Picture the engineers who built the Snapdragon X24 modem this February, announcing the world’s first 2 Gbps LTE chip, while two floors up the company they work for is fighting off the largest attempted takeover in tech history. That is Qualcomm’s 2018 in one image. The product teams kept their heads down and shipped seriously important silicon. The company around them spent the year as the prize in two enormous corporate fights, won one by getting rescued, lost the other to a clock in Beijing, and is now handing tens of billions back to shareholders because it has nothing better to do with the cash. It has been a strange year to cover this company.
Start with the one that nearly ended Qualcomm as an independent company. Broadcom, the Singapore-based chip giant, launched an unsolicited bid last November and pushed it up to roughly $117 billion, which would have been the biggest technology acquisition ever attempted. It was hostile in the real sense: Broadcom ran a proxy fight to install its own slate of directors on Qualcomm’s board, enough to take control and force the sale from the inside. The financing leaned on something like $106 billion in debt, the largest acquisition loan ever contemplated, and Broadcom’s reputation is for buying companies and gutting their research budgets to service exactly that kind of debt. That detail turned out to be the whole ballgame.
In March, President Trump killed it. An executive order, citing a recommendation from the Committee on Foreign Investment in the United States, blocked the deal outright on national-security grounds. The committee’s logic was specific and, I have to admit, hard to argue with on its own terms: Qualcomm is America’s leader in 5G, the next generation of wireless is being defined right now, and if a foreign acquirer slashed Qualcomm’s research to pay down debt, the United States could cede 5G leadership to China, with Huawei standing right there to fill the gap. So a 5G race Qualcomm has been running for years, the one that produced the X50 modem and that whole leaving-the-phone push, became the reason the US government reached in and stopped a private merger. Every Broadcom board nominee was disqualified in the same stroke. I have a lot of complicated feelings about a president personally vetoing a chip merger, but the read on Qualcomm’s strategic value was not wrong.
The deal Qualcomm actually wanted died too, and that one stings differently. Qualcomm has been trying since late 2016 to buy NXP, the Dutch maker of automotive and secure-IoT chips, for $44 billion, the move that would have turned a phone-chip company into a serious automotive-silicon player. Eight regulators around the world approved it, including the EU and South Korea. China’s regulator simply never did. As the US-China trade fight escalated through the spring and summer, Beijing let the approval sit, and sit, until Qualcomm set a hard deadline of July 25th and China let the clock run out without a word. So at the end of July Qualcomm terminated the deal, paid NXP a $2 billion break fee for the privilege of walking away, and authorized a $30 billion stock buyback instead, because returning the war chest to shareholders was the only move left. Two years of work, two billion dollars, killed by silence.
And through all of it, the chips kept coming, which is the part that actually impresses me. The Snapdragon X24 in February is the first Category 20 LTE modem, the first to hit 2 Gbps, and the first chip anyone has announced on a 7nm process, a real milestone dressed up as a routine modem launch. In May, Qualcomm opened an entirely new product tier with the Snapdragon 710, the first of the 700 series, a 10nm chip with Kryo 360 cores, an Adreno 616 GPU, the Hexagon DSP doing on-device AI, and the X15 modem at 800 Mbps. The 710’s whole reason to exist is to push premium-tier features, the AI engine, the better imaging, 4K HDR playback, down into phones that do not cost flagship money, and it is aimed straight at MediaTek in the mid-range. Qualcomm even kept expanding past phones, with the Snapdragon XR1 staking out standalone VR headsets back in the spring.
I am getting away of the Apple royalty war and the FTC case here, because this is an uglier saga and it would need its own three thousand posts. What I take away from this year is the gap between Qualcomm the company and Qualcomm the chip designer. The company spent 2018 getting fought over, rescued by a government, and frozen out of its biggest ambition by a trade war it had nothing to do with starting. The engineering org spent the same year shipping the first 7nm chip, opening a new tier, and carrying forward the on-device AI pitch the 845 made its headline. One of those Qualcomms is at the mercy of forces it cannot control. The other one is still, quietly, the best in the world at what it does. The buyback says the company knows which one is the real asset.