It appears there’s a new brain-boggling development on the semiconductor front almost every day. I don’t know about you, but I’m not sure how much more boggling my brain can take.
The first point-contact transistor was invented in 1947, which is 77 years ago as I pen these words. This was followed by the first bipolar junction transistor (BJT) in 1948. The first metal-oxide-semiconductor field-effect transistor (MOSFET) followed in 1959, which was the same year the monolithic integrated circuit first saw the light of day.
The semiconductor market grew steadily through the 1960s and 1970s, picking up pace in the 1980s, ramping up in the 1990s, and going exponential in the 2000s. Now, new advancements are coming from all directions at an ever-increasing rate.
I remember when the 1µm process was commercialized in the mid-1980s (this measurement—one millionth of a meter—represented the smallest structures that could be constructed on a silicon chip). At that time, industry pundits predicted that we were reaching the limits of what was possible. They said the same when we reached the 800nm node in 1987 (where 1nm is one billionth of a meter), the 600nm node in 1990, the 350 nm node in 1993, and… you get the idea. Now, the cutting edge is the 3nm node, with the 2nm node expected in 2025/2026 and the 1nm node potentially making an appearance circa 2028/2029.
As of June 2023, the highest transistor count in a consumer microprocessor was 134 billion transistors (in Apple’s dual-die M2 Ultra). As of 2024, highest transistor count in a graphics processing unit (GPU) is 208 billion transistors (in Nvidia’s GB200 Grace Blackwell). And don’t even get me started on wafer-scale processors like the Wafer Scale Engine 2 from Cerebras with its 2.6 trillion transistors!
Semiconductors have changed our world. Anyone younger than 30 years old cannot remember a time without the internet with its ability to provide the answer to almost any question within seconds. Anyone younger than 20 cannot remember a time without GPS to tell you how to get where you want to go and without smartphones that can play music, stream videos, and allow your mother to contact you 24/7 (“Oh, I’m sorry dear. You say it’s 3:00 a.m. where you are. I always forget that pesky 6-hour time difference!”). Yes, I do speak from experience.
We are still in the early days of mixed reality (MR), which encompasses augmented reality (AR), diminished reality (DR), virtual reality (VR), and augmented virtuality (AV). A real game changer will occur when MR is combined with generative artificial intelligence (GenAI—think ChatGPT) and mmWave 5G and 6G cellular communications. These technologies will change the way we interface with our systems, the world, and each other.
This month’s semiconductor feature offers a teasing taste of a few of the things that are happening in semiconductor space (where no one can hear you scream). The exciting thing about all this is that we haven’t seen anything yet. I, for one, cannot wait to see what comes next!