We have become used to the idea that every year computers get faster, chips get smaller and... well, our iPads get cooler.
But now we may be approaching a physical barrier to progress - and if it happens, the age of silicon might be come to a grinding halt.
According to the co-winner of this year's Millennium Technology Prize, Linux creator Linus Torvalds, the physical barrier beyond which transistors cannot get smaller, and chips cannot improve in the same way, is approaching faster than ever.
And that could have huge ramifications for computing.
The rule of thumb known as Moore's law posited in 1965 that the number of transistors able to be placed inexpensively on a circuit doubles every two years.
For about the last four decades that law has held roughly true, and the speed of computers progressed in a roughly equivalent way.
But now Torvalds, in an interview with the Huffington Post, says that we could soon reach "a completely new point" when computers cannot improve in the same way.
"I believe that within 10 to 20 years we'll hit a completely new point in computers which is when Moore's Law really stops working," he told HuffPost.
"What happens when we hit the wall of physics?"
Torvalds said that aspects of the computer industry are reliant on constant, rapid improvement - but that may be coming to an end.
"When suddenly we can't rely on computers two years from now being twice as powerful as computers today?
"I think that will change the business of technology in a big way."
The end of Moore's Law has been predicted, on and off, for several decades.
Gordon Moore himself said that the law would eventually stop holding true - and the creation of a single-atom transistor in February 2012 seemed to signify that the game was up. No transistor could be smaller than an atom - so once such components could be built cheaply how could the law be kept?
Gerhard Klimeck, head of a Purdue group which simulated the transistor, hailed the discovery in just those terms:
"To me, this is the physical limit of Moore's Law," Klimeck said. "We can't make it smaller than this."
And indeed, a 2010 study argued that the greatest period of change in computing passed in 1998, and that ever since the rate of growth had been slowing.
Torvalds recognises that the limit is not here yet ("It's still several years away" he said). But he added it's something that will only become more important:
"I just think people are seeing it looming…. I don't forsee the need to start from scratch."
Others, however, are more sceptical - including those whose job it is to improve computers (and whose co-founder came up with the law in the first place).
At a recent Intel investor's day, the company showed off its recently launched "tri-gate" three-dimensional processor, and said that there are still new avenues to explore. Others point to advances in biological computing, or the still nascent potential for artificial intelligence as new areas to explore.
"The end of Moore’s Law is always 10 years away," said Intel's lead researcher Mark Bohr. "And yes, it’s still 10 years away."
Torvalds isn't necessarily convinced - but says in the end human ingenuity will win out.
"[Eventually] people will do things differently," he said. "And at some point we'll probably have to come up with new ways of doing things.
"Unix has been around for over 30 years and Linux has been around for over 20, and we do the things we do for fairly good reasons."
"So it's not entirely impossible that in 200 years we'll still have operating systems, it's just by then hopefully no one will care because they'll be working on something new and exciting."