TRADE CHAT Intel’s crazy plan for 1nm silicon
FIVE NEW CHIP PRODUCTION NODES in four years. That was Intel’s plan to get back to technology leadership. It has now added a new 1nm lithography node to its roadmap, due in 2027. But here’s the thing: in terms of products to buy, Intel has achieved little of its original plan. So, what’s going on?
First, we need to understand the true implications of Intel’s existing plan for new chip nodes. In reality, it added up to far fewer than five nodes. Officially, the new nodes in question are Intel 7, Intel 4, Intel 3, Intel 20A, and Intel 18A.
Intel 7 is really just a rebrand and tweak of its existing and troubled 10nm node, Intel 3 is derived from Intel 4, and the same applies to Intel 18A in relation to Intel 20A. In which case, at the time of announcing that seemingly ambitious five-nodesin-four-years roadmap, Intel was only committing to two fully new nodes: Intel 4 and Intel 20A.
But it was still pretty bold, given Intel’s terrible recent track record with new silicon. Back in 2012, Intel was planning on unleashing 10nm CPUs as soon as 2015. In reality, it didn’t launch a truly commercial 10nm product until September 2019 with Ice Lake. It was then over four years before it released a CPU on the next truly new node, known as Intel 4, when its latest Meteor Lake mobile chips stumbled onto the market at the end of 2023.
But Meteor Lake only contains a slither of Intel 4 silicon. Most of the chiplets that make up a Meteor Lake CPU are produced by TSMC, not Intel. Only the compute tile is an Intel 4 chip, so you could say that all Intel has achieved since CEO Pat Gelsinger took over the company and rolled out the new plan, at least in terms of chips to buy, is a tiny volume of Meteor Lake CPU dies on that Intel 4 node.
Sure, there have been other CPU launches, including Alder Lake and Raptor Lake, but those
didn’t get Intel any closer to delivering on that roadmap, based as they are on older 10nm tech. Here we are in 2024, and Intel has until the end of next year to deliver.
The picture gets even weirder when you consider Intel’s most recently revealed plans for its chip fabs. At the Intel Foundry Direct Connect last month, Intel showed a graph mapping out its planned chip production capacity up to 2029, and it was a little bit shocking.
Intel’s capacity to produce chips is shrinking over 2023 and 2024, and won’t again exceed 2023 levels until 2027. The same graph shows that in 2025, chips produced on the latest Intel 4 and 3 nodes (which it brackets together for this data) will be a minority of overall output and, oddly, will have been slightly overtaken by the then-brand new 20A/18A node (again, bracketed together because, really, they are two versions of the same node).
Capacity for both the Intel 4/3 and 20A/18A nodes will grow slightly in 2026, but it’s not until 2027 that Intel expects those new nodes to replace its current mainstay of 10nm technology as the majority of its output.
Put another way, that bold plan back in 2021 now adds up to two nodes in four years, neither of them in large volumes for six years. In that context, Intel’s announcement at the same event that the company is plotting a new ‘10A’ node, which is equivalent to 1nm, for 2027 takes rather different implications.
Intel might be trickling out a few 1nm chips in 2027, but if its planned roll-out of the Intel 4 node is anything to go by, it won’t be making 1nm at scale until 2031— and that’s a best-case scenario. The bottom line is that Intel is at one and the same time nearing the end of its original plan, and yet still has almost everything to prove.
Here we are in 2024, and Intel has until the end of next year to deliver
Six raw 4K panels for breakfast, laced with extract of x86... Jeremy Laird eats and breathes PC technology.