Search

Arising from: C. Trisos et al. Nature 580, 496–501, https://doi.org/10.1038/s41586-020-2189-9 (2020). ABSTRACT The fundamental assumption behind this provocative study (Trisos et al. 2020) is that current and future climatic tolerance limits for a species can be inferred from the last 170 years of climate records within its current distribution. Using this approach, the authors project the probable fate, in time, of sets of supposedly co-occurring species “assemblages,” under modeled future climates. Without any doubt, the fate of species under ongoing climate change is urgent and alarming, but I suggest that both the synchrony of exposure to intolerable climate regimes and the “ecological disruptions” arising from such exposure may be substantially overestimated in by this study.

We have been following the dictum of Moore's Law for longer than most engineers have been alive. We're used to it, we know how to do it, and its focus on functionality, performance, and economics has yielded remarkable systems, from today's smartphones to supercomputers and internet servers, to engine controllers, anti-lock brakes, airbags, and the navigation systems in our vehicles. But software is notoriously buggy, and hardware isn't defect-free, either. Worse, we seem to only take security into account when some hacker perpetrates yet another outrage. The pending end of Moore's Law will greatly diminish the historical demand to replace systems every few years, which will put new pressure on system aging effects. Our military is using these same commercial systems, because they are the highest performance, most economical solutions, but this means that they, too, are now exposed to the same problems. By one estimate, more than half of automotive recalls are due to software defects. And here come self-driving cars, which could potentially suffer all of these threat axes simultaneously. Performance and efficiency enable new applications, and in the past, the computer design community could simply stop when those targets were within reach. In this talk, I will argue that we cannot get away with that attitude any longer, and need to turn our attention to areas of system design that are not good enough for how those systems will be used in the near future.

Exascale computing systems are essential for the scientific fields that will transform the 21st century global economy, including energy, biotechnology, nanotechnology, and materials science. Progress in these fields is predicated on the ability to perform advanced scientific and engineering simulations, and analyze the deluge of data. On July 29, 2013, ASCAC was charged by Patricia Dehmer, the Acting Director of the Office of Science, to assemble a subcommittee to provide advice on exascale computing. This subcommittee was directed to return a list of no more than ten technical approaches (hardware and software) that will enable the development of a system that achieves the Department's goals for exascale computing. Numerous reports over the past few years have documented the technical challenges and the non¬-viability of simply scaling existing computer designs to reach exascale. The technical challenges revolve around energy consumption, memory performance, resilience, extreme concurrency, and big data. Drawing from these reports and more recent experience, this ASCAC subcommittee has identified the top ten computing technology advancements that are critical to making a capable, economically viable, exascale system.