NVDA 股东们，狼来了一小只？# Stock
a*w
1 楼
纳米光子芯片代替 GPU：
much faster and using less than one-thousandth as much energy per operation
as conventional electronic chips.
粗简演示，潜能很大：
To demonstrate the concept, the team set the programmable nanophotonic
processor to implement a neural network that recognizes four basic vowel
sounds. Even with this rudimentary system, they were able to achieve a 77
percent accuracy level, compared to about 90 percent for conventional
systems. There are “no substantial obstacles” to scaling up the system for
greater accuracy, Soljačić says.
http://news.mit.edu/2017/new-system-allows-optical-deep-learning-0612
much faster and using less than one-thousandth as much energy per operation
as conventional electronic chips.
粗简演示，潜能很大：
To demonstrate the concept, the team set the programmable nanophotonic
processor to implement a neural network that recognizes four basic vowel
sounds. Even with this rudimentary system, they were able to achieve a 77
percent accuracy level, compared to about 90 percent for conventional
systems. There are “no substantial obstacles” to scaling up the system for
greater accuracy, Soljačić says.
http://news.mit.edu/2017/new-system-allows-optical-deep-learning-0612
