Core-to-core communication; energy harvesting with fish scales and body heat.
Most research featured in the Power/Performance Bits has far-off applications, but a team from North Carolina State University and Intel developed something that could be brought into practice today: a way to accelerate core-to-core communication.
Many important workloads incur significant core-to-core communication and are affected significantly by the costs, including pipelined packet processing which is widely used in software-based networking solutions. In these workloads, threads run on different cores and pass packets from one core to another for different stages of processing using software queues.
The team analyzed the behavior and overheads of software queue management, and proposed a method to optimize such communication.
“This approach, called the core-to-core communication acceleration framework (CAF), improves communication performance by two to 12 times,” said Yan Solihin, a professor of electrical and computer engineering at NC State. “In other words, the execution times – from start to finish – are twice as fast or faster.”
The key to the CAF design is the queue management device (QMD), a small device attached to the processor network on a chip. The QMD is capable of simple computational functions and effectively keeps track of communication requests between cores without having to rely on software routines.
The researchers also found that, because it can perform basic computation, the QMD can be used to aggregate data from multiple cores, expediting some basic computational functions by as much as 15%.
“We are now looking at developing other on-chip devices that could accelerate more multi-core computations,” Solihin said.
The paper will be presented at the Conference on Parallel Architectures and Compilation Techniques, being held Sept. 11 to 15 in Haifa, Israel.
Piezoelectric fish scales
Researchers at Jadavpur University in Koltata, India repurposed fish byproducts, harnessing the piezoelectric properties of the collagen fibers in the scales to fabricate an energy harvesting bio-piezoelectric nanogenerator.
To do this, the researchers first “collected biowaste in the form of hard, raw fish scales from a fish processing market, and then used a demineralization process to make them transparent and flexible,” explained Dipankar Mandal, assistant professor in the department of physics at Jadavpur. The collagens within the processed fish scales served as an active piezoelectric element.
“We were able to make a bio-piezoelectric nanogenerator — a.k.a. energy harvester — with electrodes on both sides, and then laminated it,” Mandal said. “We discovered that the piezoelectricity of the fish scale collagen is quite large (~5 pC/N), which we were able to confirm via direct measurement.”
Experimental and theoretical tests helped them clarify the energy scavenging performance of the bio-piezoelectric nanogenerator. It’s capable of scavenging several types of ambient mechanical energies — including body movements, machine and sound vibrations, and wind flow. Repeatedly touching the bio-piezoelectric nanogenerator with a finger could turn on more than 50 blue LEDs.
The group sees applications for their work in transparent electronics, biocompatible and biodegradable electronics, edible electronics, self-powered implantable medical devices, surgeries, e-healthcare monitoring, as well as in vitro and in vivo diagnostics, apart from its myriad uses for portable electronics.
“In the future, our goal is to implant a bio-piezoelectric nanogenerator into a heart for pacemaker devices, where it will continuously generate power from heartbeats for the device’s operation,” Mandal said. “Then it will degrade when no longer needed. Since heart tissue is also composed of collagen, our bio-piezoelectric nanogenerator is expected to be very compatible with the heart.”
Body heat as a power source
A team at the Huazhong University of Science and Technology in Wuhan, China developed a flexible, wearable thermocell that can harvest energy from body heat.
Muscle activity and metabolism cause our bodies to produce constant heat, some of which is released through the skin into the environment. Because of the relatively small temperature difference between skin (approximately 32°C, 89.6°F) and the temperature of the surroundings, it is not so easy to make use of body heat. Previous thermoelectric generators, such as those based on semiconductors, produce too little energy, are costly, or are too brittle for use in wearable systems. Thermocells with electrolyte solutions are difficult to integrate into extensive wearable systems. So the team adapted the idea, using two different gel-based electrolytes.
The researchers made use of the thermogalvanic effect: if two electrodes in contact with an electrolyte solution—or an electrolyte gel—are kept at different temperatures, a potential difference is generated. The ions of a redox pair in the electrolyte can rapidly switch between two different charge states, accepting or releasing electrons at electrodes with different temperature. In order to use this to produce a current, the scientists combined two types of cells containing two different redox pairs. Each cell consists of two tiny metal plates that act as electrodes, with an electrolyte gel in between. The first cell type contains the Fe2+/Fe3+ redox pair. The second type of cell contains the complex ions [Fe(CN)6]3-/[Fe(CN)6]4-. Because of the choice of these redox pairs, in cell type 1, the cold end gives a negative potential, while in type 2, the cold end gives a positive potential.
The researchers arranged these two types of cells into a checkerboard pattern. The cells were connected to each other by metal plates alternating above and below, to link them into a series, which was integrated into a glove. When the glove is worn, the desired temperature difference results between the upper and lower plates. This produces a voltage between neighboring cells, making it possible to generate current to power a device or charge a battery.
In an environment at 5°C (41°F), it was possible to produce 0.7 volts and about 0.3 μW. By optimizing this system, the researchers believe it should be possible to improve the power, even with smaller temperature gradients.