Hey, it's HighScalability time:

Being Google CEO: Nice. Becoming Tony Stark: Priceless (Alphabet)

• $7: WeChat's revenue per user and there are 549 million of them; 60%: Etsy users using mobile; 10: times per second a self-driving car makes a decision; 900: calories in a litre of blood, vampires have very efficient metabolisms; 5 billion: the largest feature in the universe in light years


• Quotable Quotes:
  • @sbeam: they finally had the Enigma machine. They opened the case. A card fell out. Turing picked it up. "Damn. They included a EULA." #oraclefanfic
  • kordless: compute and storage continue to track with Moore's Law but bandwidth doesn't. I keep wondering if this isn't some sort of universal limitation on this reality that will force high decentralization.
  • @SciencePorn: If you were to remove all of the empty space from the atoms that make up every human on earth, all humans would fit into an apple.
  • @adrianco: Commodity server with 1.4TB of RAM running a mix of 16GB regular DRAM and 128GB Memory1 modules.
  • @JudithNursalim: "One of the most scalable structure in history was the Roman army. Its unit: eight guys; the number of guys that fits in a tent" - Chris Fry
  • GauntletWizard: Google RPCs are fast. The RPC trace depth of many requests is >20 in miliseconds. Google RPCs are free - Nobody budgets for intradatacenter traffic. Google RPCs are reliable - Most teams hold themselves to a SLA of 4 9s, as measured by their customers, and many see >5 as the usual state of affairs.
  • @rzidane360: I am a Java library and I will start 50 threads and allocate a billion objects  on your behalf.
  • @codinghorror: From Sandy Bridge in Jan 2011 to Skylake in Aug 2015, x86 CPU perf increased ~25%. Same time for ARM mobile CPUs: ~800%.
  • @raistolo: "The cloud is not a cloud at all, it's a limited number of companies that have control over a large part of the Internet" @granick
  • Benedict Evans: since 1999 there are now roughly 10x more people online, US online revenues from ecommerce and advertising have risen 15x, and the cost of creating software companies has fallen by roughly 10x. 


• App constellations aren't working. Is this another idea the West will borrow from the East? When One App Rules Them All: The Case of WeChat and Mobile in China: Chinese apps tend to combine as many features as possible into one application. This is in stark contrast to Western apps, which lean towards “app constellations”.


• It doesn't get much more direct than this. Labellio: Scalable Cloud Architecture for Efficient Multi-GPU Deep Learning: The Labellio architecture is based on the modern distributed computing architectural concept of microservices, with some modification to achieve maximal utilization of GPU resources. At the core of Labellio is a messaging bus for deep learning training and classification tasks, which launches GPU cloud instances on demand. Running behind the web interface and API layer are a number of components including data collectors, dataset builders, model trainer controllers, metadata databases, image preprocessors, online classifiers and GPU­-based trainers and predictors. These components all run inside docker containers. Each component communicates with the others mainly through the messaging bus to maximize the computing resources of CPU, GPU and network, and share data using object storage as well as RDBMS.


• How do might your application architecture change using Lambda? Here's a nice example of Building Scalable and Responsive Big Data Interfaces with AWS Lambda. A traditional master-slave or job server model is not used, instead Lambda is used to connect streams or processes in a pipeline. Work is broken down into smaller, parallel operations on small chunks with Lambda functions doing the heavy lifting. The pipeling consists of a S3 key lister, AWS Lambda invoker/result aggregator, Web client response handle. 


• The Indie Web folks have put together a really big list of Site Deaths, that is sites that have had their plugs pulled, bits blackened, dreams dashed. Take some time, look through, and say a little something for those that have gone before.

Don't miss all that the Internet has to say on Scalability, click below and become eventually consistent with all scalability knowledge (which means this post has many more items to read so please keep on reading)...