Intel 3 is a new generation node that represents an important advance compared to Intel node 4, which, as many of our readers will remember, was used in Meteor Lake processors. It also represents a major advance in the chip giant’s plan to meet its goal of rolling out five nodes in just four years.

With the Intel node 3, the giant is halfway there if we talk about the launch, a journey that began with the Intel node 7. Before us should be the Intel 20A and Intel 18A nodes, which will be next. The Intel 20A node will experience major changes due to use RibbonFET transistors and also for implementation PowerVia power supply.

Santa Clara has confirmed that Intel Lunar Lake will be based on TSMC’s 3nm node (N3B), so for now only the Intel 3 node will be used, on Intel Xeon 6 processors, which were partially launched during this year’s Computex celebrations and which, as we told you back then, are divided into two large models, one with P cores and the other with E cores.

Intel has shared a new and interesting article on its official blog describing the most important advances it has made with Intel node 3. Compared to Intel node 4, we have an improvement in performance per watt of 18%, which means that this node is able to offer 18% higher performance with the same consumption than an Intel node 4 based processor.

This new node maintains support for high-throughput 240 nm cell libraries and introduces support for high-density 210 nm cell libraries. Support from Via Si Via is another of the added features as well as native 1.2V, Deep N-well support and 3-E pitch and 9um TSV (“through silicon”) which allows for improved 3D stack density. Intel 3 is optimized for high-performance computing and can operate at any voltage less than 0.65 volts as with voltage greater than 1.3 volts.

Intel Node 3 improves density by 10% Compared to the previous generation, it supports a flexible set of metal interconnects and is generally a big improvement, especially considering that it comes only a year apart from the Intel node 4, enough to once again place the chip giant at the top of semiconductor production.

At the moment we have no evidence that this node will be used in other processor families outside of Intel Xeon 6, but right now all doors are open in this regard and eventually will depend on success on the wafer which Intel acquires, i.e. the number of functional chips per wafer, and whether or not it is advantageous to use this node for the production of additional processors.