1965 was a game changer in our current Digital journey. We just didn’t know it then. Gordon Moore set the stage for microprocessor & computing revolution. His prediction was:
The complexity for minimum component costs has increased at a rate of roughly a factor of two per year. Certainly over the short term this rate can be expected to continue, if not to increase. Over the longer term, the rate of increase is a bit more uncertain, although there is no reason to believe it will not remain nearly constant for at least 10 years.
He was wrong! We didn’t double transistor density for 10 years while retaining the same cost. We did it for 50 years through computing innovations and we will continue to do so for the next few years, as promised by future technology predictions. Starting with 3,000 transistors on IC chip (Intel 4004) to 19.2 BILLION transistors on the 2017 AMD’s Epyc in 4 decades. Transistors are also counted by the size of the node – we progressed from 10 m in 1972 for Intel 8008 to 7 nm chips. Moore dreamt up future computer technology in terms of accessibility of chips to build modern computers but little would he have imagined that technology companies would bring technology advancement to commonly used devices. iPhone X is built on 10nm processor, just a couple of years behind cutting edge server processors that have started 7nm adoption.He was wrong! We didn’t double transistor density for 10 years while retaining the same cost. We did it for 50 years through computing innovations and we will continue to do so for the next few years, as promised by future technology predictions. Starting with 3,000 transistors on IC chip (Intel 4004) to 19.2 BILLION transistors on the 2017 AMD’s Epyc in 4 decades. Transistors are also counted by the size of the node – we progressed from 10 m in 1972 for Intel 8008 to 7 nm chips. Moore dreamt up future computer technology in terms of accessibility of chips to build modern computers but little would he have imagined that technology companies would bring technology advancement to commonly used devices. iPhone X is built on 10nm processor, just a couple of years behind cutting edge server processors that have started 7nm adoption.
Every good party comes to an end and so will our current computing journey of computing innovations in the next 4 to 5 years. Laws of Physics is to be blamed (or understood!). The issue isn’t so much as to the technical feasibility of getting higher chip density. Post 7nm, we have 5 nm, 3 nm, 2 nm and 1.5nm and that’s going to take 6-7 years to commercialize (at least until 3 nm). We run into technical challenges on the high-K dielectric layer (distance between transistors). For 10 nm density, high-K dielectric is 0.5nm or 2.5 times the diameter of the silicon atom. 5nm and 3 nm designs can be executed at tolerable densities. If we need to continue packing more transistors, we need a new technology and that’s where economics kicks in. We don’t have one but yet might get to it in the next few years.
The end of silicon chips have been a part of future technology predictions in 1999, 2014 and many years in between. We might still overcome it through new material (carbon - ?) or reaching economic feasibility with new technologies of transistor etching onto silicon.
Want better news regarding computing innovations? We are investing in alternate computing technologies and we shall examine them in the coming blogs. After all, our ability to create computing innovations is what we have always lived on.
Progress forward …