By Ed Sperling
In the race to make smart phones more attractive, three conflicting marketing needs are coming into play—and raising some serious challenges for SoC architects and engineers.

First, this new breed of devices needs to last at least a day between charges using more energy-hungry applications such as games, GPS systems and Internet searches. And, incidentally, they also have to receive voice and video calls, text messages, stay connected to multiple signals ranging from 3G to WiFi to LTE, and be able to stream video without dying mid-movie.

Second, the marketing departments have deemed a thinner phone and tablet to be more desirable for consumers. If power engineers had their way these devices would be connected to a battery the size of a brick and none of these challenges would be a problem.

Finally, these devices need to have even better performance than in the past. The trend is toward mobility, and mobility requires the same kind of search speed that’s available on a PC—or at least close enough to it not to be deemed a problem—as well as the ability to stream video without a blip. That usually means more cores that can work in sync when necessary, and which can be put into sleep mode when they’re not being used.

There are examples of this new trend. While the new iPhone is expected to have better battery life in a similar package, the new devices showing up in the Android world look increasingly slick and remarkably thinner. Samsung’s new 1.5GHz dual-ARM-core Exynos processor is expected to fit the bill for more performance, including 3D graphics, using less energy. And Amazon’s new Fire tablet uses a seven-inch touchscreen for up to 7.5 hours per charge of video.

The question now is just how much smaller these devices can get while still delivering the same kind of area-performance-power tradeoffs and meeting these three requirements. Semiconductor engineers always have viewed limitations as a series of new challenges, but at some point they will be up against the dual barriers of physics and cost. And while the physical barriers can be overcome for a price, at some point that price may be too high.

By Ed Sperling

Faster throughput, but what about power?
The bottleneck in devices is always about throughput. In the mainframe days it was data read/write and retrieve speed on the enormous spinning magnetic disk platters, with all sorts of tricks ranging from faster rotation to striping across multiple disks. In the PC era it was about faster bus architectures and a chip’s ability to get data in and out of memory and/or storage.

Fast forward to the present and it’s about the ability to get data in and out of a device through a variety of I/O ports—mostly on a single chip but in the future potentially between chips using Wide I/O. The latest entry into this camp is Thunderbolt, which has a transfer rate of 10 Gbits/sec vs. 5 Gbits/sec for USB 3.0, the emerging standard. That’s pretty fast. It’s also pretty expensive, as in $50 for a cable that just hit the market.

So far, it appears that only Apple is selling these Thunderbolt cables. The technology was developed at Intel Labs, and Apple is pushing it as a slick differentiator for data-intensive applications such as video editing. According to Intel, Thunderbolt will transfer a full-length HD movie in less than 30 seconds and back up one year of continuous MP3 playback in less than 10 minutes.

What’s interesting is that it also can reduce the amount of power needed to drive these devices because it takes less time. Time is money in computing and in I/O, providing you can run the processor and the I/O at a low enough voltage. So far there’s no talk about the power savings, though, because no one wants to mention which computers and devices are actually going to use this transfer rate. But it should raise lots of questions beyond just the cost of the cable.

Atomic switches
The University of Southampton in the U.K., the Japanese National Institute for Materials Science and Hitachi are jointly developing a low-power logic system with instant on/off logic operations. The goal of the three-year research project is to build non-volatile logic systems based on “three-terminal atom transistors hybridized with nano-MEMS switches.

The university says the technology initially will become available as an integrated logic-memory chip for portable devices, and that ultimately the memory retention capacity will allow computers and mobile phones to shrink both in size and weight.

The big question is whether you’ll still have to plug it in every night.

By Ed Sperling
Apple’s iCloud announcement this week raises some interesting market questions, but if the move is successful it will have massive implications for power.

The goal is to put more compute power and storage into the server, where data can be processed far more efficiently. This is similar to the virtualization frenzy now underway inside of all major corporate data centers, which for years had a very inefficient policy of one application per server because they didn’t trust technology. A single point of failure was considered a bad idea, and chip density hadn’t reached the point where air could no longer cool server racks. In fact, some data centers have reached a limit on how much air can be pushed through their servers because the noise it creates is already at the industrial standards limit.

Cloud computing is the next consolidation scheme, whether it’s internal or external. The big concern there is security, which is the one most cited by efforts in the EDA world to put their tools on clouds. But given the steady stream of data theft—even at large companies such as Citibank and Sony—clouds are certainly no more at risk than the best-run and most expensive operations. In fact, they are probably much more secure than many corporate data centers because security breaches are a huge public embarrassment.

But the biggest gain may be on the side of the user of this data. A thinner client—or at least one in which more cores are turned off or in deep sleep most of the time—can save enormous amounts of energy. The I/O will have to be beefed up, along with the internal electronics to smoothly handle streaming downloads and uploads. And more people will have to upgrade their broadband connections to support higher upload and download speeds. But the amount of energy necessary to drive that can be consolidated and centrally managed, which can significantly extend battery life the way virtualization has lowered the utility bills (both for cooling and powering servers) inside of large corporations.

The big question inside the processor industry is whether Apple will use its own processors or continue to use Intel chips. But for Apple customers the real shift will be in the rationalization of processing—local processing where it makes sense and centralized processing where it doesn’t. This is a dramatic change and one that is timely because of the other pieces of the puzzle that are now in place—bandwidth, I/O standards, battery limitations and growing frustration over increased functionality and limited battery life.

By Ed Sperling

The Other 3D
Intel will roll out processors using tri-gate finFET transistors at 22nm, which it says will sharply lower the operating voltage, boost performance and reduce leakage.

Multigate transistors have been the subject of research for decades, most prominently at UC Berkeley, because they can be used to reduce current leakage and increase density. Going vertical allows more transistors to be loaded onto a piece of silicon, which in the case of a processor is particularly important because more transistors can translate into better performance.

Intel claims the new structures will improve performance by 37% at low voltages. The company said that makes it ideal for small handheld devices, a market where Intel has not done very well in the past primarily because its chips are considered power hogs next to those using ARM and MIPS cores. That statement alone caused ARM’s stock to plunge 7% as speculation mounted that Intel could replace ARM cores inside of some Apple devices. This is pure speculation, of course. Apple never talks about that stuff and Intel hasn’t even intimated that. ARM’s stock recovered rather quickly, too.

Still, most companies have shied away from finFETs because they are extremely difficult to manufacture and potentially can add to the design and manufacturing cost. Intel’s big advantage in this regard is that it still owns its own fabs and develops its own manufacturing process, something that is far too costly for all but a handful of chipmakers.

An alternative to 3D structures is ultra-thin body silicon on insulator, which is now being tested by IBM, STMicroelectronics, Soitec and Globalfoundries. And there is a possibility of mixing things up to include both. But the writing is on the wall—big changes are ahead, and Intel’s move is a first big step in that direction.

TI Pushes FRAM
Microcontrollers have been used for years to reduce power in devices through such developments as multispeed motor control and intelligent sensors, but the real battle of late has been inside the microcontrollers themselves. Companies in this sector have been playing leapfrog with power numbers taking priority over performance increases.

TI’s latest rollout includes an ultra-low-power FRAM, or ferroelectric RAM (previously written as FeRAM). This type of RAM uses 250 times less power than EEPROM-based microcontrollers, according to TI, and can be written at speeds of 100 times faster. FRAM is not a new technology. It was developed in the 1990s by Ramtron, and has been manufactured by Fujitsu for more than a decade.

Apparently major strides have been made in the pricing of this technology since then. TI’s microcontroller is priced at $1.20.

By Ed Sperling
The marketing wars of instructions per second have been replaced by battles over watts per hour. How many hours a portable computing device lasts between charges and how many kw/hr are consumed by servers has suddenly taken center stage, and it shows in some of the big buzz introductions over the past week.

Apple rolled out its new iPad 2 with better energy efficiency, particularly with 3G connectivity. With WiFi Apple is advertising up to 10 hours of surfing. With 3G it can surf for up to 9 hours.

SeaMicro introduced a 512-core server for enterprises based upon Intel’s Atom chip. One of the big costs inside of large enterprises these days is electricity, both for running the servers and for cooling them. If this catches on, it could substantially lower the cost of both for IT departments. It also could have an interesting impact on Intel’s revenue stream, considering the Atom chip is significantly less expensive than its mainstream processors.

And that’s assuming Intel wins the business in the first place. VIA just introduced its low-power Eden 2 x86-based processors, which are a head-to-head rival with Atom. The focus once again is power and efficiency, and the big issue for Intel is just how extra is the Intel brand worth to customers.

By Ed Sperling
AMD jumped into the low-power market with a new version of its Fusion chip for the tablet market, which the company claims can reduce energy consumption by 40%. That puts AMD squarely in competition with Intel’s Atom, ARM’s Cortex A9 and A15, a swath of MIPS chips aimed at Android, Apple’s A4 and probably some others that sources say will be produced for localized markets. AMD also rolled out an updated parallel processing development kit, which is absolutely essential for performance.

The U.S. Air Force is developing a new ultra low-power RF transceiver to preserve battery life in military sensors, including radar and infrared cameras. The less power drawn and the lighter these devices can be made, the smaller they can be designed and the longer they can be in the air.

Toumaz, a U.K.-based developer of low-power telemetry technologies, introduced an ultra low energy radio for wireless sensor networks. The company says the device can run at 1 volt using a single button-sized cell battery and consume less than 3mV of continuous power. That should make for some interesting application possibilities.

Apple filed a patent application for what it calls “Staggered Line Inversion and Power Reduction System and Method for LCD Panels.”

Translation: Apple is looking to cut the power draw for all of its systems—iPhone, iPad and notebook computers—by altering how power is applied to the lines in the display. Considering the display is the largest consumer of power in any device, which is why the default settings power down the display rather quickly, this could be a game changer for Apple. While it can save battery life, it also can be used to extend the time a screen is operative in a device for the same charge.

Apple clearly cannot sacrifice quality in this arena. Its whole push for what it calls a Retina Display is enough pixel density so that viewers cannot see the individual pixels. That greatly improves picture quality, but it also requires more power to drive it.

Apple added another patent application to its portfolio, as well. This one is for seamless switching between radio and local media. Smooth images and movement, it appears, run deeper than just the beautiful façade and image. While it’s questionable how much power will actually be save here, it does give a glimpse into what Apple considers its competitive advantages.

–Ed Sperling

By Ed Sperling
Chang-Gyu Hwang, national chief technology officer for South Korea, sat down with Low-Power Engineering to talk about the future trends in technology, global business and power. Prior to his current role, which was created by the Korean government in April, he ran the semiconductor business at Samsung, where he spent the last 20 years in top management positions. He also is the former CTO at Samsung. What follows are excerpts of that interview.

LPE: What do you see happening next in technology?
Hwang: Over the past 20 years we’ve gone from PCs to mobile and the Internet. That will go to the mobile Internet, which will dominate over the next couple years. Then we’re expecting some fusion type of industry. There is a lot of room to improve, technology-wise and business-wise, but to satisfy and raise customer demand there will have to be some fusion industry. Inside IT many technologies will be embedded. Biotechnology will be a future technology. So will nanotechnology.

LPE: How will these technologies be used?
Hwang: In green transportation systems, for example, every country has automobiles, express trains, battery technology and battery charger systems, a smart grid and even nuclear energy. All of these are related. We’re expecting these industries will converge. Korea is relatively late in the industrial revolution, but it is strong in IT, shipbuilding, nuclear power and smart grids. We are looking for synergy in different technologies. I’m spending a lot of time in the United States. It’s a big market, and there are interesting technologies. I’m looking for partnerships from an open innovation point of view.

LPE: There’s a huge emphasis on saving power these days. What will change and why, both from a technology and a business standpoint?
Hwang: Smart grids are about using energy more intelligently and optimizing it, both from the producer and the consumer point of view. Korea is doing both. Korea has a relatively good start. We’re educating engineers. From an industry point of view, we’re well balanced in terms of nuclear and other types.

LPE: You mentioned embedding technology and convergence. Can you elaborate?
Hwang: It started with semiconductors. Many functions are now embedded into a single chip. They’re more cost-effective and use less power. At Samsung we introduced fusion technology where we embedded SRAM logic into memory chips in the OneNAND chip. There was intelligence in the software. We improved speed, reduced the chip size, used less power and boosted overall performance. That chip is dominating all markets because of its many advantages and lower cost. That kind of trend will be more prevalent. But the whole concept of an SoC was introduced because other chips were too expensive. The future will be in the fusion of emerging technology into one chip in the mobile and multimedia area. That will happen in other industries such as energy savings for automobile applications and across other industries.

LPE: Will the boundaries change between who produces what?
Hwang: Korea is relatively strong in hardware. It has a good total solution for making components, with good engineering skills. It knows how to reduce cost while maintaining relatively high performance. In the United States, companies like Google, Microsoft and Apple have different strengths. They’re strong in business and technology. Google is good in search engines, but they need server, low-power and low-latency technologies. Korea is a major supplier of components for Apple. If you apply this to other industries, there are many opportunities for collaboration. The United States has a lot of creative ideas and is very good in software, but in Korea there is more application in new fields. The two working together will drive those industries.

LPE: How about China?
Hwang: China is another variable. They are not a first mover, but they are a very effective follower—especially with low labor costs and low-cost solutions. The United States and Korean collaboration is at a higher level than what China will offer. Korea spends a lot of money on R&D, and it’s strong in core technologies in several industries. At that level we can collaborate with the United States.

LPE: What is your big challenge?
Hwang: My goal is to become the No. 5 technology country by 2020. We also need to be a first mover in some technologies. I have to figure out which industries to target, from the early design and planning stage. These should be as unique as possible, whether it was started in Korea or whether it can be developed using open innovation from other sources.

LPE: Where does the government play vs. companies?
Hwang: When you look at major Korean companies like Samsung, LG or Hyundai, they have their own plan. They are investing in R&D for the next era. But from a national R&D point of view, there’s more risk taking involved. We try to find areas that are not easy to get into, consolidate whatever research is out there, and then create new projects. Eventually companies will get involved, but initially this will be a government effort.

LPE: So this is pure research?
Hwang: Yes. More long-term research and more fundamental. I call it R&BD—research and business development. Otherwise we will not consider it seriously.

LPE: Can companies afford this kind of research on advanced SoCs?
Hwang: Normally a company cannot afford to initiate a project by itself. But at this moment we need to collect all the ideas, not only within a company. It also has to come from research laboratories, institutes and university research. There needs to be consolidation from the initial stage to the planning stage. That’s the reason we are initiating this from a national level. That’s our role.

LPE: How much of this is done through universities?
Hwang: We don’t involve them much, but we are inviting leading institutes and universities to collaborate at the initial stage.

LPE: Is there a lot of opportunity in fixing the Internet?
Hwang: The Internet still needs a lot of new technology. It’s handling massive storage. It needs low power technology. But most consumers are still asking for more speed. Low power and high speed are a contradiction, so we need to bridge these two extremes. There is an opportunity for technology. There’s also an opportunity from a business standpoint for using new technologies.

LPE: Where is the United States lagging?
Hwang: The United States is quite late for biotechnology. But there are opportunities for merging biotechnology with information technology with things like protein chips or chips that can check the status of your heart. There are many applications to find disease in the early stage at a very low cost. Korea is one of the leaders in this kind of hardware, as well as mobile phones and TVs—which are all linked to each other.

This just in: A low-power transmission of NBC’s Washington, D.C., broadcast via TV translators is going away. Residents in Virginia received a notice from Qualcomm—which specializes in low-power chips—that the service interferes with new wireless services. Thank the Northern Virginia Daily for reporting this one.

Freescale has jumped into the gallium arsenide world with four chips optimized for base-station equipment. Yes, it really is true. Bulk CMOS is running out of steam for many applications at advanced nodes.

Apple’s iPad reportedly is getting a memory boost. AppleInsider reports that at the heart of the device is an ARM processor. That explains the 10 hours of battery life.

By Ed Sperling
Intel rolled out an ultra-low power Atom processor for smart phones and tablets. The company claims it can cut power consumption during idle by more than 50 times over previous versions to as little as 100 microwatts, but mileage will vary. In some cases it will vary greatly. For example, it will only offer about 4 to 5 hours of browsing, and the numbers go down from there depending upon how much graphics-intensive computing is involved. The bottom line: This chip sleeps well and it wakes up with a roar, but beware of what you feed it.

By way of comparison—and competition—ARM’s Cortex M0 processor idles at 85 microwatts, but the software stack is significantly thinner. Legacy has its advantages–namely its ability to run everything under the sun–and its disadvantages, which are namely the same as the advantages.

AMD is pushing into the low-power space, too, although it’s not exactly clear where the company’s embedded chips will play. At the very least, it will be more power hungry than a smart phone. The new chips will have power envelopes as low as 8 watts, which isn’t something you want to stick in your pocket. It may be something you want to include in a tablet computer, however.

Apple bought Austin-based Intrinsity, which makes design tools for turning up the clock on processors without increasing the power. Industry insiders say the real reason was the engineering talent. Speculation is rampant about which companies Apple will displace as it continues developing its SoCs—and why. Incidentally, Apple’s Web site has no mention of the deal.