Intel Makes 45nm Chip

dolphinlover writes "Intel announced today that it created its first microchip using the 45 nanometer manufacturing process that it says will go into its processors in the second half of 2007. Intel said that this development provides it with a 'considerable lead over our competitors in the 45-nanometer generation'."

Why do they always screw up Moores Law

[rminsk]

The new chip makes good on Moore's Law, an industry maxim set forth by Intel co-founder Gordon Moore that stipulates the number of transistors on a chip -- and therefore its processing power -- doubles roughly every 18 months to two years.

Who added the "and therefore its processing power" to the quote? Was it the reporter or someone from Intel? Moores law has nothing to do with processing power.

Re:Says You

[the eric conspiracy]

AMD has a co-development agreement with IBM and is planning to introduce 45nm parts in 2008.

Re:Says You

[uujjj]

AMD is nearly a full year behind Intel rolling out 65nm. Intel began volume production at 65nm last summer; AMD will be ramping up in the middle of this year.

While the parent may be joking, down below you'll find a lot of posts from AMD fanboys insisting that AMD must somehow be ahead. These fanboys are as clueless as the average tech magazine reporter. You can be quite certain that AMD will not be ramping up 45nm before Intel.

CNET News article has important additional details

[MojoStan]

Here's a link to CNET News's article on this same news:

Intel shows test chips made on future processes [com.com]

One interesting (to me) bit of info from the CNET News article:

The 45-nanometer process could become particularly interesting because many chip designers believe it will be one of the more difficult transitions in years. The power consumption and performance requirements of these chips will be extremely high and chipmakers are being forced to add exotic materials and new structures to their transistors to ensure the chips function properly...

"It does get a little more challenging every time, but we come up with new technology and tricks to keep things going," said Bohr.

If a company botches the process, it could lead to product delays or recalls. Some chipmakers faced these problems during the transition to 130-nanometer chips when they swapped aluminum for copper for making interconnects--the tiny wires inside chips.

Although Intel might have a "considerable lead over our competitors in the 45-nanometer generation," it doesn't appear that this transition is expected to go as smoothly as their transtion to 65nm (which seems very smooth). Remember Intel's and IBM's difficult transitions to 90nm.

Re:What about AMDs 45nm???

[TheRaven64]

Wow, the AMD fanboys are out in force tonight. From this source [computerworld.com.au], which is three months old and so relatively recent:

• AMD's new fab, Fab 36, supports 300mm wafers (like Intels have for some time).
• It uses a 90nm process (Intel and IBM have been on 65nm for some time).
• It will transition to 65nm by the end of 2006.
• It will use 45nm and 32nm processes by the end of the decade.

It doesn't really sound like Intel is playing catch-up here.

Re:Aren't we getting close to the Theoretical Limi

[uujjj]

molecule? This is a crystal we are talking about, so the entire wafer is a "molecule". An atom of Si is about .3nm across.

Re:Why do they always screw up Moores Law

[PunkOfLinux]

double the number of cores? me thinks not. If that was the case, we'd have 32-core processors bye now -- which, needless to say, we don't.

Re:Says You

[uujjj]

"ramp up" is a common term in industry and elsewhere. it means indicates a company's acceleration of the rate of production. i don't quite understand the problem. it's like complaining "Cheesus christ man, stop saying the, *please*"

Re:Aren't we getting close to the Theoretical Limi

[WallaceAndGromit]

Most molecules are a few to a few dozen angstroms thick (from here [wikipedia.org]), and 45 nm is [google.com] 450 angstrom. So there is about another factor of 10 till we get down to the size of complex molecules. However, I do believe that most of the "stuff" used in the manufacture of chips are either pure elements or simple molecules, which are much smaller (varying from 1 to 5 angstrom [wikipedia.org])..

Right on schedule

[Anonymous Coward]

Intel's logic development is striving for a two-year cycle for each new process technology. This announcement of functional first-silicon (who knows how long they've actually had it) is part of that natural progression. Here's a table showing this announcement along with previous SRAM test chip announcements:

Process
Litho
Size
Date

P860
130 nm
18 Mbit
Mar 2000

P862
90 nm
50 Mbit
Feb 2002

P1264
65 nm
70 Mbit
Apr 2004

P1266
45 nm
153 Mbit
Jan 2006

Okay ... it's not a table...

Re:Says You

[georgewilliamherbert]

Silicon On Insulator; see the silicon on insulator [wikipedia.org] wikipedia entry for a high level summary, or google the phrase for more details.

Basically, instead of a solid slab of silicon on which you fab chip components, you put a solid slab of an insulator (sapphire / alumina for example; see silicon on sapphire [wikipedia.org] wikipedia entry) down and then an insulating silicon oxide layer, and then a thin layer of silicon on which you fab the parts. Since what's under the parts is insulator, rather than more semiconductor, it reduces the energy of switching and reduces the time to switch a transistor. Also reduces radiation effects on the semiconductor and other good stuff.

Re:What about AMDs 45nm???

[mczak]

(Intel and IBM have been on 65nm for some time)

You're right about intel (though "for some time" seems a bit exaggerated if you only count shipping products - something like a month). intel really is ahead 6 month. However, IBM is not using 65nm tech with any of its shipping products (that I'm aware of). In fact, their power5 only transitioned to 90nm very recently, though the powerpc g5 transitioned to 90nm somewhat earlier than AMD transitioned their chips to 90nm I believe. Some time ago rumours said the first dual-core g5 chip would be 65nm (the powerpc 976) but obviously this was just that - an unsubstantiated rumour, the powerpc g5 dual-core materialized as a 90nm chip (970mp).

Re:Says You

[Overly Critical Guy]

Intel is succeeding Netburst with Merom and Conroe later this year. Viiv is just a marketing name for a list of minimum system specifications for multimedia computers that happens to use a bunch of Intel-branded parts. It's unrelated to their chips.

As for who is in the hole, AMD is a year behind both 65nm and 45nm, and the Yonah is a laptop chip competing performance-wise with AMD's desktop processors. 'nuff said.

Those extra registers in 64-bit don't go that long a way (about 5%-10% on average last I checked the benchmarks). A lot of the 64-bit performance comes from the fact apps in 64-bit mode know that their chip will have at least SSE, which speeds things up. A 32-bit app optimized with SSE instructions can compete with 64-bit performance, since 64-bit is slowed down with the cache bloat and increased pointer size. 64-bit is hype designed to sell chips. It's not needed unless you actually have to access more than 4GB of RAM.

This is a benefit for the Intel Macs, whose baseline will always be the Core Duo that has SSE3, meaning all apps will be compiled with support for it, 64-bit or not. Until you need more than 4GB of RAM, 64-bit is overrated buzz that offers little.

Re:Says You

[Overly Critical Guy]

Uh, no they're not. Even the laptop Core Duo is matching the Athlon64 3800+ X2.

Is there a requirement to jerk off over AMD when you sign up to Slashdot or something? These aren't the Pentium 4 days anymore. Intel owns the mobile market, and their future roadmap kills AMD's.