Can biological computers 'extend' Moore's Law?
arts/
As we all know, traditional computers use silicon chips, but now, after several generations of computer technology, scientists are no longer satisfied with traditional computers using silicon chips - scientists have begun to study how to grow biological organic computers in test tubes, And this biological computer made of biological transistors made of genetic materials is already the sixth generation of the computer family.
DNA computing was first demonstrated in 1994 by USC professor Leonard Adleman. Using DNA alone, Professor Aardman has solved difficult problems that conventional computers cannot. After Aardman's experiments, DNA-based electronic circuits have successfully implemented Boolean logic, arithmetic calculations, and neural network calculations. Now, this field, known as molecular programming, is taking off, creating a remarkable future for computers.
After Moore's Law
In 1965, on the occasion of the 35th anniversary of its founding, "Electronic" magazine invited Moore, then director of Fairchild Semiconductor's research and development laboratory, to write an observational review for it, predicting the future of the microchip industry. At this time, the global semiconductor industry was just budding, Intel had not yet been established, and there were only a handful of chips produced and sold on the market. Based on limited data, Moore boldly proposed a roadmap that has been regarded as a standard by future generations - the number of transistors that can be accommodated on an integrated circuit chip will double every 18-24 months, and the performance of microprocessors will double. or half the price . This is the famous "Moore's Law".
In the past half century, "Moore's Law" has made a huge contribution to the development of computing power and even productivity. At the same time, it has also enabled the entire information technology to achieve comprehensive iteration and update, becoming the law of technological innovation and even economics. No matter how much controversy there is, there is no doubt that Moore's Law has been the golden rule of the semiconductor industry since it was proposed for more than 50 years, guiding the development of the industry .
The 4004, the first processor released by Intel in 1971, was produced using a 10-micron process and contained just over 2,300 transistors. Subsequently, the process nodes of transistors decreased at a rate of 0.7 times, and 90nm, 65nm, 45nm, 32nm, 22nm, 16nm, 10nm, 7nm, etc. were successfully developed one after another. Now, transistors are already breaking through to 5nm and 3nm. Perhaps, even Moore himself did not expect that the effect of this law is so durable.
But while the semiconductor industry is advancing by leaps and bounds, it is clear that it is impossible to maintain this growth indefinitely. The "doubling" cycle is all 18 months, meaning that the number of transistors increases by a factor of one hundred every ten years. This is why, for half a century, scientists have also been thinking about the development of new computer models - the technological manufacturing technology of electronic computers will eventually reach the limit .
In the research of exploring non-traditional new computer models, biological computers have attracted the attention of scientists. The biological computer known as the sixth generation computer, its main raw material is protein molecules produced by means of bioengineering technology (especially protein engineering), which is used as a biological integrated circuit - a biological chip.
In biochips produced by protein engineering technology, information is transmitted in the form of waves along the structural sequence changes of single and double bonds in the protein molecular chain. Protein molecules are much smaller than electronic components on silicon wafers and are located in close proximity to each other. Therefore, biological components can be as small as several billionths of a meter, and the density of components can reach 10 to 100 trillion, or even 1,000 trillion gate circuits per square centimeter.
This means that the time required for a biological computer to complete an operation is only one ten thousandth of the current silicon integrated circuit computer. This is indeed the case. The time required for a biological computer to complete an operation is only 1 × 10-11 seconds, which is 1 million times faster than the speed of human thinking.
Moreover, unlike ordinary computers, because the raw materials of biochips are protein molecules, biocomputers have both self-repairing functions and can be directly combined with living organisms. At the same time, the biochip has the advantages of less heat generation, low power consumption, and no signal interference between circuits.
Advanced storage, amazing computing
We already know that the calculation of biological computer refers to the calculation model with biological macromolecules as "data", which is mainly divided into three types: protein calculation, RNA calculation and DNA calculation.
The study of protein computing model began in the mid-1980s, when Conrad first proposed a biological computing model using proteins as computing devices. In 1995, Birge discovered that the bacteriorhodopsin molecule has a good "dimorphism", and planned to design and manufacture a protein computer. Subsequently, Birge's colleagues and other researchers at the University of Syracuse used the prototype protein to create an optoelectronic device that can store information 300 times more than the memory of current electronic computers. This device contains bacteriorhodopsin protein, Information is written and read using a laser beam.
Unlike protein computing, RNA computing and DNA computing are computational models that utilize biochemical reactions, or more precisely, specific hybridization between nucleic acid molecules . However, compared to RNA molecules, DNA molecules are easier to operate experimentally and more convenient to process information on the molecular structure.
Specifically, DNA computing is a computing model that uses DNA molecules and related biological enzymes as basic materials and biochemical reactions as the basic process of information processing . The DNA computing model was first proposed by Dr. Armand in 1994. At that time, Armand published a paper proposing a technique for solving a practical mathematical problem by means of DNA computing. The problem is: connect 7 cities by 14 one-way streets, find the nearest path to all the above cities, and can't go back.
This is a classic problem in mathematics. And the computer can't solve it, because when the number of cities increases, the possible connection paths also increase, and the growth rate far exceeds the growth rate of the city. This would be overwhelming for an electronic computer - it would have to find all possible paths and then compare them to find the shortest path.
Armand's biological computer uses a similar approach, but it operates faster than an electronic computer. Because different base combination molecules are used to define the path, the path code is exactly complementary to the city code; then put these molecules and enzymes into the test tube and let them combine freely, the answer can be combined in just a few seconds, but it is correct mixed with wrong.
Next, Armand spent seven days picking out the correct answer. He first selects the DNA strands of the same length, from which he selects the DNA strands containing the first city, then selects the DNA strands containing the second city from these, and so on. After 7 screenings, he finally obtains the answer to this question. Although the efficiency is not considered at first, it proves that biological computing is not a fantasy after all .
It can be seen that the DNA computer model overcomes the two serious deficiencies of the small storage capacity of the electronic computer and the slow operation speed. First, DNA, as the carrier of information, has a huge storage capacity, and 1 cubic meter of DNA solution can store 1 trillion There are billions of binary data, far exceeding the total storage capacity of all current electronic computers in the world; secondly, it has a high degree of parallelism and fast computing speed. The computing volume of a DNA computer in one week is equivalent to the total computing volume of all electronic computers since the advent of Third, the energy consumed by a DNA computer is only one billionth of the energy consumed by an electronic computer to complete the same calculation; finally, the synthesized DNA molecules have certain biological activities, especially the interaction between molecular hydrogen bonds. Gravity still exists. This ensures specific hybridization function between DNA molecules.
It can be said that every breakthrough in DNA computing will definitely bring immeasurable contributions to the development of human society. The DNA computer based on the DNA computing model will inevitably have massive storage capacity and amazing operation. speed.
very useful biological computer
At present, the biological computer is still in the stage of technological breakthrough. For example, the Massachusetts Institute of Technology and the Singapore University of Technology and Design announced a breakthrough discovery, using an organic virus, they have been able to develop faster and more efficient biological organic computer. Columbia University also announced that they have stored a complete computer operating system on a single piece of DNA.
And researchers from Microsoft and the University of Washington have demonstrated the first fully automated DNA system that can be used to store and retrieve data. The ultimate goal, says Microsoft researcher Karin Strauss, is to put a system into production that will have an end-user experience like any other cloud storage service. .
You know, a tiny DNA smear can hold 10,000 GB of data, which means a data center the size of a shopping mall can be shrunk down to the size of a sugar cube.
DNA is cheap and easy to synthesize, and computing with DNA requires far less energy than silicon processors. A Google data center might consume millions of dollars' worth of energy a year, while a bioorganic computer might only need a few cheap metabolites to run.
In addition to utilizing DNA for data storage, bioorganic computers have potential roles where silicon computers cannot. A research team from ETH Zurich has used CRISPR gene editing technology to build a functioning dual-core biological computer inside human cells.
Imagine having a living computer inside our body that monitors our health, repairs damaged tissue, and regulates our bodily functions. We might even use bioorganic computers to boost our intelligence. DNA computers can also interact with the biochemical environment, allowing us to deliver drugs and treatments inside living biological tissue.
In fact, Israeli researchers have taken the next step by successfully building biological circuits inside cockroaches. The DNA they created can be folded together like Japanese origami, which allows nanobots to deliver some payload. Those payloads are likely to be a molecule, an enzyme, or an antibody. And each payload can activate or deactivate the next nanorobot along the chain, so they build a circuit inside a living cell.
Such biological circuits can be used in many ways. For example, in one experiment, scientists used this biological circuit to identify whether certain cells were cancerous or not, and then they sent a self-destructing signal to the cancer cells. In the future, similar biocomputing devices may be used to non-invasively monitor tumor development, while also delivering targeted drug delivery to specific sites.
However, despite this advantage, it may be a long time before we see biological computers on our desktops. These studies are only at the beginning of a long road. Fortunately, with the maturity of related sciences, bio-organic computers are still striving to become a general-purpose computer, and gradually develop more unique applications suitable for the capabilities of bio-organic computers. .