Netflix opens the floodgates on Super HD streaming support

Sorry, I could not read the content fromt this page.

View the original article here

Lapsing of Moore's Law opens up opportunity in chip design

As the trend famously codified by Intel co-founder Gordon Moore—that the number of transistors on an integrated circuit would double every two years—seems to be flagging, one top engineer suggests that it is time to rethink chip design to buy performance increases.

Instead of just focusing on reducing chip size and cranking up processor speeds, engineers need to look at making tweaks or possibly even change fundamental microprocessor architecture to ensure chips are faster and cheaper to produce, said Robert Colwell, director of the microsystems technology office at DARPA (Defense Advanced Research Projects Agency), in a talk on Monday at the Hot Chips conference in Stanford, California.

Colwell dismissed arguments that Moore’s Law will continue to hold, and said engineers should give serious thought to the design and economics of chip making.

Robert Colwell, director of the microsystems technology office at DARPA

“I pick 2020 as the earliest ... when we can call it dead. That’s only seven years away,” Colwell said. “I’m thinking 7-nanometers. You can talk me into 2022, you might be able to even talk me into 1-nanometer. But you’re not going to into 1-nanometer... I think physics dictates against that.”

There have been different interpretations of Moore’s Law, with the most common one being that the number of transistors on a chip will double every two years, which will make chips faster. But Colwell tried to clarify the definition, saying that in 1965 Intel’s Gordon Moore focused more on the economics related to cost-per-transistor, which would drop with scaling.

“What it really is ... if you’re going to integrate a lot of components on a chip, there’s an optimal place on that curve where you should do that. You can go beyond the optimal point, but it will cost you in terms of price per component,” Colwell said, adding there’s a sweet spot where maximum profit can be eked out assuming sales are relative to the number of chips made.

It is true that beyond a certain geometry it will hard to make chips smaller, but Colwell said economics, and not physics, would ultimately end Moore’s Law. The day chip makers can’t get return on the billions invested in making chips smaller is the day Moore’s Law will break. Instead of waiting for chip economics to crash, innovation should start now.

Chips now have billions of transistors, and the ability to push clock speed and performance will reached its limit, and after silicon engine stalls, small tricks and incremental tweaks like power gating and turbo will happen for a short time to improve chips. But chip designers should start early, and an effort like changing fundamental chip design could help before and after Moore’s Law ends.

One approach to consider might be separating the instruction set architecture, microarchitecture, circuits, functional blocks and other parts now integrated onto a chip, and tweaking them for specific applications, Colwell said.

“I think the end of Moore’s Law opens the door to designing special-purpose things again,” Colwell said, adding that in the 1970s, one could make specialized floating point arrays with vector processors. DARPA is doing research in the areas of quantum computing, nanotechnology and distributed computing, he said.

Researchers at universities and chip companies are also looking at new materials to replace silicon, and also advanced manufacturing technologies. Colwell said that the new technologies are far from practical implementation, and chip makers will have to rely on technologies like CMOS, which has no practical replacement in sight.

“CMOS is really good stuff,” Colwell said. “There are [only] two or three [new technologies] that are promising at all. It’s just hard to beat CMOS.”

Some ways to buy performance in the meanwhile could be through the use of new materials, photonics, optics and 3-D stacking, in which transistors are placed on top of each other.

Outside of the computer industry, the auto industry will feel the biggest impact of the end of Moore’s Law, Colwell said. The last 30 years of innovation in cars such as navigation systems, antilock breaks, guidance systems and others have all been driven by semiconductors.

“I think that’s really cool but all of it is based on computers. If we stall out, what are they going to do differently from generation to generation?” Colwell said. “I think they have been living off the electronics for the last 20 to 30 years, and if we don’t continually feed them huge increases, it’s not clear what they will do next.”

Colwell also threw a dart at his former employer, Intel, where he was the chief architect for all Pentium chips.

“Intel is terrible at anticipating. They don’t look down the road and say ‘five years from now the rules will be different, I need to react today, I’m going to put some bets on the table’. There’s some of that, but not a lot. But what they are really good at is reacting,” said Colwell.

Agam Shah is a reporter for the IDG News Service in New York. He covers hardware including PCs, servers, tablets, chips, semiconductors, consumer electronics and peripherals.
More by Agam Shah, IDG News Service

View the original article here

Rapidly growing DigitalOcean opens second NYC data center

Continuing its rapid expansion, cloud service provider DigitalOcean has opened a second facility in the greater New York metropolitan area -- a data center located in Google's gargantuan Manhattan facility.

The additional location is another sign of the fast-growing success the company has had in offering a low-cost, easy-to-use IaaS (infrastructure as a service).

"There's a huge market for developers out there who just want a server online as fast as possible, and that is what DigitalOcean provides," said Mitch Wainer, a co-founder and the chief marketing officer of DigitalOcean.

Although Wainer disputed that DigitalOcean's first New York City-area data facility, actually located in northern New Jersey, ran out of space, new users who have signed up in the past month or so were instructed that they could not host their accounts there, and were instructed to choose DigitalOcean's outlets in Amsterdam or San Francisco.

In any case, the company's user base is growing quickly. The company has accrued more than 35,000 customers since its launch in 2011, and its revenue is increasing by about 30 percent to 50 percent per month, Wainer said.

Internet service market analysis firm Netcraft noted that in June 2013 DigitalOcean had more than 7,000 Web-facing servers, a dramatic increase from the prior December when Netcraft counted only 100 DigitalOcean Web-facing servers.

This newest set of servers is being colocated with the Telx data center company, which leases space in Google's 3 million-square-foot building. The building takes up an entire block on 8th Avenue, in the Chelsea district of Manhattan.

DigitalOcean was founded with the idea of offering cloud hosting that would be easier to set up than what was offered by the market leader, Amazon Web Services (AWS).

"There was a huge void in the market," Wainer said. Amazon and Rackspace are focused on enterprise-scaled businesses, but businesses in the lower-tier to midlevel market, which don't require advanced feature sets and don't have full engineering teams to manage infrastructure, are completely ignored, he said.

With its wide range of services, Amazon Web Services recently launched a series of tutorials that can guide users into setting up virtual machines.

In contrast, DigitalOcean keeps its deployment options relatively simple -- and inexpensive. Accounts are billed either on a monthly or an hourly basis. The least expensive "droplet," which is what DigitalOcean calls a virtual machine, costs US$5 a month, or about $0.007 cents per minute, and features a single CPU core between 2GHz and 3GHz, 512MB of working memory, 20GB of storage and 1 terabyte of data transfers.

In contrast, the least expensive month-to-month option on AWS, a T1.micro image running Linux, would run about $14.96 per month, and does not come with any storage.

The company claims that users can set up a droplet, or resize it to another pricing tier, in less than a minute. It offers virtual machine images running basic versions of either the Ubuntu, Debian, CentOS, Arch or Fedora Linux distributions.

The company is using a number of new technologies for competitive performance, including SSDs (solid state drives), which can boost I/O performance. For virtualization, DigitalOcean uses Red Hat's KVM hypervisor.

Just as Microsoft is doing with its Windows Azure service, DigitalOcean is initially pursuing the market of developers who need modestly priced machines to develop and test code.

Wainer points to the rapidly growing user base of GitHub, an online source-code repository, as proof of the potential size of the developer market. Founded in 2008, GitHub now hosts 7.7 million repositories from 3.3 million users.

"You can see from that market size alone, that this is a billion-dollar market opportunity," Wainer said.

While DigitalOcean has been courting the developer market, it is also suited for startup businesses as well, Wainer said. For instance, one startup using DigitalOcean is the RSS reader service NewsBlur, which has accrued a significant user base after Google shut down its own RSS reader.

Not that the company hasn't had a few bumps. Last month, one user found that the service was using identical SSH fingerprints for multiple Ubuntu droplets generated within a single account, which was a significant security vulnerability. The company has since fixed the problem.

"That's one of the things about being a startup -- you can't be perfect from the get-go," Wainer said, though adding, "Whenever there is an issue, we've been responsive."

DigitalOcean, which is based in New York City, now has 30 employees, though it is looking to hire network engineers and Linux system administrators. It is also expecting to get a significant investment from an outside party shortly, which will be used "to drive new features and acquire capacity in additional regions," Wainer said.

The company is planning to offer virtual private networking and object storage, similar to Amazon's Simple Storage Service (S3). It will also broaden its array of off-the-shelf virtual machines, including those with WordPress or other popular applications installed and ready for use.

Joab Jackson covers enterprise software and general technology breaking news for The IDG News Service. Follow Joab on Twitter at @Joab_Jackson. Joab's e-mail address is Joab_Jackson@idg.com

Joab Jackson covers enterprise software and general technology breaking news for the IDG News Service.
More by Joab Jackson

View the original article here