This site may earn affiliate commissions from the links on this folio. Terms of utilise.

Silicon wafer

Nosotros've discussed AMD'due south 7nm plans several times this year, including the company'due south conclusion to source its 7nm silicon for Vega's machine learning iteration from TSMC. Every bit Extreme Ultraviolet Lithography (EUV) ramps up and begins to enter production, a number of foundries including TSMC, Samsung, GlobalFoundries, and Intel are pursuing their own strategies and introduction plans. One interesting tidbit in a contempo report sheds a bit more than calorie-free on AMD's 7nm plans and suggests we'll come across the company pursuing dual-source strategy well into next year.

In an commodity at EETimes, GlobalFoundries admits using like pitches and SRAM cells to TSMC's 7nm, precisely considering information technology gives AMD more flexibility. The visitor'due south chief technologist, Gary Patton, told EET's Rick Merritt that AMD "volition have more need than nosotros take capacity, so I have no problems with that." The same article also confirmed that AMD volition provide its get-go 7nm chip with tapeout expected later this year. IBM hardware and other ASICs will follow in 2022.

When I toured GlobalFoundries earlier this year, the company spoke to the divergence between its strategy for 7nm FinFET designs, where EUV is being introduced first, and its plans for fully depleted SOI (FD-SOI). Right at present, GF offers fully depleted SOI at 22nm, with 12nm ramping upwardly into full production. GF, however, made it clear that it's transitioning to FinFET technology nodes as a fast follower, while its focusing on building upward its 22FDX and 12FDX as an manufacture differentiator — a unique capability that GF has and other companies don't.

cmos-product-roadmap-overview

While a full word of the differences between FD-SOI and FinFET is beyond our scope hither, FD-SOI has some capabilities that brand information technology attractive to analog and RF circuits, which don't calibration well on FinFET designs. Information technology also offers much lower blueprint costs (wafer costs estimates and comparisons vary) and, in the instance of GF's 12FDX, can burst upward to providing FinFET-equivalent functioning over brusk periods of time. It'southward potentially favored for low-ability IoT devices and automotive applications, and given the stiff competition GF faces from TSMC and Samsung, information technology makes sense that the company would be looking for an alternative way to stand up out.

But it's also interesting to run across GF moving to align its 7nm roadmap in a way that makes it easier for AMD to design at both TSMC and GF, considering that'due south precisely what foundries typically don't do. With a few abortive exceptions, like the Common Foundry Platform (now defunct), a chip built at one foundry but tin't be built at another without a total redesign. This is why companies don't typically move production from i firm to another, even in the face of substantial delays. The implication here is that while AMD and GF are no longer formally attached or affiliated with each other, AMD remains i of GF's largest and most important customers.

The optics of GF's decision to fast-follow at 7nm FinFET are clouded past the fact that mainstream CPU and GPU designs aren't the major drivers of new nodes anymore. Ten years ago, it was common to see AMD and Nvidia GPUs as some of the kickoff cutting-edge node shipments when TSMC had a new process to play with. Mobile hardware now dominates these early on shifts. It'south not clear which 7nm hardware will send first, only mobile SoCs, not GPUs or CPUs, have been the pipecleaners of choice since 20nm.

The implications of Patton'southward annotate, generally speaking, is that AMD volition return to split 7nm manufacturing between TSMC and GF, with its APUs and CPUs probable built at one visitor and its GPUs at another. Betwixt the ii, it may make more sense to build its 7nm GPUs at TSMC, if that's the company too handling Nvidia'south. While its CPU competition with Intel remains critical to its overall lesser line, AMD likewise knows that Intel's 10nm procedure is delayed. AMD has ever preferred GlobalFoundries for its CPUs — information technology moved Krishna and Wichita to TSMC but later the GF design failed — and we suspect the visitor will go on to employ GF for CPUs and APUs at 7nm and below.