Breaking News
President Trump Finally Admits It
ALL 21 STATES. ALL 1,799 COUNTIES. EVERY SINGLE ONE SHIFTED BLUE.
The Crisis in Poland
Jefferson, Secession, and "All Men Are Created Equal"
Top Tech News
Prism puts an eye-controlled Linux computer on your face
Remember Yesterday: AI For Today and Tomorrow
Robot Mowers Invade America's Golf Courses, Putting Groundskeepers On Notice
Monstrous 13-foot mecha craves the dangerous jobs humans avoid
Scentronic tech lets you smell the internet
Virgin Galactic's new rocket plane revives its bid for a space airline
The metal kept moving east while the sellers ran, and Ronnie called the flush a week early
Chinese Company Develops $1,000 AI-Powered Laser System To Kill Mosquitoes
PRP vs Exosomes vs Stem Cells: Which Fits Your Condition?
SHOCKING: The Secret (Frantzve) Family Business
Three Eras of Machine Learning and Predicting the Future of AI
They show :
before 2010 training compute grew in line with Moore's law, doubling roughly every 20 months.
Deep Learning started in the early 2010s and the scaling of training compute has accelerated, doubling approximately every 6 months.
In late 2015, a new trend emerged as firms developed large-scale ML models with 10 to 100-fold larger requirements in training compute.
Based on these observations they split the history of compute in ML into three eras: the Pre Deep Learning Era, the Deep Learning Era and the Large-Scale Era . Overall, the work highlights the fast-growing compute requirements for training advanced ML systems.
They have detailed investigation into the compute demand of milestone ML models over time. They make the following contributions:
1. They curate a dataset of 123 milestone Machine Learning systems, annotated with the compute it took to train them.