It continues to be hard to buy high-end PC components for nearly 6 months. There are a number of causes of this, including pandemic-related impacts, the related surge in demand for all computing hardware, and supply shortages. A lot of eyeballs happen to be trained on foundries like TSMC, to the stage that national governments have put pressure around the foundry to improve automotive chip production. AMD is one of the companies most-impacted by these production shortages, because AMD designed the SoCs for the recently-launched PlayStation 5 and Xbox Series S|X along with its own RDNA2 GPUs and Ryzen CPUs.

At first glance, this decision to rely on TSMC for its products may seem shortsighted as well as an obvious reason for weakness within the company’s supply strategy. That’s the argument Mark Campbell makes at Overclock3D, where he writes:

Had AMD had use of more fab space, the company could produce more chips and generate more sales by having more sellable products on the market. AMD’s overreliance on TSMC’s 7nm node is hurting them. This is a problem for TSMC, as their lack of 7nm fab space gives their competitors more room to grow and it is leaving demand/money around the table… Later on, AMD will need to consider diversifying its product stack, utilising competing nodes from the likes of Samsung to manufacture a number of its products.

Campbell’s research into the situation makes some strong points. AMD factually lost share of the market in Q4, also it lost share because it’s supply constrained. If TSMC had more 7nm capacity online, AMD might have sold more chips and therefore made more income. He acknowledges that AMD can’t quickly move its products from TSMC, and his call for AMD to maneuver to Samsung “within the future” acknowledges just how long it requires to transition designs from one foundry to a different, even if I didn’t quote it above.

Where I disagree with Campbell is his statement the unique circumstances has “hampered AMD’s growth prospects.” We are able to only evaluate the growth prospects of a company with regards to the rest of the market because it currently exists. It’s correct that additional available 7nm capacity at TSMC would have given AMD more processors to market, however these are not usual times. TMSC has been absolutely slammed with orders as a result of the pandemic, and production worldwide has been restricted to shortages of materials like ABF.

While we don’t have proof of this, there have been rumors that the reason Nvidia opted for Samsung rather than TSMC this cycle was to ensure better supply of its Ampere hardware compared to what TSMC could guarantee. Let’s assume, at the moment, that this is broadly true. Nvidia’s decision to work with Samsung rather than TSMC didn't keep Ampere on store shelves in late 2021. More recent industry reports, filed in the past couple of months, claim that the reason we’re seeing silicon shortages through the market is partly because the foundries themselves are unable to secure sufficient supplies of necessary materials. Initial Ampere issues were blamed on low yields, but that situation can also be rumored to have improved late last fall. If Samsung and TSMC already are manufacturing at peak capacity, an AMD decision to port designs to Samsung would not have solved their shortage problem.

Furthermore, there are disadvantages in dealing with multiple foundry partners. AMD could be paying twice to create the same CPU, and every version would have to be customized towards the foundry it had been built on. Samsung’s 5nm node is different from TSMC’s, also it doesn’t provide the same improvements relative to TSMC’s 5nm node. If AMD wanted to ship hardware at both TSMC and Samsung on the future 3nm node, it would need to design one form of a CPU to make use of Samsung’s gate-all-around FETs (GAAFETs) and something to make use of FinFETs at TSMC.

Dual Sourcing Could Have Made the Pandemic Shortages Worse

There are a couple of potential risks in dual sourcing. First, AMD would be betting that Samsung and TSMC could both hit its desired clock and power consumption targets. If a person foundry were built with a problem doing so, AMD might have no choice but to lower its goals to be able to ship a homogenous product (from a consumer perspective). This isn’t just theoretical. When Apple dual-sourced the A9 from TSMC and Samsung, the TSMC chip had significantly better battery life. If this built phones using modems from Qualcomm and Intel, it limited the performance of the Qualcomm modem therefore the Intel modem wouldn’t look slower. AMD could end up in a situation where it had to limit the performance of a future Ryzen CPU to complement exactly what the weaker of its two foundry partners could ship.

5nm was a full node shrink for TSMC and essentially a quarter-node optimization for Samsung, based on Wikichip. Samsung’s 6LPP, 5LPE and 4LPE are further optimizations and improvements to FinFET technology for businesses who don't want to pay the price of designing for GAAFETs.

The second risk of dual sourcing is more subtle. Imagine a hypothetical situation by which AMD had split its 7nm orders between TSMC and Samsung heading into 2021. For that purposes of this experiment, assume the pandemic plays out more-or-less as it did. With Team Red taking in less of TSMC’s 7nm capacity, Nvidia decides to build Ampere there, instead. The pandemic hits. AMD, like several semiconductor companies, attempts to ramp capacity while introducing services, only to discover one foundry is having difficulty with yields. There isn't any choice to increase orders in the other, because both are already building every chip they can produce for every other customer they’ve got.

In this, attempting to dual-source products actually winds up being worse for AMD, not better. It needed to pay twice to design Zen 3, because Samsung’s 7nm uses EUV and AMD’s design at TSMC does not. It's far fewer chips available of computer would if it had only used the higher-yielding foundry. It has no option to increase production, because said higher-yielding foundry has no additional fab capacity available. In situations like we saw in 2021, where total semiconductor production is constricted globally because of unprecedented circumstances, it’s possible for dual sourcing to become a liability rather than a strength.

Relying on TSMC hasn't hurt AMD’s growth prospects. What hurt AMD’s growth prospects in 2021 was an unprecedented pandemic that sent shockwaves with the semiconductor market by driving a massive demand surge at the same time that production got real complicated.

Campbell concludes his piece by writing “Intel is able to react to changes in demand in a manner that AMD cannot, allowing them to grow regardless of the strength of the competition.” This ignores the fact that Intel suffered its own significant chip shortage back in 2021 – 2021.

The reason Intel had the capacity to handle impact of 2021 today happens because Intel was literally building out new convenience of yesteryear 3 years. Fab 42, Intel’s new 10nm production facility, only came online last October. In 2021 – 2021, Intel also boosted its 14nm capacity by installing new lines in existing factories. When 2021 hit, Intel was able to make use of the proven fact that its new capacity was on the internet and in-place. This was, obviously, a very good thing for Intel — but it’s the best thing that happened because Intel was struck by shortages in 2021 – 2021 and couldn’t immediately respond.

Intel got lucky. Its very own reaction to the shortages of 2021 and 2021 was sufficiently strong to permit it to soak a few of the increased interest in semiconductors in 2021. AMD’s growth prospects have not been restricted to its reliance on TSMC and there are intrinsic risks to dual sourcing at a time when capacity is scarce and production shortages limit production. All the major foundry companies are planning capacity expansions and/or building new facilities, therefore the concept that AMD must jump ship from TSMC to be able to boost production in the long term is untrue.