The Week in Review
Here is a collection of highlights from this week’s news stream as reported by HPCwire.
Look out silicon chips, here comes … DNA?
How cool is this? Computer chips made out of synthesized strands of DNA. Chris Dwyer, assistant professor of electrical and computer engineering at Duke’s Pratt School of Engineering, has been working on just such a technology.
In recent experiments, Dwyer demonstrated that by mixing customized snippets of DNA and other molecules, he could create billions of identical, tiny structures that look like miniature waffles. These nanostructures can be used as the building blocks for applications ranging from the biomedical to the computational. In fact, the self-assembly of molecular nanostructures is expected to one day form the basis of future high-speed integrated circuits.
The nanostructures function as optical logic circuits, which use light instead of electricity to signal a “yes or no”, “zero or one” type of binary response. This light-based DNA switching occurs more rapidly than with a traditional circuit that uses an electrical current.
Dwyer believes that these self-assembled, light-sensing molecular circuits can be produced inexpensively in virtually limitless quantities. In fact, in one day, one scientist in a lab can produce more of these simple logic circuits than the world’s entire output of silicon chips in a month. More complex circuits can be created using various combinations of DNA strands.
This is the first demonstration of such an active and rapid processing and sensing capacity at the molecular level. Conventional technology has reached its physical limits. The ability to cheaply produce virtually unlimited supplies of these tiny circuits seems to me to be the next logical step.
In HPC compiler news…
There are a couple of compiler announcements this week. First up, Impulse Accelerated Technologies and DRC Computer have announced that the Impulse C-to-FPGA tools have been integrated with the DRC Accelium, Xilinx-based coprocessor card. This will allow software developers to access FPGA-based hardware acceleration using familiar C programming tools. The integration work was performed by Synective Labs, a long-time partner of both DRC and Impulse.
According to the announcement, many design teams are finding that their investment in software development is outpacing their investment in hardware development, so there is an increasing need for programming tools that enable the quick deployment of future platforms as they emerge. The enhanced tools are said to support such a migration to future FPGA-based accelerators since most if not all of the design code is abstracted from hardware, allowing the design and test code to be easily retargeted to future platforms.
And from the Portland Group: PGI compilers are now available across the entire Cray product line. PGI compilers are already installed on nearly all Cray XT systems, but now Cray will resell PGI compilers and development tools with the Cray CX1 and CX1000 systems. The optimizing Fortran, C, and C++ compilers and development tools support the latest 64-bit processors from AMD and Intel and include the new PGI Accelerator features for NVIDIA GPUs, allowing them to leverage the GPU capabilities of the Cray CX line.
Douglas Miles, director of the Portland Group, said:
Giving Cray’s customers the flexibility and simplicity of purchasing our high-performance compilers directly from Cray for the Cray CX1 and Cray CX1000 systems is a big win for everyone. Together we can offer a uniform PGI compiler environment on the Cray CX machines running Linux or Windows up through the high-end scalable Cray XT machines and provide significant ease of migration and application upscaling benefits for existing and future Cray users.
PGI compilers and tools are used by developers to create high-performance computing applications in such fields as weather forecasting, geophysical processing, aerodynamic simulation, structural analysis, automotive crash-testing, and computational chemistry.