Putting together semiconductor companies is the fad of the season, but success may be fleeting. Multi-billion dollar mergers rarely yield the sorts of success promised to investors, especially in a market where tech is treading water and new innovations – I'm looking at you Internet of Things (IoT) – are as much hype as substance. 

Qualcomm is shelling out $39 billion in cash to purchase NXP Semiconductors, giving Qualcomm a larger footprint into the world of automotive, IoT, security and networking. The combined company is expected to generate annual revenues of more than $30 billion with "leadership positions across mobile, automotive, IoT, RF, security and networking." In theory, Qualcomm will be able to offer more complete system solutions in the aforementioned mobile/auto/IoT and Security, and networking (MAISN) areas, plus better go-to-market capabilities.

From the merger, Qualcomm expected to generate $500 million of annualized cost synergies – i.e. reduced workforce, streamlined marketing and sales, and other cuts – within two years after the deal closes. It's an interesting argument because if I'm doing the math correctly, the megadeal will yield all of 1.67 percent cost savings per year across those $30 billion of revenues.

Some of those cuts may fall more heavily across the part of NXP formerly known as Freescale, since there's a lot of diversity in plant and hardware from the merger between NXP and Freescale. I also have to raise an eyebrow since the March 2015 NXP/Freescale merger announcement also claimed that NXP would find its way to "a clear path of $500 million of annual cost synergies."

The biggest problems on future sales for Qualcomm are a stalled-out mobile market, an automobile market that won't be as lucrative in terms of "devices" as smartphones, an IoT space that has serious security issues and a glut of different business models, and a networking world that is on the move from proprietary hardware to open systems.  There's not a lot of excitement in smartphones, other than Samsung's latest phones bursting on fire and Google trying to convince people it can do mass market with Pixel while Samsung recovers.  Similarly, 5G is less of a standard and more of a dream with about three more years of R&D ahead before carriers have a true handle as to the costs for deployment.  

I'm not sure if we've hit "peak auto," but the automotive industry will be the first to tell you that the combination of ride sharing, vehicle sharing, and autonomous cars is leading to an uncertain future where the U.S. model of two cars in every garage is fading fast.  And new safety and autonomous tech in new cars has find a level between adding cost and better features, with too much cost driving away buyers.  You'll never see as many cars sold as smartphones and while the bill of materials -- in this case, number of chips in vehicle -- per car will be higher than any single cell phone, the sales and profit ratios between automotive silicon and mobile phones have yet to be fully determined. 

IoT is a mess.

More specifically, IoT security needs to get better quickly before fears about a botnet for every home start killing IoT sales at the consumer end and carriers start taking proactive measures to protect networks that start "bricking" IoT devices, resulting in further consumer distrust. I also believe the IoT community needs to move away from its walled-garden approach of owning a proprietary cloud and not talking to other devices, into a more open standards interoperable plug-and-play environment. IoT is a long-term play, not a get rich quick one.

The world of networking is going through its own consolidation, with Broadcom (formerly Avago) snapping up Brocade for $5.5 billion this week.  But the networking world faces its own pressures as Facebook and others continue to push for bare metal hardware and open source software for switches, routers, and other network devices. 

Facebook, through the vehicle of Telecom Infra Project (TIP), rolled out its first white box optical switch for open packet DWDM optical networks this week. TIP members include Intel, Nokia, and a group of network operators including Deutsche Telekom and SK Telecom, with working groups both in core networking and wireless.  Whatever savings Broadcom and Brocade plus Qualcomm and NXP generate through mergers will be pressed by the lower cost of hardware coming out of open projects, making it hard for either new entity to generate substantial growth in the networking sector.