Dr Peper
Moore’s Law is coming to an end, and many people don’t know how to feel about it. In fairness, the end of Moore’s Law is not something that crept up on us out of nowhere. Industry experts predicted the termination of Moore’s Law years ago. They observed its gradual decline and forecasted a grim future for Moore’s Law that has since proved to be an accurate calculation.
But the question remains… why is Moore’s Law ending? And why should you care?
I’m kicking off the start of the year with a podcast episode that focuses on endings. That is, the end of Moore’s Law and why it matters. As always, I focus on AI in simple terms so that whether you’re new to AI or a seasoned pro, you can follow along fully immersed!
Listen to this episode of Short and Sweet AI below or read on for more…
What is Moore’s Law?
Have you ever wondered why and how technology keeps getting better, faster, and cheaper?
Gordon Moore, the co-founder of Intel, believes that the number of transistors on a microchip doubles every 18 to 24 months. This theory became known as Moore’s Law, and it’s been around for decades.
Moore first postulated the theory back in 1965, and it stemmed from his observation that the computer industry kept creating chips with more transistors on smaller pieces of silicon. Doing so helped to significantly increase their power while keeping the cost of developing chips relatively low. It became cheaper to make the chips despite them having more power. This cycle occurred every 18 to 24 months with no signs of slowing down at the time.
Moore didn’t expect his observation to transform into Moore’s Law, but it happened nonetheless and continued for the last 50 years. Moore’s Law is the reason that you and I can afford to have powerful devices in our pockets, such as an iPhone. To give you some idea of just how powerful the iPhone is, it is 100,000 times more powerful than the Apollo spacecraft’s computer that went to the moon!
Google’s big “OMG” moment
Google has the answer to everything, right? Whenever you need to check something or ask a question, Google can usually answer within seconds.
What you might not know about Google is that the giant search engine has some of the most complex neural networks used today. Five years ago, Google discovered that it potentially had 100 million customers lining up to use their speech and recognition apps. That is a lot of people for one organization to handle, even one as huge and successful as Google.
Google then realized that if these 100 million customers used the apps for just three minutes per day, they would have to double their amount of data centers and microprocessors. It seemed like the multinational technology company had quite the problem on their hands until they had an “OMG” moment that would change everything.
Building all of the data centers they needed to accommodate their users would take years. So, Google came up with a solution to solve their problem. They made a chip solely for deep neural networks ten times more powerful and effective than standard microprocessors. This chip became known as a Tensor Processing Unit (TPU).
Why TPU is the Helen of Troy of AI
Google’s epiphany to create an AI accelerator application-specific integrates circuit, which is essentially what a TPU is, led to Moore’s Law’s inevitable downfall.
Computer pioneer and academic Dave Patterson explained that if Google no longer needs to buy Intel chips, Moore’s Law has ended. Not only is Google not having to purchase Intel’s chips, but other organizations and tech companies are following in Google’s virtual footsteps.
Patterson nicknamed the TPU the “Helen of Troy of AI” because it helped launch a thousand chips. Huge companies like Amazon, Microsoft, and Alibaba, have all jumped on the TPU bandwagon and developed custom machine learning computer chips for their own internal use, just like Google.
So, how did Intel respond to Google’s power move?
Well, they ended up spending billions of dollars purchasing tech companies and investing in computer hardware companies that specialize in creating microprocessors. This came as a shock to many because it was a complete shift of focus for Intel that not many people saw coming.
The end of Moore’s Law: What’s next?
It’s hard to believe that a few decades ago, very few could foresee the end of Moore’s Law. Yet, here we are. Moore’s Law coming to an end means that the type of innovation the tech industry has used for the last 50 years is also ending.
This leads to my next question. If Moore’s Law is ending, what will replace it?
Nobody has a definitive answer. However, experts are leaning towards edge computing and edge AI, where data is processed by the device rather than the cloud. This enables the device to work faster.
Experts are also discussing features surrounding the creation of smart homes using IoT (internet of things) and the rise of quantum computers.
When I think of the end of Moore’s Law, this quote comes to mind:
“What a strange and marvelous world this is, when you stop needing to wrap yourself only in the things you already know.”
While nobody knows for sure what will happen when Moore’s Law ends for good and what shall replace it, it should be viewed as a time for positive and forward-thinking innovation.