Ready to make your mark as a networking innovator and leader? Interested in learning how today’s open source and open standards initiatives are coming together? Want to know how networking is revolutionizing hot markets such as micro-services and Internet of Things (IoT)? Feeling the need to reach out to executives, technologists, vendors, and investors? If... Read more →
A shared appetite for high performance computing hardware and frameworks is pushing both supercomputing and deep learning into the same territory. This has been happening in earnest over the last year, and while most efforts have been confined to software and applications research, some supercomputer centers are spinning out new machines dedicated exclusively to deep learning.
When it comes to such supercomputing sites on the bleeding edge, Japan’s RIKEN Advanced Institute for Computational Science is at the top of the list. The center’s Fujitsu-built K Computer is the seventh fastest machine on the planet according to the Top 500 rankings …
Japan to Unveil Pascal GPU-Based AI Supercomputer was written by Nicole Hemsoth at The Next Platform.
The post Worth Reading: Five jobs robots will take first appeared first on 'net work.
Spiceworks survey shows positive workplace relationships rank above money as key to job satisfaction.
Sigfox often deploys antennas on billboards to speed deployment.
Its secret sauce has been its relationships with operators.
Buffering packets in a network is both good and bad. It is good because a buffer can hold packets from one stream while another stream’s packets are being processed, to take up and release short bursts of traffic, to hold and then release packets when there is a very short interruption on the wire (or during a route change), and in many other situations. However, queues are bad when there is a standing queue, which means a particular flow always has some number of packets in a particular queue along the path between the source and the destination. This normally occurs at the narrowest point along the path, or rather the link with the lowest bandwidth. In a previous post, I looked at BBR, a change to the way TCP computes its window sizes, that attempts to reduce the amount of traffic “in flight” between a sender and receiver to reduce the number of packets being held in a particular buffer along the way.
This post will consider another solution: CoDel. CoDel is essentially an improved tail drop mechanism that provides the correct signals to TCP to slow down its send rate, or rather to reduce the window size (and Continue reading