Adaptable transistors could reduce CPU transistor count by 85 percent

[Shawn Knight]

What just happened? Researchers from the Vienna University of Technology have developed an adaptive transistor designed to provide more flexibility during run-time. The revolutionary new transistor was produced using germanium, an element on the periodic table with an atomic number of 32. To explain how it works, one needs to start with a basic understanding of an ordinary transistor.

As SciTechDaily highlights, a transistor is a tiny component that either allows current to flow, or blocks its flow, based on whether or not electric voltage is applied to a control electrode. This structure allows for the creation of simple logic circuits, and when you squeeze billions of them into a single package, you get something akin to a modern processor (in the simplest terms, of course).

In standard single-electrode transistors, free-moving electrons carry a negative charge. Individual atoms with an electron removed are positively charged. The researchers’ prototype works a bit differently.

In addition to the usual control gate (red) there is also a program gate (blue).

“We connect two electrodes with an extremely thin wire made of germanium, via extremely clean high-quality interfaces,” said Dr. Masiar Sistani, one of the researchers that worked on the project.

“Above the germanium segment, we place a gate electrode like the ones found in conventional transistors. What is decisive is that our transistor features a further control electrode, which is placed on the interfaces between germanium and metal. It can dynamically program the function of the transistor,” Sistani added.

The researcher further explained that germanium was chosen due to its special properties.

“When you apply voltage, the current flow initially increases, as you would expect. After a certain threshold, however, the current flow decreases again – this is called negative differential resistance. With the help of the control electrode, we can modulate at which voltage this threshold lies.”

The design, Sistani said, enables new degrees of freedom that can give the transistor the exact properties it needs at any given time.

Walter Weber, Masiar Sistani and Raphael Böckle (left to right).

Professor Walter Weber, another member of the team, said an arithmetic operation that previously required 160 transistors is now possible with just 24 transistors thanks to the new design. At that rate, it doesn’t take much imagination to envision how this breakthrough could be scaled to significantly impact efficiency and operating frequency.

“We don’t want to completely replace the well-established silicon based transistor technology with our new transistor, that would be presumptuous,” said Sistani. “The new technology is more likely to be incorporated into computer chips as an add-on in the future.”

More information on the new transistor can be found in a recently published paper on the American Chemical Society’s website.

Image credit: Miguel A. Padrinan

Permalink to story.

https://www.techspot.com/news/92775-adaptable-transistors-could-reduce-cpu-transistor-count-85.html

 

[fps4ever]

But can it be mass produced enough at an economic level...

 

[Raytrace3D]

My money is, not less transistor counts, but more powerful computers with the same transistor density as before. I look back at PC power requirements and laugh over the years. Seems the more power efficient they become, the more power they require because the thermal limits can be increased. Maybe this will the case in laptops or in servers, but certainly not on the desktop market where seems like the sky's the limit. lol

 

[emmzo]

Germanium is pretty hard to find.

 

[yRaz]

Germanium is pretty hard to find.

Germanium isn't that hard to find, it's just hard to purify. It's mixed in with lots of things we already mine like Zinc and Coal. But in the amounts that would be used in chips now that transistors are literally getting to point where we can count the atoms that make them. I don't think Germanium would influence the cost of chips. We already use gold to coat contacts inside our PCs and that really isn't a big deal. 1 ton of germanium could likely make millions of processors.

Another note, reducing transistor count isn't the big deal here. The big deal is that we could pack almost 10X as many transistors on the same die.

 

[umbala]

But can it be mass produced enough at an economic level...

Nothing that is new and revolutionary can be produced at an economic level when they first discover it. It takes years or even decades for that to happen, or it never happens at all. We probably won't know for a while if this new approach is viable in the long run.

 

[fps4ever]

Nothing that is new and revolutionary can be produced at an economic level when they first discover it. It takes years or even decades for that to happen, or it never happens at all. We probably won't know for a while if this new approach is viable in the long run.

Yeah, that is what I pretty much posted as a shorter question. Look at the cost to produce it per ounce vs. others already in mass use and you can see that is a tall hill to climb.

 

[SixTymes]

This is interesting, and sounds promissing. I hope to see this implemented in my life time.

 

[colemar]

Sounds like the pentode concept which was introduced in thermionic tube tecnology to enhance the quality of the triode device.
It seems surprising that nobody tought of this before for transistors. I bet somebody already considered the idea and deemed it impractical.

 

[Danny101]

Seems like too much of a failure point. Like air plane manufacturers going back to buttons rather than the touchscreen that they tried to implement because the system was too...touchy.

 

[human7]

This could be useful when we approach the limits of transistor density. When we get to the point where we can't make transistors smaller than, say 0.34nm (per another article), then reducing the number of transistors that is needed to solve a problem is the next logical step.