One of my electronics engineer colleague complained (jokingly) that in the last 10-15 years there have been no revolutionary "ideas" in the world of electronics, but that we have only annually limited ourselves to reducing the size of transistors more and more so as to have smaller and faster chips.

What do you think?

 

I think that you need to laugh at his/her joke, even if it is not funny

He/she might end up being your boss one day

 

Nonsense!

In an apocryphal tale from the late 19th century, the head of the US Patent Office wrote in his letter of resignation to President McKinley that the US Patent Office should be closed because "everything that could be invented has been invented". It would be an understatement to say that this particular conceit was premature.

 

Todays GPU based systems are a good example:

https://images.nvidia.com/aem-dam/e...ure/NVIDIA-Turing-Architecture-Whitepaper.pdf

 

in the last 10-15 years there have been no revolutionary "ideas" in the world of electronics,

I would think that the concept of quantum communication and quantum electronics (whether it can be made practical or not) can be considered revolutionary.

 

Revolutionary? No, I agree with him. Most of the basic circuits we use today were done originally with tubes.

I have would say microcontrollers have revolutionized electronics across a broad spectrum of applications, but I can’t think of anything new in the last 10 to 15 years with such wide usage.

 

What do you think?

I think your colleague is uninformed because he hasn't worked in the field.

One of my electronics engineer colleague complained (jokingly) that in the last 10-15 years there have been no revolutionary "ideas" in the world of electronics

FinFETs were invented by Intel for the 22nm node. Since then things have progressed to GAA (Gate All Around) and contacted gate; just to name a couple.

At the far end of your 15 year limit (maybe a little beyond),

1. multiple VT devices were implemented,
2. strained silicon was implemented to improve P device performance, then N (gate working metals were changed from polysilicon).
3. Many lithographical methods were required to be able to continue to manufacture smaller and smaller feature sizes using 193nm light sources.
4. Copper metal was introduced (this was a big deal due to the risks from copper contamination - I remember seeing signs restricting where copper could be in the fab).
5. lowK dielectrics were developed
6. New architectural features are always being implemented.
7. Intel implemented a gate last manufacturing process (for better gate characteristics control) that others didn't do because it was too difficult.
8. We found that diodes used to protect gate oxides during manufacture became ineffective and gated NAC diodes had to be implemented.
9. Nwell resistors were replaced with gate blocked nwell resistors that allowed better control of the resistance.
10. MIM caps were introduced to get supply decoupling without taking up die area.
11. power throttling got more sophisticated, entire cores could be turned on and off
12. new memory technologies: Optane (came and went), STT (not sure if it's available yet, it was a big deal when I retired)
13. FLASH is using an absurd number of layers
14. after being very very late, EUV finally became mainstream (sort of)

EDIT: edit for readability

we have only annually limited ourselves to reducing the size of transistors more and more so as to have smaller and faster chips

Reducing feature size also reduced power requirements and allowed lower operating voltages. Manufacturers gave up on pursuing higher clock rates and focused on more transistors instead.

 

SiC and GaN devices for high voltage applications are revolutionary developments in both material science and electronics.

 

I think we have a semantic problem here. I would call all of the mentioned technologies incremental, not revolutionary.

Solid state was revolutionary. It replaced nearly every application of tubes over the years.

Same with ICs replacing discrete transistors.

And I already mentioned micros, which have replaced discrete logic.

To me, revolutionary means it changes everything.

 

Well I'd say ultracapacitors are somewhat revolutionary.

 

I would say they are the definition of incremental.

 

I would say they are the definition of incremental.

With hindsight, I must agree with you!

 

The technology was also originally developed in the 1950s, which surprised me. So not new in the last 10to 15 years.

 

The GPU Architecture and Programming Models have IMO revolutionized supercomputing.

https://www.nature.com/articles/s42256-022-00463-x

 

Programmable logic, PAL, GAL, and FPGA have also revolutionized some segments.

 

Gallium Nitride Transistors for 60GHz+ operation

Silicon Carbide and GalliumNitride for High ELectron Mobility Transistors for fast switching of high current applications. These are just getting into consumer electrons. Allows very small, high-current power supplies. Silicon carbide also work well in high temp environments.

Some incredible devices are already on the market but, based on demand for Gallium-based raw materials, we'll see more (cooler and faster) soon.

 

I think we have a semantic problem here. I would call all of the mentioned technologies incremental, not revolutionary.

Solid state was revolutionary. It replaced nearly every application of tubes over the years.

Same with ICs replacing discrete transistors.

And I already mentioned micros, which have replaced discrete logic.

To me, revolutionary means it changes everything.

Exactly! I think he meant something like this. Very interesting!

 

smaller is revolutionary already

 

The definition of revolutionary doesn't require that "it changes everything".

 

Definition c is the one that applies. I agree that it does not have to change everything, but at least it should substantially replace something in current use in most cases.

The iphone was revolutionary, it relegated most old phone designs to the junk pile. No other phone since has been revolutionary, despite maketing ckaims.

Gallium nitride transistors, for example are not going to replace most transistors.