In his presentation at GigaOM’s Structure conference on Wednesday, Feldman explained that the data center is not only the cloud, it’s providing the value for most of the phones, tablets and myriad devices we carry every day.

“The demand for compute has left the client side and moved into the data center,” said Feldman. “Over a three-year period we went from 3 percent to a third of the U.S. population owning a tablet … We now spend hours and hours a day in the cloud where before, we were on the couch.”

This change means we’re not just changing computing, but also networking and storage. He said IT has become software-defined. And the building blocks aren’t the only thing changing, the buildings where we house the compute is changing as well. Even where we put those buildings is changing.

“We used to put data centers in urban environments but where do we put them now? In Eastern Washington or along river banks in Oregon to take advantage of lower power,” said Feldman.

“The data center now does the compute for the client side. Millions of millions of users each with parallel work. We don’t ask it to do CAD/CAM …. the vast majority of the work we ask it to do is simple parallel work for the client side. And that work is very different.”

It’s not about CPU performance, which means that work requires a different type of processor. “And in the future we believe it’s going to be an ARM processor,” Feldman said.

So Feldman called for the industry to rethink how it designs servers to make them more efficient. The server world should also embrace open source hardware like what the Open Compute Project wants to offer. He left the audience with the thought that in the 60 year history of computing, smaller, higher-volume parts have always won. That used to be x86-based processors. But in 2012 more than 8 billion ARM CPUs were shipped, more than twenty times the x86 volume.

It’s too bad Feldman didn’t spend some time talking about how AMD plans to adapt to the realities of an ARM-based chip world, where dozens of vendors have the ability to design and build ARM-based chips. That’s a big shift from building x86 chips that only two vendors can sell.

Check out the rest of our Structure 2013 coverage here, and a video embed of the session follows below:

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The role of converged infrastructure in the data center
• How tomorrow’s mobile-centric data centers will look
• Webscale and cloud are changing the server value chain. Who wins?

The news of the new EU industrial strategy came just a couple of days after the Geneva-headquartered, French-Italian manufacturer STMicroelectronics launched its own three-year project, worth €360 million ($463 million), aimed at creating a European microelectronics design ecosystem based around its fully-depleted silicon-on-insulator (FD-SOI) manufacturing process.

Some in the industry, such as chipmaker GlobalFoundries, have previously urged European authorities to back electronics manufacturing on the continent in order to counteract the vast influence of Asia and (to a lesser extent) the U.S. in this field.

Cheaper, faster, smarter

The European Commission’s strategy, announced on Thursday, is intended to make chips cheaper, faster and smarter. It will concentrate on shoring up three existing electronics clusters, namely those in Dresden (Germany), Eindhoven (Netherlands) and Leuven (Belgium), and Grenoble (France). Connections will also be made with other clusters such as that in Cambridge (UK), which is big in the wireless sector.

“I want to double our chip production to around 20 percent of global production,” Digital Agenda Commissioner Neelie Kroes said in a statement. “I want Europe to produce more chips in Europe than the United States produces domestically. It’s a realistic goal if we channel our investments properly.”

As per usual, this isn’t a simple public cash splurge. €5 billion in public funds – 30 percent from the EU with the rest coming from national and regional funds – will go to R&D, in order to help stimulate the sector. Overall, the Commission says, industry has indicated it will stump up €100 billion over the seven years: €15 billion in capital expenditure and €85 billion in operational costs.

The kind of electronics we’re talking about could be used in desktop and handheld computers, but the main thrust is for embedded systems and “internet of things” devices, from sensors and smart grids to new healthcare technologies. As Kroes said in a speech, “this isn’t about computers.”

Targeting these areas plays to Europe’s strengths. According to the Commission, Europe already pumps out half of global automotive electronics, 40 percent of electronics used in energy applications, and 35 percent of those used for industrial automation – this will be a reference to the output of companies such as Bosch, which are hugely active despite often being somewhat under-the-radar. Then we also have smaller manufacturers working in high-growth niches, such as health implants and sensors.

And jobs?

The purpose of all this is to make Europe less reliant on manufacturers outside the continent, but job creation is also a major factor. The Commission reckons the European electronics industry already employs 200,000 people directly and supports a further million jobs indirectly.

That said, the Commission also pointed out in its statement that demand for skilled workers in these fields is higher than supply – if this whole strategy is to work, the implication runs, Europe will need to attract more skilled workers. The statement talks of coordinating public efforts across Europe. Perhaps that will mean tweaking immigration rules: something the U.S. tech sector is also heavily vocal about these days.

Meanwhile, STMicro’s push – called, incredibly, “Pilot Lines for Advanced CMOS Enhanced by SOI in 2x nodes, Built in Europe” (Places2Be) – also takes place in the context of a wider European project, the nanoelectronics-focused ENIAC. In a briefing note accompanying Thursday’s announcement, the Commission insisted that ENIAC and ARTEMIS (another project focusing on embedded computing) had been a success, and that the new drive did not denote failure of those two schemes.

The Commission said the new joint undertaking would build on “lessons learnt” from ENIAC and ARTEMIS while providing a “simplified funding structure”.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Google and the Ghost of Silicon Valley Past
• Key technologies for the smart city
• Will Standardizing the Cloud Cause Clarity or Confusion?

Well, those startups now have seriously heavyweight competition in the form of LogMeIn, the remote connectivity specialist, and ARM, the British firm whose low-power chip designs underpin the vast majority of mobile devices, and which is now competing with Intel to own the IoT space.

LogMeIn has just launched its own PaaS for the internet of things, calling it Xively (the beta version was known as Cosm). And developers wanting to start creating connected devices on this platform are being offered the Xively Jumpstart Kit, which combines Xively with ARM’s mbed platform, for building devices using ARM’s microcontrollers. With this kit, the companies promise, developers can “rapidly progress from prototyping to volume deployment”.

Xively is based on LogMeIn’s Gravity infrastructure – the same one used to support the company’s cloud storage offering, Cubby — and it comes with development tools for writing and prototyping services, a provisioning engine for deployment and a scalable management console. It supports real-time messaging and directory and data services, as well as analytics, and it uses a “pay-as-you-grow” pricing model that should make the platform attractive to startups.

The directory services extend to a “commons” named the Xively Connected Object Cloud, through which different companies’ devices can interconnect. According to LogMeIn, a “fundamental philosophy” baked into the Xively terms of service states that “customers own their data and can choose whether or not to share all, part, or none [of] it.”

A showcase page for the platform shows early projects built on Xively that include the Visualight smart lightbulb and even some of the post-Fukushima crowdsourced radiation-monitoring efforts (which used an earlier iteration of the platform, called Pachube at the time).

While the Xively Jumpstart Kit should help inventors and developers gravitate in ARM’s direction, it’s not like Intel is sleeping. Intel said in February that its own Intelligent Systems Framework – a set of specifications for connecting, managing and securing IoT devices – had been used to support more than 50 products. The company also released new software tools for, you guessed it, reducing time to market.

Although ARM does benefit from a much broader ecosystem than Intel, it’s too early to call that race. However, those startups trying to build their own PaaSes for the internet of things had better get a move on. LogMeIn’s offering is already pretty mature for this space and, given the momentum rapidly building behind the IoT movement, its timing is exquisite.

(Incidentally, the internet of things is a subject that will be discussed at our Structure 2013 event in San Francisco on 19-20 June, so be there.)

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

While multiple 802.11ac devices and routers have recently hit the market, Quantenna is delving deeper into the 802.11ac standard, using multiple input-multiple output (MIMO) smart antennas to deliver four parallel streams of data over the same frequencies while aggregating both the 2.4 GHz and 5 GHz Wi-Fi frequencies to create connections as fast as 2 Gbps.

The company already sells its 802.11n chips to router makers like Netgear, but the deal with STMicro could expose its technology to a broader range of devices beyond home and business wireless routers. STMicro sells processors and communications chips to the many industries including automotive, entertainment and security markets. Through its joint venture with Ericsson, it also makes integrated silicon for mobile handsets.

The focus on multi-gigabit Wi-Fi might seem a bit ridiculous given that the typical home broadband connection could never match such speeds, but there are plenty of signs that such high-capacity networks will be necessary. As home networks become more complex and sophisticated, smart TVs and entertainment hubs will be pushing big video files around the house, which will require workhorse local area networks.

And while gigabit broadband connections are only just emerging in the U.S. households, public Wi-Fi hotspots using high-speed backhaul links are gaining popularity among carriers, municipalities and other service providers. Such shared public connections could allow hundreds of people to access the same Wi-Fi nodes, taking the data load off cellular networks.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• How to deliver the next-generation web experience
• How tomorrow’s mobile-centric data centers will look
• Research In Motion: future scenarios for its fate

Welcome to the curse of the automotive industry: the lead time on new car designs and manufacturing schedules mean that the technology you’re buying today was developed years earlier. What’s more, that technology effectively becomes locked down in your vehicle. As soon as your drive off the lot the connected car system you have is the one you’re stuck with for years. (For more details on the connected car technology check out our infographic.)

Mobile processor maker Nvidia, however, is proposing a solution to that problem: why not make an upgradable connected car system. We “upgrade” our smartphones and tablets every year or two by buying completely new devices, but that’s not really an option for an automobile.

However, with processors based on its Tegra designs, Nvidia wants to empower automakers to build cars that not only have top-of-the-line computing components when they roll off the lot but also can be upgraded periodically during their long lives.

In short, Nvidia wants to help automakers make connected cars that never become obsolete.

Meet Jetson

According to Nvidia Director of Automotive Danny Shapiro, the company designed its automotive processors, called Visual Computing Modules, around a flexible framework that allows automakers to work future processor technology into what are typically three-year long development cycles. Rather than design a connected car system years away from production using today’s chips, engineers can design tomorrow’s cars using tomorrow’s chips, Shapiro said.

Tesla Model S

That program is already seeing some pretty significant results, Shapiro said. Within a month of shipping in Google’s flagship tablet, the Nexus 7, the Tegra 3 debuted in the Tesla Model S, powering its impressive infotainment system (along with a separate Tegra 2 processor to handle the instrument cluster).

That solves the first problem – making an infotainment system that’s not obsolete before it hits the show floor. Solving the next problem — making a connected car system that keeps up with the pace of consumer electronics innovation — is much trickier.

To tackle it, Nvidia recently launched a new automotive architecture called Jetson, which tries to solve more than just the problem of obsolescence. First, Jetson is powerful, incorporating Nvidia’s pixel-crunching graphics processing units alongside its Tegra VCM chips. Nvidia is hoping that its silicon won’t just be the brains of your infotainment system but an extra set of eyes on the road.

Danny Shapiro

Nvidia wants to power the advanced driver assistance systems (ADAS) emerging in the next-generation of cars, Shapiro said. Moving beyond adaptive cruise control and proximity detection, cars will eventually sport omnidirectional cameras that will “see” the road in all directions and possibly even scanning lasers that can model a vehicle’s surrounding in 3D. The art of processing image and spatial data just happens to be Nvidia’s sweet spot.

In addition, Nvidia has crafted Jetson to be a development platform that builds on its earlier work with its VCM chips. “Automakers can simulate future designs,” Shapiro said. “They can get their development done now, preparing for the next-generation chips and next-generation car apps.”

Finally, Jetson is modular. The core processing unit is designed to be swappable. That means an automaker can easily incorporate the latest and greatest version of Jetson into their existing connected car and infotainment systems each successive years. It also means, Shapiro said, that one day we could upgrade our car’s dashboard computers much like we’d upgrade an old PC.

Pimp my ride’s CPU

Unless you’re one of those folks that can afford to buy a new car every time the ashtrays get full, chances are any new vehicle purchase is going to be a long-term investment. Six years is not an unreasonable time to spend driving the same car, but that’s an eternity in the world of consumer electronics. Six years ago, what we now think of as a smartphone didn’t exist, and no one had yet developed many technologies we now take for granted such as speech-powered virtual assistants, 3D mapping and location-based social networks.

Many automakers have decided that trying to keeping up with the day-to-day advances of that technology is an exercise in futility and have built their connected car strategies around the smartphone itself. Ford and Chevy, for instance have designed their connected infotainment systems as extensions of the driver’s handset. So as the smartphone becomes more powerful, so do their cars’ dashboards.

An upgradable CPU would solve part of that problem, but not the whole problem. It doesn’t matter if your new car dash computer can process hi-rez images in real time if it doesn’t have the sensors to collect those images.

But the auto industry is trying to solve that problem as well. Ford has launched an open-source hardware program called OpenXC, which could let us upgrade components like heads up displays and sensor arrays in our future cars.

I’m not saying you’ll be able to turn your old jalopy into KITT from Knight Rider, but who knows? One day maybe we could customize our cars so they behave like new even if they don’t look like new.

Mouse car image courtesy of Shutterstock user Mopic

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• CES 2012: a recap and analysis
• Why Google Android’s Electric Vehicle Deal With GM Matters
• The fourth quarter of 2012 in cleantech

Qualcomm’s baseband chips and integrated applications processors have long supported all cellular technologies and bands, but they’ve never been able to produce a truely global phone. That’s because the other hardware components of the phone have never supported the same breadth of frequencies. Consequently, LTE devices have always been region-specific. Even Apple had to can its usual of strategy of producing a single global device and design three different variants of the iPhone 5.

Wireless Intelligence projects 38 distinct LTE bands in 2015

But Qualcomm’s new front-end chip, called the RF360, can supposedly support up to 40 LTE bands, both the time division and frequency division variants of LTE and all legacy 3G and 2G technologies to boot. Qualcomm created a 3D chip that utilizes a separate sophisticated antenna tuner that can latch onto any of 40 LTE frequencies between 600 MHz and 2.7 GHz – pretty much the entire range of current 4G spectrum.

This technology will be a key element in creating the future universal LTE phone, but — before you get too excited — it’s not the only necessary element. Other components in the RF chain such as the antenna will need to catch up before a device could feasibly work on every LTE network in the world. Smart antenna makers like SkyCross and Ethertronics have designed antennas that can support a dozen bands or so, but they’re not quite ready for 40.

But Qualcomm EVP and co-president of mobile and computing technologies Murthy Renduchintala said that the RF360 would allow device makers to make far fewer variants of their phones. In order to cover all of the world’s LTE networks, a vendor is faced with the prospect of designing as many as 10 different devices. The capabilities of RF360 could cut that number down to as few as three, he said.

“There will always be more problems to solve,” Renduchintala said an interview with GigaOM. “What we’ve done here is remove one of the most enormous obstacles.”

One of the problems this technology could overcome is the 4G fragmentation problem that’s already emerging in the U.S. All four of major operators are deploying LTE on different frequencies, while the rural and many regional operators are off on their lonesome in a neglected portion of the 700 MHz band. Clearwire isn’t just launching LTE on it’s own 2.5 GHz band, it’s the only U.S. carrier using TD-LTE. Maybe the RF360 can’t yet produce a global 4G phone, but it could produce a universal phone for the U.S. — and maybe ensure that smaller operators aren’t left out of the 4G revolution.

Also, Apple could conceivably use the technology to combine all of its iPhone 5 variants into a single device, but it still wouldn’t have a universal iPhone. Apple’s three iPhone models still leave out a good deal of the world’s current LTE frequencies, and with current technology it couldn’t cram 30 or 40 bands into a single device.

But wait, there’s more! Qualcomm has also introduced its own envelope tracking technology into the module, which will help sate LTE device’s notorious hunger for power. Envelope tracking helps control the enormous energy spikes inherent in LTE, reducing device power consumption by as much as 30 percent. Other silicon vendors like Broadcom and Altair Semiconductor have announced support for envelope tracking in their new super-chips, but support doesn’t necessarily equate inclusion. Qualcomm developed its technology in-house and is embedding envelope trackers directly into its future RF products.

Renduchintala said the module has already begun sampling and is in the hands of phone manufacturers. The first commercial devices with the new capabilities should start appearing in the latter half of the year.

This post was updated at 12:05 PM, Thursday, with new information on the implications of the technology for the iPhone.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• Mobile 2012 and beyond
• CES 2012: a recap and analysis
• How to manage the signaling storm in 2013

As I wrote earlier this week, LTE-Advanced is a much-abused term, used increasingly throughout the industry to make LTE products seem much more significant than they actually are. Carriers and vendors have latched onto a single technique in LTE-Advanced standard to justify their use of the moniker.

Altair is no exception, though to be fair its new super-chips are more advanced that others. It’s incorporated into its designs two techniques from the LTE-Advanced standard: carrier aggregation, which bonds together disparate swathes of spectrum into one big super-carrier, and enhanced inter-cell interference coordination (eICIC), which will allow small cells and big macrocells to coexist in the same airwaves.

What’s more, Altair co-founder and marketing VP Eran Eshed said that whatever LTE-Advanced techniques its chips don’t support today will be supported in the future through software upgrades. “In contrast to competitive solutions, Altair’s solution is based on a very advanced and powerful SDR (Software Defined Radio) architecture which means that we have the ability to deploy a chipset and upgrade its features as standards evolve,” Eshed told me via email.

Perhaps the most notable detail in Altair’s new chip specs, though, is its use of envelope tracking. It’s an obscure little technology being developed by companies like Nujira and Quantance, but envelope tracking has the potential to significantly boost 4G-device battery life by tempering LTE’s innate power hunger. Eshed wouldn’t tell me whose envelope tracking technology Altair is using, but this is the first implementation of the technology I’ve seen in a chipset.

Correction: I was wrong about Altair being the first to support envelope tracking. Broadcom’s new LTE chip, announced last week, also supports the new technology.

Photo courtesy of Shutterstock user alphaspirit

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The big theme of MWC: How to live in a connected world
• Updated: Forecast: global mobile subscribers, 2010-2015
• Sector RoadMap: Social customer service in 2013

This week LSI announced its first ARM-based chip for the mobile base station. You thought Nvidia and Qualcomm’s quad-core smartphone processors were impressive, well LSI is embedding 16 ARM Cortex A15 cores, along with LSI’s networking accelerators and ARM’s low-latency CoreLink interconnect technology, onto a single 28-nanometer chip.

The chip family is designed for base stations of all sizes, scaling from the macrocell down to the picocell, making similar to the flexible and modular platforms offered by competitors Texas Instruments (TXN) and Freescale. Both Freescale and TI have begun incorporating ARM cores into their base station chips, though neither one is a complete ARM convert. Freescale leans heavily on the PowerPC architecture, while TI is pairing ARM cores with its bread-and-butter digital signal processors (DSPs). But ARM is definitely taking bigger and bigger strides into the mobile network with its increasingly powerful but energy-efficient silicon designs.

One company that’s hoping to join ARM within the guts of the mobile network is Intel, which is no stranger to skirmishes with the U.K. silicon giant in the infrastructure market. Intel is trying to establish a foothold for itself in the emerging technology cloud-RAN (RAN stands for radio access network). Cloud-RAN would separate the base station from the tower and move baseband processing into the cloud.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The fourth quarter of 2012 in cleantech
• The future of mobile: a segment analysis by GigaOM Pro
• How tomorrow’s mobile-centric data centers will look

But Huang didn’t have the same enthusiasm for smartphones, reflecting the fact that Nvidia has had trouble penetrating that potentially lucrative market despite the attractiveness of the Tegra line. Why? Nvidia offers the two major silicon components necessary to power any smartphone. It has the quad-core Tegra 3 processor itself — which has made it into a handful of high-end smartphones like the HTC One X – and thanks to its acquisition of Icera in 2011, it has the radio modem necessary to connect the phone to the network.

But what Nvidia doesn’t have is an integrated applications processor and modem, which Huang readily admitted is its key impediment. Here’s an excerpt from Huang’s comments at Nvidia’s earnings call (you can read the full transcript at Seeking Alpha):

“There is no standalone modem business anymore and in many of these new 4G connected device marketplace, an integrated approach is necessary and that’s the reason why we bought Icera and that’s the reason why we’re investing in LTE.

“… with an LTE modem, the Tegra processor and our software capability, we will be able to address a much larger phone opportunity going forward. And so we’ll have some phone success this year, but we’re not expecting to have a whole lot of phone design wins until we engage the market with LTE.”

There are a lot of reasons why phone makers prefer integrated designs. Fewer components mean fewer suppliers and fewer parts to cram into the limited space of a smartphone. Fewer components mean less complexity in design and ultimately a lower cost to manufacture. But one of the biggest reasons 4G smartphone makers have become so keen on integrated chips is because they drain less power than a split-silicon solution. By sharing the same chipset and reducing the overall number of circuits, the design takes much less of a toll on the battery.

That’s a big concern for smartphone makers as the first generation of LTE handsets proved to be battery killers. Last year before Mobile World Congress, Qualcomm SVP of Product Management Raj Talluri predicted that the power-saving benefits of Qualcomm’s then-new integrated Snapdragon S4 processors would give it a tremendous advantage in the LTE smartphone market. So far his prediction has proven true. Qualcomm has dominated the market.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• CES 2012: a recap and analysis
• The future of mobile: a segment analysis by GigaOM Pro
• Tablet market to hit over 377 million units by 2016

What Gartner’s calculations found mean that Apple is no longer the world’s biggest spender on semiconductors. However, lower spending doesn’t necessarily mean Apple is buying fewer chips. It could be that Apple is getting better deals on them, or it could have to do with the type of chips it is buying versus what Samsung is. Here’s how the two rank among the rest of the world’s biggest spenders on semiconductors:

Because they are pricier, PC chips are still providing the bulk of the demand for semiconductors, according to Gartner. But as that market continues to shrink, it led to an overall down year for chip purchases. Demand for chips actually shrank 3 percent in 2012, but was helped out by Samsung and Apple’s seemingly insatiable appetite for semiconductors. Together the two accounted for 15 percent of semiconductor spending as they duke it out over the future of consumer mobile devices.

Related research and analysis from GigaOM Pro:
Subscriber content. Sign up for a free trial.

• The future of mobile: a segment analysis by GigaOM Pro
• Mobile Q2: Smartphone growth surges; iPad’s rule continues
• What Google’s Honeycomb Means for Apple and Microsoft