IBM announced a breakthrough: chips with 0.7nm transistors — a leap to 100 billion transistors on a fingernail
IBM Corporation has introduced a chip manufacturing technology with a transistor architecture of 0.7 nm, equivalent to 7 angstroms. This step marks a new stage in the miniaturization of semiconductor components, where traditional planar structures are giving way to an innovative approach — nanosheet. In this architecture, transistors are not placed on a single plane but in multiple layers, fundamentally changing the layout density.
Record Density and Energy Efficiency
According to IBM estimates, this approach allows placing nearly 100 billion transistors on a chip the size of a fingernail. For comparison, modern advanced 2-nm technologies, which are just beginning to be introduced into commercial production, offer much more modest figures. The new architecture promises a performance increase of up to 50% or an improvement in energy efficiency of up to 70% compared to 2-nm solutions from 2021. This means that for the same computing tasks, power consumption could be nearly halved, which is critically important for data centers and mobile devices.
Commercialization Prospects
IBM notes that commercial production of chips using the new technology could begin within five years. However, this is an ambitious timeline, given that the transition from a 2-nm to a 0.7-nm process requires not only a new transistor architecture but also adaptation of manufacturing equipment, materials, and lithography methods. Nevertheless, if the stated parameters are confirmed, it will pose a serious challenge to competitors, including TSMC and Samsung, which are currently focusing on mastering 3-nm and 2-nm nodes.
My expert perspective: Nanosheet technology is not just an evolution but a potential turning point in the semiconductor industry. If IBM succeeds in implementing commercial production within the stated timeframe, we will witness the emergence of chips that will force a rethinking of many modern approaches to processor design, especially in the fields of AI and high-performance computing. However, the path from a laboratory prototype to mass production is fraught with technological and economic risks, so the five-year forecast should be viewed with cautious optimism.