Our review of the 12th Gen ‘Alder Lake’ Intel Core microprocessors leaves no room for doubt: these chips perform outstandingly, but they just beg to be manufactured using more advanced integration technology. Intel is producing them in its 10 nm node, and, although this company ensures that its technology is comparable to the 7 nm node of TSMC or Samsung, its consumption and temperature figures cast reasonable doubts.
During our tests, the high-performance cores of the ‘Alder Lake’ processors have proven to be real thread eaters (threads), but its fabulous productivity has been somewhat overshadowed by significantly higher peak power consumption than comparable AMD Ryzen 5000 processors. And furthermore, the Intel chips they also get hotter.
The design of the microarchitecture is to some extent responsible for these figures, but there is no doubt that photolithography also has a direct impact on both the consumption of a microprocessor and the energy it dissipates in the form of heat. Intel knows what it has on its hands better than anyone, and for this reason it has unveiled some of the innovations it is working on in a clear attempt not only to give its integration technology the accolade it needs, but also to sustain the moore’s law beyond 2025.
These are the innovations in integration technology that Intel has in hand
Before proceeding further, it is worth pausing a moment to review what Moore’s Law says. What Gordon Moore, the co-founder of Intel, realized just over five and a half decades ago was that the number of transistors in integrated circuits would double every year, and, at the same time, its relative cost would be drastically reduced.
Ten years later, he amended his observation by increasing the time required for this integration technology development to take place, putting it at 24 months, not a year. And since then his forecast has been met with more than reasonable precision. However, to sustain this trend over time it has been necessary to innovate. Innovate a lot and well.
ASML, Intel, TSMC, Samsung, GlobalFoundries and other companies involved in the integrated circuit manufacturing industry dedicate a significant portion of their resources to research and development of new technologies. There are no secrets here: improving integration technology requires innovation. And it is precisely what Intel is doing. These are the three pillars you are working on to achieve a photolithography that allows you to make a difference, and along the way, dilate Moore’s law:
- We owe a lot to silicon. Very much. It is the chemical element that has brought us here and the one that for decades has made possible a technological development that still lasts. The problem is that their physicochemical properties impose restrictions that get closer and closer as integration technology develops. Intel claims to be working on new materials that will only have a few atoms thick and that will allow you to fine-tune transistors designed to overcome the limitations imposed by silicon.
- One of the goals of semiconductor manufacturers is to find a way to introduce more transistors per square millimeter, and one way to achieve this is by not just placing them next to each other; It is also possible to stack them on top of each other using tiny plates that allow double the integration density. This is one of the primary objectives that the RibbonFET transistor architecture will have, which is set to replace the current FinFET, and which could be ready in 2024.
- The third essential area that Intel engineers are working on is none other than optimizing the packaging of their integrated circuits. What they are looking for is increase interconnection density, and the Foveros Direct technology, which according to this company will be ready in 2023, seeks to multiply it by 10 making possible the direct connection of copper to copper conductors, which allows, incidentally, to minimize resistance.
As we have just seen, the innovations that Intel claims to be working on on paper look very good. And furthermore, some of them might be ready in no more than two or three years. Hopefully that is the case because there is no doubt that what interests us as users is that semiconductor manufacturers, whatever they are called, have the most competitive technologies at their fingertips. And if, in addition, these innovations come from the hand of the end of the crisis of the integrated circuits, honey on flakes.