Get email delivery of the Cadence blog featured here
The fires in Northern California and floods in the east coast and the heat waves affecting most of the northern hemisphere this summer have me thinking about climate change and climate modeling. After writing last week’s post about quantum computing, my head is spinning a little bit. Climate modeling is a huge subject.
But let’s step back from that a little bit.
Think of all the neat things that you can do with your handy laptop. Buried deep in the processing cores of your computer are a couple billion logic gates. So, for example, if the chip inside your computer were the size of the earth, each logic gate would be the size of a garden gate on a house, give or take. That’s how tiny—and yet vastly powerful—your computer can be.
But, as I pointed out last week, this isn’t enough processing power to model even the simplest of systems. Consider a system with only ten variables with ten possible values for each variable—there are 1010, or one TRILLION ways that system could be configured. Your laptop can’t process that many bits of logic in any reasonable amount of time. Hence the need for quantum computers that can perform that kind of number-crunching all at once.
Now consider a system as complicated as a chip. Obviously, there is no way to effectively model 2 billion logic gates with all the configurations possible before building the thing. So chip designers need to simulate that chip, as best as they possibly can, given the tools that they have. Maybe they have a few servers at their disposal. Maybe a row of servers in a datacenter. Maybe they even have a Palladium of their very own. Maybe they’re hooked into the Cadence Cloud and can perform their simulations in some vast datacenter in the sky. But no matter how the simulations are performed, there will always be bugs, there will always be corner cases, there will always be ways of designing chips (and boards and systems) that are more efficient.
But think of the possibilities if quantum computing becomes ubiquitous! Simulations could basically be unlimited by processing power. Designers, architects, and engineers will know before the chip (or board or system) is ever built that their chip (or board or system) is literally as efficient and as low-power as it is possible to be, taking a fraction of the design time to create it.
I heard some special effects people talking about explosions recently. They were saying how amazing the 9/11 disaster was, purely from the standpoint of design. They were astounded by all of the data they saw in the collapse of the towers. They learned, by watching the scenes over and over, about how ineffective it is to model an expanding cloud of debris by “chunking” the cloud into pieces and manipulating these pieces as units. To make an explosion truly believable, you must model each hair, each grain of sand, each speck of dust, each draft of air in the smoke, and each one has to move independently of the other.
Without the increase in computing power in the last few years, the computer animated films that have come out since 2001 wouldn’t be nearly as realistic as they have become.
But let’s follow this line of thinking to its logical conclusion. Suppose we had practically unlimited computing power from quantum computers, and the only limit to creating animated film was… well, no limits at all. Special effects could be created at the mote-of-dust level, and characters can be created to include every hair, every pore, every miniscule muscle twitch, every subtlety that exists in a character’s face, modeled perfectly and completely. What’s to prevent filmmakers from making computer-animated films so realistic that a viewer couldn’t tell whether it was filmed from live action, or created completely in the mind of some designers?
Nothing, that’s what. Should quantum computing become an accepted reality in the entertainment industry, I think you should expect to see just that.
Being able to predict extreme weather events and the impact of extreme weather phenomena will become a necessary tool for all sorts of circumstances, from city planning to a myriad of logistics to crop rotation, not to mention how to plan your next vacation and whether you need to carry your umbrella next week.
Think of the thought experiment to illustrate chaos theory, about how the flutter of a butterfly wing in China can result in a hurricane in Texas. Isn’t it true that the more variables that you can identify and isolate and predict, the more accurate your climate model could become? The hard part is identifying all the disparate factors; the computations could become easy, using quantum computing.
I wonder if quantum computing could mean the end of chaos theory entirely? I mean, if I could identify all the forces and weights and measures of a leaf blowing in the wind, what’s to prevent me from hooking into the local quantum computer to predict exactly how that leaf will behave as it is falling to the ground? Or, more relevantly, could I eventually predict everything from pork futures to fluid dynamics to traffic patterns to how a particular enzyme may interact with another enzyme? With modeling all that is possible to be known about a system, our results will only be as good as the models we use to simulate them, and how many times we run the simulation.
Being that just the three-body problem still has yet to be solved, we have a long, long way to go. In the meantime, the first thing that we have to do is to imagine the possibilities and follow each line of thinking to its logical conclusion. What’s the end-game of quantum computing? Could it become a modern-day oracle? And do we want it to be?