14 DAY TRIAL // A new computer chip that is being developed by technology giant IBM allows a computer to actually learn the game of Pong and the strategies that lead to victory and might have a big impact on the gaming world in the long run.
The chip is called Systems of Neuromorphic Adaptive Plastic Scalable Electronics or SyNAPSE and it apparently uses an array of digital synapses as it tries to simulate the actual workings of the human brain.
This means that all processing tasks are distributed among artificial synapses and, just like in real brains, some of the pathways are left unused in the long term while others are strengthened and created learning processes.
The creation of a cognitive computer that uses processes similar to those in the human brain is the aim of a contract that IBM has with the Defense Advanced Research Projects Agency, worth 41 million dollars.
It will create a 10 billion neuron, 100 million synapse machine.
Dharmendra Modha, who is the project leader for IBM Research working on the new chip, has stated, “We want to extend and complement the traditional von Neumann computer for realtime uncertain environments. Cognitive computers must integrate the inputs from multiple sensors in a context dependent fashion in order to close the realtime sensory-motor feedback loop.”
The fact that a computer chip is able to use a self-created strategy to win a game of Pong means that in the future more powerful chips built around the same concepts could learn how to win more complex games.
This could revolutionize the gaming industry and could lead to gaming systems that have the ability to adapt to the strategies used by the human players and provide them with a stronger challenge than current machines can.
IBM is in the early stages of development for SyNAPSE, so it could be 10 years or more before we see any of these applications in the real world.