Reflections Of The Void: 05/01/2017

Tuesday, May 30, 2017

TFS: Some Interesting concepts behind TFS - Next-generation file system. The Design goals are ambitious & well thought. However, Machine learning for caching can make sense but I will still prefer to see how it cope in real life over a long period of time. Especially when there is historical access pattern that span days - months.
Stanford DAWN Project : While I know some cool tech will come out of it public cloud provider will probably shape the future of machine learning infrastructure at faster pace & with shorter iteration
Bolt : this is really cool, the authors of the paper are pretty much pushing the noisy neighbour to 11. They propose a technique that offers stealth denial of service on public cloud infrastructure.

Intel Manufacturing Conf : sets of slide deck giving a peek into Intel manufacturing process and the upcoming wave of 10 nm chips. It seems Intel is currently keeping up with Moore's law, not by reducing the transistor size but by also increasing transistor density.
Introduction to Topological Quantum Computation : introduce the concept of quantum computing with anyons which allow a more resilient quantum computing system.
Serverless Conf video : almost all video are now available. Check out the Serverless at Nordstrom video. This is, in my view, the best of the bunch. It's an actual practical talk by the dev who implemented it and without marketing spin.

Exascale Computing Project : An update on the state on the US exascale project. It always amazes me how HPC system and software are such a different beast from industrial solution. I wonder if it isn't artificially maintained as such more because of bias and ego than the actual requirements. [slides]
Pravega : Where stream is the new storage. It's pretty much taking the log based storage to an extreme and combine messaging and persistence in an elegant approach. It's really aimed at tackling modern distributed systems and micro service problem. A concept to watch[github]
ServerlessConf : If you couldn't attend here is the least of the key takeway / summaries you can find online:

In-Datacenter Performance Analysis of a Tensor Processing Unit : How custom deep learning hardware behave in real datacenter, implication and gain associated with the use of such custom solution.
Push on Green : Great article on google roll out policy and process. A lot of common sense, and also some less common but as important. This is a great read for anybody involved in software delivery and especially if you are aiming at having an efficient CI/DI system.
BBR : google congestion protocol for maximising bandwidth usage. It's new TCP scheduling algorithm to fight buffer-bloat at the TCP level. Since the majority of internet traffic is TCP, wide adoption would cause a big improvement. TCP scheduling only affects outgoing packets

Serverless Design Patterns : Basic serverless pattern, nothing fancy but useful to keep in mind.
How to Read a Paper : the short version is quite simple: Abstract -> conclusion ->figure, then start iterating through the paper, one layer at a time.
Billion node graph inference: iterative processing on The Machine : I still believe that the machine is vaporware. It was promised for years. Still at the stage of nice plastic prototype and everybody developing/designing for it use simulator/emulator. But hey, it's nice to see what you can theoretically achieve on "the machine" even if it's a Superdome X in reality.

Machine Learning Cheat Sheet : Ok it's more than one pager, but it contains pretty much all the essential machine learning info.
A Large Self-Annotated Corpus for Sarcasm : If you want to make your chatbot even more annoying .. This is what you need.
Resilient Functional Service Design : Hardening your distributed architecture the right way