I am considering a move to Milan, so testing waters for a good job there as an r&d mechanical engineer or a data scientist. Market is flooded by job agencies and body rental shops, therefore I need direct contacts with firms through my own 2016 database, which is made of 5000+ listings in Central and Northern Italy. Neither an optmist nor a pessimist here, just going to have a good look in June and see what happens. First step there may be renting a flat with a good wi-fi service in the recently rejuvenated quartiere Isola for one month.
Deep learning benchmark from Fast.ai course Lesson 1 with a GTX 1070 8GB from my Win 10 laptop with factory settings: 292s stable (260-300s with 1070, 380-430s with a 1060 6GB, 120-150s with a 1080 Ti 11GB). Laptop ready, wearing the helmet, off I go: Fast.ai course, Kaggle.com competitions, own studies.
I am finding difficult to perform a clean install of Python packages (Theano, Tensorflow, Keras on top of Anaconda and xgboost) for deep learning on my Win 10 native laptop with a GTX 1070 8GB. Procedures are cumbersome and the most recent drivers do not work, adding to compiling errors. I am now looking at this tutorial by Phil Ferriere updated to May 2017, which seems reasonable and clean. Alternatively, an Ubuntu 16.04 partition on my second data hard disk should allow a smoother install. I’ll go that route, Windows is too fragile to risk flooding my SSD again. EDIT: the tutorial by Phil worked and I now have all the packages I need up and running on Win 10. Woohoo!
CERN started its third year with LHC experiments on May 23. Nuclear physics is pretty different from nuclear engineering but more important in the grand scheme of human knowledge. From their press release linked above: “Physicists hope to be able to identify disparities between their measurements and the Standard Model. This is one of the ways in which the unknown can be probed. Although it describes a lot of the phenomena of the infinitely small precisely, the Standard Model leaves many questions unanswered. For example, it describes only 5% of the universe; the rest is formed of dark matter and dark energy, the nature of which are as yet unknown.” Good luck!
The DronesBench Index is a novel parameter for the efficiency of drones developed by DPM Elettronica, Foggia, Italy, through the DronesBench testing machine. Next DronesBench public events: 1) The 2017 IEEE International Instrumentation and Measurement Technology Conference, Torino, Italy, 22-25 May; 2) The Mini Maker Faire of Torino, Italy, 27-28 May; 3) The Commercial UAV Expo Europe, Brussels, Belgium, 20-22 June; 4) The 4th IEEE International Workshop on Metrology for Aerospace, Padua, Italy, 21-23 June. The most promising market for DronesBench at the moment is the educational sector: high schools, universities, labs in Italy and abroad.
The joke about nuclear fusion is it is always 20 years away. The best project, the biggest and the most funded today is from the global consortium ITER in Cadarache, France, which is hosting an open day on May 20. ITER is expected to achieve first plasma in 2025 and full operation in 2035. On the other end of the spectrum, startups are trying to get fusion at small scale, for example Tri Alpha Energy in the US, Tokamak Energy in the UK (first plasma achieved already!), General Fusion in Canada. Who will get commercially there first, the mammoth or the mouses? The answer, as usual, is still 20 years away.
There is an explosion of applied research and science coming from high performance computing. This is leading to a great amount of papers being submitted to journals and conferences. Peer reviewing novel results is difficult and time consuming, so the industry leaders are not available for this unpaid burden. Journals and conferences are therefore trying to exploit this tsunami by charging a reading or presentation fee, in that circumventing peer review. As a result, top institutions are keeping outsiders away, in that creating two scientific worlds: one rich for the elites, another poor for the rest of the world.
My Twitter account NukePep is pretty busy following the most established players from the nuclear industry, media and research labs since 2013. A few critical voices are more involved with weapons and geopolitics, Bulletin of the Atomic Scientists among them. They are a small think tank from the US lobbying for climate attention, security awareness and sustainable technological development. Their latest annual report 2016 has been released three days ago and is an interesting read, a necessary link between hard science and media narratives. I am not sure I agree with them in full but their concern is sound.
Spent a lot of time last year on Kaggle, learning the tricks of competitive data science. Kagglers were in love with this three-layer architecture: input) data cleaning and feature engineering; 1) a number of xgboost and deep learning models; 2) a number of blending and optimisation functions; 3) weighted averages; output) final submission. Making use of Python open source packages, AWS cloud images and EC2 instances, you were done. Planning to join again for deep learning competitions.
It is almost time to get rid of AWS and its p2.xlarge spot instances, so a new laptop will be needed. The best compromise today has an Nvidia GTX 1070 GPU, which couples well with an Intel i7-6700 CPU. I also want a 15.6 screen, low temperatures under extreme workload and upgradability for RAM and HDs. From this Dec 2016 comparative review by NotebookCheck, it seems to me that the Acer Predator 15 G9-593 is as good, expensive and bulky as a same class desktop. At slightly below €2000, it is a benchmark for the future. EDIT: for CPU thermal throttling, new thermal paste would solve or even undervolting core by -0.120V.