Two recent announcements highlight a growing trend towards partnership and innovation aimed at space-based technical computing and storage infrastructure. The former seeks a 100X increase in computational power via a High-Performance Spaceflight Computing (HPSC) processor, and the latter is exploring appropriate storage media for low-earth orbit satellite focal planes and RF sensor data.
MLCommons Adds Edge/Embedded AI Inference Benchmark
Alex Norton and Bob Sorensen
MLCommons, an international artificial intelligence (AI) standards body formed in 2018, launched MLPerf Tiny, their first benchmark targeted at the inference capabilities of edge and embedded devices, or what they call "intelligence in everyday devices". The new benchmark is now part of the overall MLPerf benchmark suite, which measures AI training and inference performance on a wide variety of workloads, including natural language processing and image recognition. The benchmark covers four machine learning (ML) tasks focused on camera and microphone sensors as inputs: keyword spotting, visual wake words, tiny image classification, and anomaly detection. Some important use cases include smart home security, virtual assistants, and predictive maintenance.
6 202021 | HYP_Link
Opportunity for DNA as a New Archive Storage Medium
Mark Nossokoff and Bob Sorensen
Using biological building blocks in place of traditional materials to assemble computers has been a research topic for many years, but recently the first potential commercial use cases have begun to emerge, centered on storage for large data sets. The DNA Storage Alliance, created to promote a storage ecosystem based on synthesized DNA strands, recently shared their aspirations for the emerging technology that offers significant promise in durability, simplicity, cost, and density over traditional magnetic counterparts. The initial goals of the alliance are to educate the public and raise awareness about DNA-based storage. Further out, the alliance may pursue the creation of specifications and standards, such as encoding, physical interfaces, retention, and file systems, to ensure that DNA-based solutions complement existing storage hierarchies. The alliance notes that expectations for the growth rate of current storage mechanisms cannot keep pace with the rising demand for data storage, particularly where growing data retention and related data mining efforts are driving the need to save increasingly larger data sets for longer periods of time. Such requirements are well suited to DNA-based archive storage characteristics in applications including digital content creation, robotics, smart cities, autonomous vehicles, healthcare, astronomy, and climate science.