Archive

Archive for October, 2011

Coupled systems lecture 1…

October 28, 2011 Leave a comment

Coupled systems lecture 1 – Model Predictive Control http://t.co/CYYFZaOt via @creativecommons

Categories: News

http://t.co/NPM3CuVQ Mobi…

October 25, 2011 Leave a comment

http://t.co/NPM3CuVQ Mobile synaptic systems – because they can learn to do just right and many things in parallel.

Categories: News

Tips for hard times – car…

October 20, 2011 Leave a comment

Tips for hard times – caring for talent you can’t hire now pays back later! http://t.co/41naZ72c

Categories: News

Will money or air-time wi…

October 18, 2011 Leave a comment

Will money or air-time win the micropayment business? Money http://t.co/bveNjw5S vs. air-time http://t.co/6aakkKk0.

Categories: News

Good engineering – Sustainable business

October 17, 2011 Leave a comment

Sustainable engineering of consumer products means a much reduced amount of production waste, careful choice of good commodity components, integration of technologies with robust and scalable performance , and in general designing for a life together with the customer rather than a brief financial transaction.

Below is a neat introduction worth watching before each valuable day of good engineering:

Categories: Uncategorized

Leave your wallet at home – Android tap-to-pay!

October 15, 2011 Leave a comment

Android now supports tap-to-pay!

Categories: News

Massively parallel processing and optimization

October 15, 2011 Leave a comment

Massively parallel processing and optimization is easy to do with Python using the playdoh module. It can easily vectorize tasks to use all the CPUs and GPUs available on your network.

Network of distributed CPUs and GPUs

Playdoh network of distributed CPUs and GPUs

The module is documented at http://lpp.psycho.univ-paris5.fr/pdf/3022.pdf and available for download at http://playdoh.googlecode.com.

Categories: News

Real-time brain simulation on a PC

October 15, 2011 Leave a comment

Richert et al. report fast and realistic brain simulation for 138k neurons and 30M synapses using a high-end graphics desktop PC.

Spiking frequency in different simulated brain regions in response to different stimulus

Spike-frequency response in simulated brain areas trained on differently oriented input patterns.

Edgars Lazdins has started the process to integrate such a brain into a humanoid robot called iCub. The brain will be very minimal but with some help with the training or more efficient simulator we could soon see it toddling around and explore our world through its own eyes!

Categories: News

Nanosynapses using conventional PCM mass-storage technology

October 15, 2011 Leave a comment

Kuzum et al. suggest that a conventional mass-storage technology using phase-change memories (PCM) can perform like a young rat’s visual cortex.

Biological versus PCM-based synaptic network

Categories: News

Wondered why associative …

October 13, 2011 Leave a comment

Wondered why associative thinking matures so late during adolescence? If we smartened up sooner, we’d end up dumber – http://t.co/OW3LGyTF

Categories: News