There was no talk of the production being close at once to the physical limits, which Neomon gave a refinement. Now it looks like the new actions predpokldan bariry pror, but it’s a little bit too much looking at vc.
Going back to the past, we can see how much progress the production has made. In 1985, when Intel built the i386, it needed manufacturing technology of 1 micron, 1,000 nanometers. Twenty years later, we were able to cross the 100 nm limit, and now we’re under 10 nm.
Here we have to talk a little about the sizes we are talking about. For example, the width of a human hair is about 80,000,000 nanometers, the diameter of human DNA is 2.5 nanometers, and the diameter of one gold atom is about one part of a hundred nanometers. The first size of the atom is Brno as the limiting factor, which cannot give a variable transistor that it consists of.
This is one reason why IBM is now notifying the sample produced. On the other hand, it is not easy to compare the current generation of production with the previous generation, also because the so-called 3D transistors are used in a number of modern IP addresses compared to the previous generation 2D. Anyway, we can talk about the progress that can change the ipu space with the same sweat transistor, which not only saves space, but also changes the power consumption and speeds you up.
IBM In sv zprvAssuming the new product can deliver up to 45 percent speed and a quarter of energy savings, compared to the currently widely used 7nm production process the modernity uses a production technology called Extreme Ultra Violet (EUV), that is, extreme UV lithography, used In two productions, with only short wavelengths.
Mon production and a few years
As is usual in any similar modernity, first prototype and series production, these are two different things. For example, 5nm IBM ip architecture Announced May 2017 a prvn ip s touto technologi se dostal do spotebitelskho zazen And last fall.
It may take several years, and a factory cannot be built, as they can handle all the pitfalls that make such a pin. Starting a new production is not an easy task, technologically and financially. Advanced production machines similar to such intellectual property cost billions of crowns and require a perfectly trained operator.
IBM will not produce such chips. It focuses on developing intellectual property in its development lab in Albany, USA, and producing it from money in others. Especially in the current lack of IP, when its manufacturers are far from being able to meet the demand, every hour is more effective.