A given chip has to face many problems when imported into a high-quality product. From heat to varying vibrations, to the frequencies it reaches, it all becomes increasingly complicated to work because reducing transistors and increasing population may require testing, measuring, and testing of each chip in a very complex way, that is, each lithograph adds some problems. .
5nm as a 7nm follower represents only some important changes
The biggest among them is that the transistor fin itself must be physically strong. Reducing each end size and gate requires a series of larger larger layers if standard FinFET transistors are maintained.
The twins and gates extend, becoming taller, so the number of materials and their placement is important to maintain the current flow uniform on the three walls of the said door. We can assume that by increasing the amplitude, it is likely to exceed the maximum current flow of each transistor, but the fact is that the ratio of diameter and height is key to increasing its efficiency, since the ratio is not proportional to the performance. .
This poses a control problem on the walls to ensure that the transistor controlling process is correct. To avoid this, artificial intelligence is used to machine learning to detect specific problems before a liar completes some important processes, so that it can be known in advance if it is a malicious failure.
The second problem facing 5nm is curiosity, since it is the transfer of data from the wafer and chip itself. Here again, Deep Learning takes the lead, because it can detect through machine vision images and through specific algorithms the problems and points of interest that developers have to focus on to perform the node production process.
This fact has created a timely shortening of the delivery of this 5 nm lithographic process, because thanks to deep readings TSMC has been able to detect and anticipate where the most serious problems were without having to spend huge hours based on evidence and critical recording errors.
3nm could be a paradigm shift for a new generation of places
