Could particles (molecules), each of which is a computer component in itself, be able to bring even greater growth in computing power in the next six decades?
Over the past six decades, record miniaturization of transistors has provided unprecedented growth in computing power. It is possible that microprocessors, each of which is a computer component in its own right, could bring even greater growth in computing power over the next six decades.
Atomic-scale computing, in which computing processes are carried out in a single cell or using circuits containing single-atomic surfaces, holds enormous promise for the microelectronics industry. This allows computers to increase their computing power through the development of components in nano- and pico-metric dimensions. In theory, atomic-scale computing could provide computers more powerful than today's supercomputers.
"Researchers in the field of atomic computing are now at the same point where transistor researchers were before 1947. No one knows where we are headed," notes Christian Joachim of the Center for the Development of Materials and Structural Studies (CEMES) at the French National Center for Scientific Research (CNRS) in Toulouse, France.
"Joachim, head of the group for nanosciences and picotechnology, currently coordinates a team of researchers from fifteen industrial and academic research institutes in Europe whose groundbreaking work on a molecular alternative to transistors brought the vision for atomic computing one step closer to reality. Their efforts, the continuation of the work that began back in 1990, are financed today by the European Union.
In ordinary microprocessors - the "engine" of modern computers - transistors are the essential structural units for digital circuits by creating logic gates that process "false" (0) or "true" (1) signals. To produce a single logic gate several transistors are required, and modern microprocessors contain billions of them, each of which is 100 nanometers in size.
Transistors have continued to be miniaturized since Intel founder Gordon E. Moore famously predicted in 1965 that the number that could be placed on a single processor would roughly double every two years. However, it is not impossible that there will come a point in time when the laws of quantum physics will limit and limit any possibility of further minimization using the methods that exist today. At this point, atomic-scale computers begin to operate using completely different methods to solve this limitation.
"Nanotechnology is the ability to take a physical component and minimize it to the smallest possible size. It's a "top-down" approach," the researcher notes. He and his research team turn this approach on its head, starting with the single, isolated atom and understanding whether such a tiny part of matter could be a logical gate, a memory store, and the like. "This is a "bottom-up" approach or, as we call it, "bottom-down" because we are not interested in reaching the scale of the material itself," he explains.
Joachim's group focused on taking a single cell and building computing components "on top" of it, with their final goal being the construction of a logic gate in a single cell.
How many atoms do you need to build a computer?
"The question we asked ourselves was how many atoms do you need to build a computer?" Joachim explains. "That's something we can't answer right now, but we're gaining a better understanding of it."
The team was able to build a simple 30-atom logic gate capable of performing a task equivalent to fourteen transistors, while investigating the architecture, technology and chemistry required to achieve such computing power within a single die.
The group focuses on two main structures: one that imitates the typical set-up of a logic gate, but in an atomic way, including nodes, circuits, networks, etc. and the second, a more complex one - a process that relies on changes in the isolated configuration to receive the input of the logic gate and quantum mechanics to perform the calculations themselves.
The logic gates are connected to each other using a Scanning Tunneling Microscope (STM) and an Atomic Force Microscope (AFM) - devices with a resolution of a hundredth of a nanometer capable of measuring and moving individual atoms. As a side project, which partly stemmed from pure fun and partly from the interest to develop new avenues of research, the researcher and his team used the method to build nanomachines such as rims, gears, motors and nanovehicles that each contain a single particle.
"Place logic gates in them and they will be able to decide where to move," explains Joachim, pointing to what could be the first global application of atomic-scale robotics.
The importance of this group's research has led to widespread recognition in the scientific community, although Joachim is careful and points out that this is a very basic field of research. It will take some time before commercial applications result from this research. But they have full confidence in this.
"Microelectronics needs us if logic gates - and following them microprocessors - will continue to be miniaturized," notes Joachim.
8 תגובות
rummy:
There is a huge difference between the claim that everything that can be invented has already been invented and the claim that it is impossible to invent anything that one wants to.
While the first claim is apparently wrong, the second is certainly true!
You are trying to attack the second claim using a proven wrong quote that supports the first.
This is a false argument.
Do you think there will ever be a device that will allow you to reach the past and prevent your birth?
The question (a question, not an argument) is whether it is possible to invent as much as we can. It seems you claim yes, where do you get it? This is a physical question and not an engineering question.
It can be argued on the other hand, that nature does not provide materials according to requirements, and there are no other suppliers on the list...
One point, I want to remind you of the famous story about the director of the US Patent Office who declared before the US Congress at the beginning of the century that humanity has already invented everything that can be invented!!! As he was wrong, your argument is also inaccurate - there are not many things that cannot be invented.
Regarding a breakthrough in quantum computing technologies.
I have an interesting question, can it be said that there is such a technological law that says that every breakthrough will eventually come, or not, and then we just waste time and hope (..) trying to create something that is impossible.
Pico, because the channels in question are of the order of electrons. (Each electron affects). I think so..
As written in the article, Higgs, in the matter of wiring, today AFM is used to create nano or wires.
What I don't understand is the concept of picotechnology: if a single atom grows to the size of a single angstrom (tenth of a nanometer), how can technology be created from materials that are a hundred times smaller than the size of a small atom? The concept could be a translation error.
On the other hand, I don't understand the great innovation in a thirty-atom logic gate. Already today we are talking about 45 nm transistors. I guess 30 atoms is about that size and maybe even bigger than that.
We are waiting and expecting a real breakthrough in quantum computing that will boost modern technology to a huge degree, as happened with the beginning of the use of electricity in the era of the technological revolution at the end of the last millennium.
Greetings friends,
Ami Bachar
Unfathomable for practical understanding. (How do you say? Above and beyond or nano from below?).
Nice, but how do you solve the problem of linking between the components, the wiring must be of a different type as well, otherwise there is not much point in it