When it comes to manufacturing in North America we make food, cars, and import just about everything else. Our hunger for high tech has gone into hyper active mode over the last decade, and the thousands of new products pouring into our local Best Buy are usually made in China, Taiwan, or just about anywhere inexpensive labour can be found. The repetitive tasks and long hours have led to a PR nightmare for companies such as Apple, who in earnest, have offered up several guided tours to help highlight the positive aspects of life at these city sized plants responsible for making our gadgets. These tours have been guided, carefully staged events however, and many have wondered what a candid un-staged video might look like. If that’s you, wonder no longer. A manufacturing slipup by Quanta has given us a rare glimpse at one of the HP manufacturing lines, and it’s pretty darn dull.
The industry-wide move to the 28nm manufacturing process has been slowed by the 28nm manufacturing struggles suffered by TSMC and other for-hire chip fabricators. Poor 28nm yields have adversely affected product availability for several of TSMC's partners, to the point that Nvidia and Qualcomm were rumored to be threatening to take their business elsewhere. Those dark days may be (mostly) behind, however, as TSMC's 28nm production skyrocketed in the second quarter.
Intel’s chip plant in Kiryat Gat, Israel, is about to be upgraded to 22nm production capability, the chip maker said at a news conference. The upgrade will see the company invest around $2.7 billion, including a $210 million grant that was recently approved by the Israeli government. The fab is expected to begin production on 22nm process technology in December, which is in keeping with the late 2011/early 2012 launch of Ivy Bridge processors -- 22nm die shrink of Sandy Bridge. A few months back, Intel announced that it would spend up to $8 billion on similar upgrades to four of its existing plants in Oregon and Arizona and the construction of a new 22nm fab in Oregon.
Exactly how low can Intel's Atom processors go? At least down to 15nm, according to Intel's latest roadmap. Think about that for a moment. At just 15nm, the manufacturing process is about the size of 60 atoms. By comparison, human hair measures 100,000nm in diameter. Pretty amazing, isn't it?
Intel's plans were revealed in a slide being shown at the Intel Developer Forum (IDF), which also showed 32nm and 22nm parts filling in the gap between 15nm and currently shipping 45nm Atom processors.
The slide is short on details, but interestingly it looks as though Intel will crank out more Atom chip offerings as the manufacturing process continues to shrink. At 15nm, Intel shows five chips each for netbooks, nettops, and consumer electronics, and six chips each for handheld "Z" series and embedded "E" series products.
Toshiba today announced it has begun mass producing NAND flash chips using a 24nm CMOS manufacturing process, representing the smallest geometry and highest density yet in NAND flash, the company said.
The announcement steals a bit of thunder from IM Flash -- a joint venture between Intel and Micron -- which said it would begin churning out 25nm-based NAND chips by the end of 2010.
"Toshiba leads the industry in fabricating high density, small die size NAND flash memory chips," Toshiba said in a statement. "Application of the 24nm generation process technology will further shrink chip size, allowing Toshiba to boost productivity and bring further enhancements to the high density, small sized products. The 24nm process products are also equipped with Toggle DDR, which enhances data transfer speed."
Toshiba says its latest technology has already been applied to 2 bit-per-cell 64Gb chips that are the world's smallest on a single chip (8GB), and will also add 32Gb and 3 bit-per-cell products fabricated on a 24nm process soon.
OLED displays are widely accepted to provide some of the best image quality money can buy. The problem is that it does take a lot of money to buy them. The current generation of manufacturing tech means that small OLEDs, like the one found on the Nexus One, are really at the upper limit of cost effectiveness. A start up called Kateeva wants to change that. They are developing a system for printing OLED displays.
Kateeva’s manufacturing process has been shown to be capable of printing 1.8 by 1.5 meter OLED displays. They estimate the costs to be roughly 60% of current methods. Don’t get too excited yet. The OLED printer is set to be tested by display manufacturers next year. Just imagine, in a few short years you may be tossing out your tired old plasma or LCD HDTV and buying an OLED version.
Development for 32nm is going well for Intel, so well that the chip maker has decided to axe its 45nm Havendale chips before they reached volume production and will make the move to the 32nm Clarkdale instead, according to DigiTimes. Havendale was originally scheduled to launch by the end of the year, but Intel will instead go forward with 32nm Clarkdale in the first quarter of 2010.
Citing sources at motherboard makers, DigiTimes says Intel also plans to mark several processors as EOL (end of life) in the second half of 2009 and through the first quarter of 2010. Among them will be the Core 2 Extreme QX9775, Core i7 940, and a bunch of Core 2 Quad, Pentium, and Celeron CPUs. The chip maker will also begin discontinuing both the Atom 330 and Atom 220 in April 2010.
Meanwhile, the sources say Intel plans to launch the Core 2 Quad Q9505S, a quad-core CPU designed specifically for all-in-on PCs.
Intel co-founder Gordon Moore once predicted that the number of transistors on an integrated circuit would double every 18 to 24 months, a prediction which has been famously dubbed Moore's Law. But according to market research firm iSuppli, the move to 18nm will signal the end of Moore's Law.
"The usable limit for semiconductor process technology will be reached when chip process geometries shrink to be smaller than 20nm, to 18nm nodes," said Len Jelinek, director and chief analyst, semiconductor manufacturing, for iSuppli. "At those nodes, the industry will start getting to the point where semiconductor manufacturing tools are too expensive to depreciate with volume production, i.e., their costs will be so high, that the value of their lifetime productivity can never justify it."
So when exactly will it happen? According to iSupply, in the year 2014. In 2007, Gordon Moore said his prediction could be upheld for at least another decade. Five years from now, one of them is going to be wrong.
When it comes to Moore’s law these days, it seems like everyone’s a cynic. However, now there’s one more reason to be optimistic about the future of miniaturization, as researchers have published a paper describing a lithography technique which may provide a new means of producing chip features smaller than 32nm.
The technique involves the use of quasiparticles called plasmons to focus light at an incredibly high resolution. Chris Lee at Ars Technica describes the technology: “A lens, based on plasmons, can be created by a set of concentric metal rings. The fields from the plasmons in each ring act in such a way as to create a tightly focused spot of light. In principle, these lenses could focus light tightly enough to create features about five to ten nanometers in size.”
The problem with plasmon lenses is that they must be positioned at just 20 nm away from the wafer. The scientists claim to have overcome this hurdle with their new technique, which uses air pressure to control the lens’s distance from the wafer.
Significantly, the new technique eliminates the need to create a new photomask for each revision to the chip, potentially lowering costs and speeding up development.
It might not be as well publicized as Micheal Phelps' race to 14 gold medals, but there's another kind of race going on in the chip industry, and that's to see who will be the first to reach 22nm. But it might not be Intel leading the way, and instead it looks as though IBM may be emerging as the front runner.
Unlike the path to 45nm and 32nm, getting to 22nm presents some significant challenges for chip makers, one of which includes getting the circuits "printed" in a process called photolithography. As IBM engineer Subu Iyer notes, "Once the wavelength of light becomes comparable to the size of the thing you're trying to print, things break down. The challenge is to use a light wavelength of 192 nanometers because 'extreme ultraviolet' radiation is still impractical."
Iyer went on to say that in terms of physics, getting to 22nm is a tall order requiring a tremendous amount of computation. To help with that, IBM has developed what it's calling the Computational Scaling (CS) initiative, which includes support from several of IBM's partners. If nothing else, this collaboration puts added heat on Intel, who downplayed IBM's foray into 22nm earlier this summer.
Might IBM beat Intel to the punch? Hit the jump and make your prediction.