... much like AMD Processors do.

Think of this comparison, a 2 seat passenger sports car at 3.6Ghz speed compared to a 50+ passenger bus going at 2.0ghz. Not as fast but can do a whole lot more in the same distance. Ok so maybe not the best comparison but it popped into my head pretty quick while reading your comment.

Blah blah, not only that, but Penryn are clocked closer to the Pentium 4's than ever before. You can be sure that not too many refreshes more from now, the clockspeed will start increasing too! I personally want a 5GHz then 10GHz processor.

(lylesback2 said @ #3)
What happened to Intel? They were up to 3.6Ghz processors, and now all they produce is 2.0Ghz (Range) processors.... much like AMD Processors do.

The 3.6 GHz processors were among the slowest when it came to anything except video rendering. It had a horribly long pipeline architecture, so the best analogy would be having a bit faster car which could only turn every 3 blocks compared to the AMD/Core 2 which could turn anywhere you want but was a bit slower. Although you may get to a point faster with the faster one, you are pretty much at a disadvantage in most situations.

Both AMD and Intel only seem to be about the "Core's" now. While most people don't run more then one application (basic users) who are purchasing these dual-core, and quad-core processors, they aren't even taking advantage of the multi-threading part of the chip.

I agree with Thrawn, I wanna see 10Ghz chips, and then go for super multi-core chips that Intel promised by 2010.

I hope within 3 years, the Ghz's actually move from this stand still at 2.0-2.6Ghz range, and finally break the 3 or even 4Ghz marker with multi-core chips

mmmmm, 64bit with 64cores, on a 6.4Ghz processor

(ivanz said @ #3.4)
The 3.6 GHz processors were among the slowest when it came to anything except video rendering. It had a horribly long pipeline architecture, so the best analogy would be having a bit faster car which could only turn every 3 blocks compared to the AMD/Core 2 which could turn anywhere you want but was a bit slower. Although you may get to a point faster with the faster one, you are pretty much at a disadvantage in most situations.

I'd propose a better metaphor being thusly:

A 150hp car engine runs at 3000-6000 RPM.

The 4000hp engine in a diesel locomotive runs at 1000 RPM or so.

I don't think there's any application which really wins on a P4 CLOCK-FOR-CLOCK. The only reason the P4 was competitive at all was how high it clocked.

(lylesback2 said @ #3.5)
Both AMD and Intel only seem to be about the "Core's" now. While most people don't run more then one application (basic users) who are purchasing these dual-core, and quad-core processors, they aren't even taking advantage of the multi-threading part of the chip.

I agree with Thrawn, I wanna see 10Ghz chips, and then go for super multi-core chips that Intel promised by 2010.

I hope within 3 years, the Ghz's actually move from this stand still at 2.0-2.6Ghz range, and finally break the 3 or even 4Ghz marker with multi-core chips

mmmmm, 64bit with 64cores, on a 6.4Ghz processor :D

That's literally one of the most uninformed comments ever, no offense. It's not just about the quantity of cores. It's about the efficiency of the clock per clock processing ability.

Clock speed means nothing now. It's merely an indication of the operating frequency that affects processing power within the same microarchitecture. Other than that, between architectures the clock frequency has little relevance to do with performance; meaning that, it's not about how high the frequency is, it's about how much the processor can do per clock frequency.

I can hire 20 monkeys to help type a book, or I can hire 10 professional typists to do the same thing. Just because I have more monkeys doesn't mean it can do more work. It's about how much work each one can do.

I can take my Core 2 Duo E6600, disable the second core, and still beat your "3.6GHz" Prescott with relative ease.

(chconline said @ #3.8)
Clock speed means nothing now.

I can take my Core 2 Duo E6600, disable the second core, and still beat your "3.6GHz" Prescott with relative ease.

No one is saying that Prescott was better than Conroe. And in ref. to clock speeds, note that within an architecture, many workloads scale linearly with clockspeed. Linear is very far from "nothing".

(Hak Foo said @ #3.6)
I'd propose a better metaphor being thusly:

A 150hp car engine runs at 3000-6000 RPM.

The 4000hp engine in a diesel locomotive runs at 1000 RPM or so.

lol, nice :P

(Thrawn said @ #3.9)

No one is saying that Prescott was better than Conroe. And in ref. to clock speeds, note that within an architecture, many workloads scale linearly with clockspeed. Linear is very far from "nothing".

Please read my entire post. Don't pull one sentence out of the entire thing.

What the article should mention is that the vast majority of those transistors are allocated to its cache. Still impressive feat.

Sorry replied to wrong post was in relation to vvv post. [/edit]

it's for itanium processors dude they don't need 3.0GHz if you notice itaniums usually come in clusters anyway.