Beyond Moore’s Law

We’ve known (referred to as “Moore’s Law”) that the number of transistors on a device will double every 18 months almost since the beginning of computers.

We all understand that there has to be a limit to this, though. A transistor that is smaller than, say, an atom cannot be constructed.

However, as of right now, the “law” is still in place.

There are indications that it is slowing down, though, and Intel predicts that it will finish in or around 2029.

After then, computers won’t continue to get cheaper and quicker over time.

This will be a weird thing to get used to.

We’ve all learned to rely on this consistent growth, and our industry will change significantly in the absence of it.

Will mobile phone manufacturers, for instance, still convince consumers to buy new models every few years even if the old ones are not any better or less expensive than the originals?

The most likely response is that software will need to advance and become more intelligent, requiring more computations to be performed by distant servers. Although we can already kind of see that tendency, it needs to continue much farther.

Additionally, I would anticipate some architectural changes in computer design, such as a greater focus on processors like GPUs, which are groupings of comparatively basic machines operating in enormous parallel fashion.