Readable Ideas that may change the speed of boinc and Research - add yours - firstly the Berkeley labs 100 award for science list

https://science.energy.gov/about/honors-and-awards/rd-100-awards/2017-RD-100-Award-Finalists

Office of Science Laboratory: Lawrence Berkeley National Laboratory

Other Partners: N/A

Name of Project: Double Barcoded Shotgun Expression Library Sequencing (Dub-seq)

Double Barcoded Shotgun Expression Library Sequencing (Dub-seq) is a technology for discovering the functions of genes in microbes under different environmental conditions. Because Dub-seq can process large amounts of genetic information at once, it is faster, cheaper, more flexible, and requires less work than previous genetic analysis technologies. Scientists can adapt it to a variety of biotechnologies, such as discovering new enzymes, finding new cancer drugs, gaining insight into resistance to viruses, and understanding how antibiotics act on microbes that cause disease.

****

Office of Science Laboratory: Lawrence Berkeley National Laboratory

Other Partners: University of Illinois-Champaign

Name of Project: CrunchFlow

CrunchFlow is a software package that simulates how chemical reactions occur and change as fluids travel underground. CrunchFlow includes a number of chemical and physical processes that similar products do not, such as changes in how easily water can move through rocks. All of these features are available in a single package that users with a variety of expertise can run on a desktop computer. With CrunchFlow's computational efficiency, scientists can achieve high spatial resolution while extending simulations far back in geologic time. By improving the accuracy of a range of Earth and environmental sciences applications, CrunchFlow helps scientists better understand current and past ecological systems below the Earth's surface.

****

Office of Science Laboratory: Oak Ridge National Laboratory

Other Partners: SepQuant, Inc.

Name of Project: dropletProbe Surface Sampling System for Mass Spectrometry

The dropletProbe system, developed with support from Oak Ridge National Laboratory, is a completely new means of surface sampling for mass spectrometry, a major scientific technique for measuring the masses of chemicals in a sample. The dropletProbe system provides rapid, simple chemical extraction and analysis for a host of scientific applications. It is a low cost, low-maintenance, and nondestructive method for sampling complex analytical surfaces, such as biological tissue samples. It provides scientists with a high degree of precision for targeting specific areas on the sample. By reducing cost and improving accuracy, this tool should help increase the pace of scientific discovery.

****

Office of Science Laboratories: Pacific Northwest National Laboratory; Lawrence Berkeley National Laboratory

Other Partners: National Energy Technology Laboratory; Los Alamos National Laboratory; Lawrence Livermore National Laboratory

Name of Project: National Risk Assessment Partnership (NRAP) Toolset

Deep underground geologic formations offer promising places to safely and effectively store large volumes of carbon dioxide (CO2) generated from burning coal, oil, and natural gas. The National Risk Assessment Partnership (NRAP) Toolset is the first complete suite of computer software that models possible environmental risks from potential storage sites, such as fluid leakage and earthquakes. The Tool-set draws on the expertise of five DOE national laboratories and is being used by more than 250 stakeholders from academia, regulatory agencies, and industry.

You caught my eye, but it's unclear to me what you are really asking for here. The locus of confusion appears to be "readable", but the examples aren't helping. I also think there is some motivational problem here. As a volunteer, my priorities are for convenience and lack of interference with other things my computers are doing. As a researcher in some application field, your priority might be for speed, but why would we volunteers care how quickly you can generate publishable results?

Putting on my dusty computer scientist hat, I would revert to my concerns with computational efficiency and look to OS principles, where the primary focus is usually on identifying the scarce or bottleneck resources and arranging or rescheduling things to optimize overall throughput. Seems like your bottleneck resource is network bandwidth, but you have a huge surplus of CPUs, so BOINC should seek to reduce the data transfers even if it results in some waste of computation. I could do a deeper analysis, but right now I'm too engulfed in flying dust to think harder about it.

---

#1 Freedom = (Meaningful - Constrained) Choice{5} != (Beer^3 | Speech)

thank you for the meaningful reply, much appreciated RS

of interest = genetic & cancer

topic = work

for = developers + propagation

the Can-DLE project is deeply meaningful to cancer research .... and genetics : please learn more about their theorems and programming development ...

https://cbiit.cancer.gov/ncip/hpc/candle


the following is a long winded article mostly going on about how much work they did watching the screen as the HPC Exo-scale did the work debugging code ;P but interesting non the less

https://www.hpcwire.com/2017/11/07/sc17-ai-machine-learning-central-computational-attack-cancer/

RS

TCell function controller's explored

https://www.scientificamerican.com/custom-media/exploring-the-tumor-microenvironment/

have you explored the MS link to the immune-system inhibitor ?

(c)RS