"Silicon Frontline": Intel Restarts Packaging Plant—This Move Has Really Changed

This time, with Intel restarting Fab 9, I’d rather see it as a “comeback by shoring up weaknesses” than simply a recovery of production capacity. For decades, Intel’s strongest playbook was always designing its own CPUs, manufacturing them itself, and then maximizing profits through process technology and scale. But as TSMC’s lead in advanced process nodes gradually widened, Intel also began adjusting its strategy, pushing advanced packaging into a more prominent position. For hardware watchers, this is more than just factory news—it could very well affect the future supply rhythm of AI chips, server chips, and even game-related cloud services.

Interestingly, packaging was not regarded as a “highlight segment” for a long time. Over the past decade or so, Intel shut down quite a few packaging plants, or outsourced them directly to other companies, showing that the industry generally believed at the time that this part offered limited technical sophistication and profit potential. But now the situation is completely different: advanced packaging has gone from a supporting role to a key factor in determining the balance between performance, power consumption, and cost. The article mentions that Intel’s EMIB advanced packaging technology is no less capable than TSMC’s, and this is also one of the few chips it currently has to compete head-on in tough battles.

Fab 9 itself carries a strong sense of its era. This plant, located in New Mexico, USA, was originally built in the 1980s, shut down in 2007, and then reactivated in January 2024 as a packaging facility. From the perspective of industry storytelling, it feels very much like an old factory entering a second playthrough—but this time, the mission objective is no longer traditional manufacturing, but serving advanced chip packaging. Intel has poured in billions of dollars for this, including a $500 million subsidy from the U.S. CHIPS Act, which makes it clear this is not a trial run, but a formal push backed by both policy and capital.

What deserves even more attention is that Fab 9 and the neighboring Fab 11X are already seen as key facilities for Intel’s advanced packaging business. One line in the original article is especially on point: “Fab 9 and the nearby Fab 11X plant are now key facilities for Intel’s low-profile but fast-growing advanced chip packaging business.” The real emphasis behind this sentence is not “low-profile,” but “fast-growing.” This is because, at this year’s January earnings call, Intel management already raised its revenue forecast for this segment from less than $1 billion to the level of several billions of dollars. The scale of that adjustment alone shows that internal expectations have changed significantly.

From the perspective of industry competition, the value of this move may even materialize sooner than Intel’s hoped-for revival of its chip manufacturing business. The reason is straightforward: catching up with TSMC in advanced manufacturing requires time, yield, and customer confidence, whereas advanced packaging looks more like a breakthrough point that can be monetized more quickly. Google and Amazon have both been named as potential major clients because they have in-house chip development needs while still requiring external packaging support. Although neither side has commented, the market has clearly picked up on the room for possible collaboration. For the gaming community, changes like this will not directly turn into “cheaper graphics cards” or “plunging CPU prices” in the short term, but in the medium to long term, they could affect AI infrastructure costs, and in turn influence the computing supply for cloud gaming, generative content, and large-scale online services.

If I were to make a judgment on the trend, my view is very clear: this time, Intel is not trying to replicate its old path to victory, but is instead looking for a more realistic new route. It may not be able to use packaging alone to completely reverse its lag on the manufacturing side, but it has at least found a fulcrum that can generate revenue more quickly and make it easier to build customer relationships. In other words, the restart of Fab 9 is not about nostalgia—it is more like a new campaign Intel has opened for itself; whether it can clear this stage depends not just on technology, but on whether it can turn its “packaging advantage” into stable orders.