DailyPost 2944
ALPHACHIP – DESIGNING THE AI WORLD!
While we have remained glued to our real world, the real world is fast shifting into a totally different arena. It is like the distorted reality field of Steve Jobs. His distorted reality field became reality and the world’s reality was decimated. It is the broad framework of the world we live today and nobody in his wildest dreams can call it unreal. The same process is unfolding in many more fields today by the all pervasive tech called artificial intelligence. We have been celebrating ChatGPT and indirectly LLMs as the bee-all and end-all of artificial intelligence and predicting protein structures walks away with the Nobel Prize.
Suffice to say, that the fields where AI is breaking the glass ceiling are difficult to count, the most remarkable ones are coming out in the open. One such area is chip design. One AI tool from DeepMind fold has transformed computer chip designing for all times to come and is progressing exponentially as we speak. This may be one of the most fascinating tech revelations of the AI age; “Computer chips have fueled remarkable progress in artificial intelligence, and AlphaChip returns the favor by using AI to accelerate and optimise chip design.”
People familiar with DeepMind’s Alpha story would be sure by now as to what future AlphaChip holds for the chip world. It all started in 2020 with the introduction of a novel reinforcement learning method to design chip layouts. The model has now been named AlphaChip.This method has already been in use to design superhuman chip layouts in Google’s custom AI accelerator, the Tensor Processing Unit (TPU). AphaChip’s another claim to fame is that it was one of the first reinforcement learning approaches used to solve real world engineering problem.
Chip designing is extremely complex and for this reason chip designers have literally struggled to automate the chip floor planning process for over sixty years. The beauty of this tool, as in the case of AlphaGo and AlphaZero, is that starting from a blank grid, “AlphaChip places one circuit at a time until it’s done placing all the components. The novel “edge-based” graph neural network allows AlphaChip “to learn the relationships between interconnected chip components and to generalise across chips, letting AlphaChip to improve with each layout it designs.” Chips are at the core of the AI world, it heralds a bold new age of computing. AI fuels its own growth, giving exponential push to all its components while simultaneously increasing its concentric circles; the potential area of influence.
DEEPMIND BELIEVES IN CRACKING INTELLIGENCE TO CRACK EVERYTHING.
Have a nice evening.