Using clouds of ultra-cold atoms and a pair of lasers operating at optical wavelengths, researchers have reached a quantum network milestone: entangling light with an optical atomic coherence composed of interacting atoms in two different states. The development could help pave the way for functional, multi-node quantum networks.

The research, done at the Georgia Institute of Technology, used a new type of optical trap that simultaneously confined both ground-state and highly-excited (Rydberg) atoms of the element rubidium. The large size of the Rydberg atoms – which have a radius of about one micron instead of a usual sub-nanometer size – gives them exaggerated electromagnetic properties and allows them to interact strongly with one another.

A single Rydberg atom can block the formation of additional Rydberg atoms within an ensemble of atoms, allowing scientists to create single photons on demand. Georgia Tech professor Alex Kuzmich and collaborators published a report on the Rydberg single-photon source in the journal Science in April 2012, and in a subsequent Nature Physics article, demonstrated for the first time many-body Rabi oscillations of an atomic ensemble.

In the new research, the state-insensitive trap allowed the researchers to increase the rate at which they could generate photons by a factor of 100 compared to their previous work.

“We want to allow photons to propagate to distant locations so we can develop scalable protocols to entangle more and more nodes,” said Kuzmich, a professor in Georgia Tech’s School of Physics. “If you can have coherence between the ground and Rydberg atoms, they can interact strongly while emitting light in a cooperative fashion. The combination of strong atomic interactions and collective light emissions results in entanglement between atoms and light. We think that this approach is quite promising for quantum networking.”

The research was reported June 19 in the early edition of the journal Nature. The research has been supported by the Atomic Physics Program and the Quantum Memories Multidisciplinary University Research Initiative (MURI) of the Air Force Office of Scientific Research, and by the National Science Foundation.

Generating, distributing and controlling entanglement across quantum networks are the primary goals of quantum information science being pursued at research laboratories around the world. In earlier work, ground states of single atoms or atomic ensembles have been entangled with spontaneously-emitted light, but the production of those photons has been through a probabilistic approach – which generated photons infrequently.

This spontaneous emission process requires a relatively long time to create entanglement and limits the potential quantum network to just two nodes. To expand the potential for multi-mode networks, researchers have explored other approaches, including entanglement between light fields and atoms in quantum superpositions of the ground and highly-excited Rydberg electronic states. This latter approach allows the deterministic generation of photons that produces entanglement at a much higher rate.

However, until now, Rydberg atoms could not be excited to that state while confined to optical traps, so the traps had to be turned off for that step. That allowed the confined atoms to escape, preventing realization of atom-light entanglement.

Based on a suggestion from MURI colleagues at the University of Wisconsin, the Georgia Tech team developed a solution to that problem: a state-insensitive optical trap able to confine both ground-state and Rydberg atoms coherently. In this trap, atoms persist for as much as 80 milliseconds while being excited into the Rydberg state – and the researchers believe that can be extended with additional improvements. However, even the current atomic confinement time would be enough to operate complex protocols that might be part of a quantum network.

“The system we have realized is closer to being a node in a quantum network than what we have been able to do before,” said Kuzmich. “It is certainly a promising improvement.”

Key to the improved system is operation of an optical trap at wavelengths of 1,004 and 1,012 nanometers, so-called “magic” wavelengths tuned to both the Rydberg atoms and the ground state atoms, noted Lin Li, a graduate student in the Kuzmich Laboratory.

“We have experimentally demonstrated that in such a trap, the quantum coherence can be well preserved for a few microseconds and that we can confine atoms for as long as 80 milliseconds,” Li said. “There are ways that we can improve this, but with the help of this state-insensitive trap, we have achieved entanglement between light and the Rydberg excitation.”

The rate of generating entangled photons increased from a few photons per second with the earlier approaches to as many as 5,000 photons per second with the new technique, Kuzmich said. That will allow the researchers to pursue future research goals – such as demonstration of quantum gates – as they optimize their technique.

Experimentally, the research works as follows: (1) an ultra-cold gas of rubidium atoms was confined in a one-dimensional optical lattice using lasers operating at 1,004nm and 1,012nm wavelengths. The atomic ensemble was driven from the collective ground state into a single excited state; (2) By applying a laser field, an entangled state was generated. The retrieved field was mixed with the coherent field using polarizing beam-splitters, followed by measurement at single-photon detectors; (3) The remaining spin wave was mapped into a field by a laser field.

According to Kuzmich, the success demonstrates the value of collaboration through the MURI supported by the Air Force Office of Scientific Research, which in 2012 awarded $8.5 million to a consortium of seven U.S. universities that are working together to determine the best approach for creating quantum memories based on the interaction between light and matter.

Through the MURI, a team of universities is considering three different approaches for creating entangled quantum memories that could facilitate long-distance transmission of secure information. Among the collaborators in the five-year program are Mark Saffman and Thad Walker at the University of Wisconsin, Mikhail Lukin of Harvard, and Luming Duan of the University of Michigan, who at the beginning of this century made pioneering proposals which formed the basis of the approach that Kuzmich, Li and colleague Yaroslav Dudin used to create the entanglement between light and the Rydberg excitatio

Today, everywhere we turn we hear speakers give presentations at conferences and industry events despairing how the rise in silicon design costs is hampering the semiconductor industries growth path. As part of this problem, we now recognize that software design costs have eclipsed silicon design efforts and have become the largest portion of the SoC creation effort. In addition, IP integration costs are now rising as more discrete IP blocks are infused into SoC designs today.

The design landscape has also changed due to rising design complexity and lengthening design cycle times, especially in the System-on-a-Chip (SoC) market. These changes are having an impact on the SoC Design Start market delaying and preventing, to some degree, the architectural refreshes silicon designers undertake periodically to bring their solutions into line with changing market requirements and rising customer expectations.

Semico Research Corp. has looked at these issues and their impact on SoC silicon and software design cost efforts and encapsulated these changes and trends in a new report titled: SoC Silicon and Software Design Cost Analysis: Costs for Higher Complexity Continue to Rise, May, 2013.

Forecasts for SoC silicon design costs and software design costs are given from the 90nm node out through the 10nm node focusing on Advanced Performance Multicore SoCs, Value Multicore SoCs and Basic SoCs. Categories of effort with definitions for silicon and software designs are established with a forecast given at each node for the three types of SoC silicon. A forecast for Derivative SoC design costs at the 28nm node from 2011 through 2017 is also given.

Some of the data discussed in 58 pages with 15 tables and 33 graphs are:

• Total SoC design costs increased 48 percent from the 28nm node to the 20nm node and are expected to increase 31 percent again at the 14nm node and 35 percent at the 10nm node.
• Total SoC silicon design costs increased 78 percent at the 28nm node from the 40nm node.
• Total Software design costs increased 102 percent at the 28nm node and are forecast to show a CAGR of 79 percent through the 10nm node.
• Advanced Performance Multicore SoCs represent the most expensive silicon designs with Value Multicore SoCs and Basic SoCs exhibiting lower design costs.
• Derivative SoC silicon designs allow designers to accomplish their solutions at a fraction of the cost compared to first time efforts at the same process node when it first becomes commercially available.
• Costs for an Advanced Performance Multicore SoC design, continuously done at the 45nm node will experience a negative CAGR of 12.7 percent by the time the 14nm process geometry becomes commercially available, showing that subsequent designs at the same node become less expensive over time.
• 20nm silicon with a $20.00 ASP is required to ship 9.238M units to reach the breakeven point.
• The cost to integrate all the discrete IP blocks used in contemporary SoC designs is also rising for both the silicon and software efforts, showing a CAGR of 77.2 percent.
• Discussion of initiatives by EDA vendors to create tools allowing software designers to reduce design costs and more fully integrate their effort with silicon designers.

GLOBALFOUNDRIES and Fuzhou Rockchip Electronics Co., Ltd. today announced that Rockchip’s next-generation mobile processors are ramping to production on GLOBALFOUNDRIES’ 28nm High-K Metal Gate (HKMG) process technology. Based on a multi-core ARM Cortex-A9 design, the RK3188 and RK3168 chips are optimized for tomorrow’s high-performance, low-cost tablets that require long-lasting battery life (see product specifications in annex).

The combination of Rockchip’s design and GLOBALFOUNDRIES’ 28nm HKMG process technology resulted in a mainstream tablet System-on-Chip (SoC) capable of operating at up to 1.8 GHz performance, while still maintaining the power efficiency expected by mobile device users. The chips began sampling to OEMs in early 2013 and are now ramping to support a wide range of manufacturers.

“Collaborative foundry partnerships are critical for us to differentiate ourselves in the competitive market for mainstream mobile SoCs,” said Chen Feng, vice president of Rockchip. “We have chosen GLOBALFOUNDRIES as our strategic source partner of 28nm HKMG because their 28nm HKMG process has allowed us to ramp our products with very high yields in a relatively short timeframe. This partnership is a true demonstration of GLOBALFOUNDRIES’ unique approach to Collaborative Device Manufacturing.”

“At GLOBALFOUNDRIES, we are constantly seeking opportunities to offer our customers innovative silicon solutions to help them get the most benefits from their SoC designs,” said Mike Noonen, executive vice president of marketing, sales, design and quality at GLOBALFOUNDRIES. “Our partnership with Rockchip is a great example of how early collaboration can result in better performance and power characteristics with reduced time-to-market. We are excited to see Rockchip successfully leveraging this technology on our production-proven HKMG process.”

GLOBALFOUNDRIES’ 28nm-SLP technology is based on GLOBALFOUNDRIES’ “Gate First” approach to HKMG, which has been in volume production for more than two years.

The market for hard disk drives (HDDs) used in video surveillance will hit the billion-dollar level in less than five years, as safety concerns and the requirement for higher image quality spur demand for more data storage, according to a Storage Space Market Brief from information and analytics provider IHS.

Revenue for both internal and external HDDs in video-surveillance applications will rise from $638.7 million this year to $1.0 billion by 2017, a remarkable 57 percent increase. Growth this year alone is forecast to reach 23 percent from 2012 revenue of $521.1 million, and double-digit-percentage revenue expansion will ensue each year for the next four years.

The revenue figures translate to 7.3 million units in shipments by 2017, up from 2.4 million units in 2012 and a projected 3.5 million units this year.

“The HDD industry as a whole will reap the benefits of a fast-growing video surveillance industry that requires ample storage, with the need for higher-quality video, network connectivity and cloud storage also driving the market,” said Fang Zhang, analyst for storage systems at IHS. “At present, internal HDDs that combine storage capacity with the recording system in one unit have a larger market than external HDDs in which the drive is separate. Shipments of HDDs for internal storage were higher than those for HDDs in external storage during 2012—a feat that will be replicated this year.”

Products on the market today that use internal HDDs for video surveillance include internal direct attached storage (DAS), enterprise digital video recording (DVR), box appliance network video recorder (NVR) and PC-based network video recorders.

Next year, however, the tables get turned permanently as external storage HDD shipments take the lead. From a 48 percent share of the market in shipments last year of the total HDD space for video surveillance, external HDDs leap over internal HDDs in 2014 with a slight majority of 51 percent share of the market. And while internal HDDs continue to retain a viable portion of the market, external HDDs will keep gaining in the market with their share hitting approximately 54 percent by 2017.

Internal vs. external storage

While internal storage is a cost-effective way of storing video data, external storage boasts larger capacities that can be added flexibly to a system as the need arises.

External storage also has more versatile applications as it can be connected or viewed anywhere, such as in the cloud, where the potential for external storage is plentiful and abundant. The capability of external systems to scale and accommodate large amounts of video data is one reason why more hard disk drives will be needed in the coming years, pushing both shipments and revenues up. Another reason is that high-quality video itself will demand more storage space, which also will help boost HDD use in video surveillance.

All told, total HDD shipments for video surveillance will grow at a compound annual growth rate of 29 percent from 2012 to 2017, with revenue likewise increasing at a lower but still-solid rate of 14 percent during the same period.