Posted on February 28, 2024 by elankovan sundararajan
New Atlas; Paul McClure (February 22, 2024)
University of California, Riverside researchers developed a framework to increase computer processing speed while decreasing energy consumption. The new framework, called simultaneous and heterogeneous multithreading (SHMT), is a type of parallel processing in which the computational function of multiple components is broken up and shared. SHMT’s runtime system separates virtual operations into one or more high-level operations (HLOPs) to take advantage of multiple hardware resources simultaneously, allocating HLOPs to the target hardware’s task queues and adjusting task assignments as necessary. In tests, SHMT was found to be 1.95 times faster and use 51% less energy than the baseline method.
—MORE—
Filed under: Computer Science, Source: ACM TechNews | Leave a comment »
Posted on February 28, 2024 by elankovan sundararajan
Holly Chik (February 22, 2024) South China Morning Post;
A team of researchers from China’s University of Shanghai for Science and Technology, Peking University, and the Chinese Academy of Sciences developed a technology that allows about 5.8 billion indexed Web pages to be stored in a device the size of a desktop computer. The researchers created 3D architecture to store data across hundreds of layers in a disk, resulting in optical data storage capacity reaching the petabyte level.
—MORE—
Filed under: Computer Science, Source: ACM TechNews | Leave a comment »
Posted on February 15, 2024 by elankovan sundararajan
GeekWire; Taylor Soper (January 23, 2024)
An analysis of 153 million lines of code changed by GitClear, a developer analytics tool built in Seattle, found that “code churn,” or the percentage of lines thrown out less than two weeks after being authored, is on the rise. It also found that the percentage of “copy/pasted code” is increasing faster than “updated,” “deleted,” or “moved” code. Said GitClear’s Bill Harding, “In this regard, the composition of AI-generated code is similar to a short-term developer that doesn’t thoughtfully integrate their work into the broader project.”
—MORE—
Filed under: Computer Science, Source: ACM TechNews | Leave a comment »
Posted on December 26, 2023 by elankovan sundararajan
University of Stirling (U.K.)
December 11, 2023
Researchers at the U.K.’s University of Stirling leveraged ChatGPT to improve the speed and reliability of a software program. The researchers asked ChatGPT to update software automatically to improve computer coding. Stirling’s Sandy Brownlee said, “We found that, on the open source project we used as a case study, a LLM [large language model] was able to produce faster versions of the program around 15% of the time, which is half as good again as the previous approach.”
—MORE—
Filed under: Artificial Intelligence | Leave a comment »
Posted on December 26, 2023 by elankovan sundararajan
HPCwire
December 14, 2023
IBM is partnering with Japan’s Keio University and The University of Tokyo, South Korea’s Yonsei University and Seoul National University, and The University of Chicago to support quantum education activities. IBM intends to deliver educational offerings, in combination with contributions from each of the participating universities, to advance the training of up to 40,000 students over the next decade to prepare them for the quantum workforce.
—MORE—
Filed under: Computer Science, Quantum Computing | Leave a comment »
Posted on December 26, 2023 by elankovan sundararajan
NVIDIA has unveiled a more advanced graphics processing unit (GPU), the H200, which is nearly twice as fast as its H100 standalone accelerator. Based on the Hopper architecture, the H200 is a powerful upgrade aimed at accelerating high-performance computing (HPC) and models powering the generative artificial intelligence boom.
—MORE—
Filed under: Artificial Intelligence, Computer Science, High Performance Computing, High Throughput COmputing, Machine Learning, Source: Interesting Engineering | Tagged: Source: Interesting Engineering | Leave a comment »
Posted on December 26, 2023 by elankovan sundararajan
Interesting Engineering
Sejal Sharma
December 13, 2023
Researchers at Australia’s International Centre for Neuromorphic Systems (ICNS) at Western Sydney University have developed a neuromorphic supercomputer than can perform 228 trillion synaptic operations per second, similar to the number the human brain can handle. The DeepSouth supercomputer, expected to be operational by April, is notable for being smaller than conventional supercomputers and able to process substantial amounts of data at a rapid pace using much less power.
—MORE—
Filed under: Artificial Intelligence, Computer Science, High Performance Computing, High Throughput COmputing, Source: ACM TechNews | Leave a comment »
Posted on December 26, 2023 by elankovan sundararajan
Reuters
David Lague
December 14, 2023
When Q-day, the day quantum computers are able to defeat current encryption methods, will occur is up for debate, given that quantum computing is still in its early days. In the meantime, nations including the U.S. and China reportedly are harvesting vast amounts of encrypted data in hopes of decrypting it later. Meanwhile, the U.S. and its allies are working on post-quantum cryptography, and China is working on a theoretically hack-proof quantum communications network. The World Economic Forum predicts that 20 billion devices will need to be upgraded or replaced over the next 20 years to meet quantum security standards.
—MORE—
Filed under: Computer Science, Quantum Computing, Source: ACM TechNews | Leave a comment »
Posted on July 29, 2023 by elankovan sundararajan
By David Linthicum,
InfoWorld | Jul 28, 2023
AI-driven coding is now in wide use, but we may not know all the risks of using it until the damage has been done. Think security problems and code that wastes resources.
—MORE—
Filed under: Cloud Computing, Infoworld | Leave a comment »
Posted on July 29, 2023 by elankovan sundararajan
InfoWorld
Serdar Yegulalp
July 17, 2023
After three years in the works, version 3.0 of Cython, the Python library for compiling Python code to C, has been released. The latest version of Cython aims to simplify the process of writing C extensions for Python. It no longer supports Python 2, but supports newer Python features up to Python 3.12 while expanding “pure Python mode,” which lets developers use existing Python linting and code analysis tools on Cython. Version 3.0 also improves NumPy support to allow developers to write NumPy ufuncs directly in Cython. Meanwhile, a guaranteed stable subset of Python’s APIs is accessible via the new “limited API” for Python, which will eliminate the need to recompile Cython extension modules for use in future versions of Python.
—MORE—
Filed under: Computer Science, Programming language, Python, Source: ACM TechNews | Leave a comment »