The IPO launch of Twitter (TWTR) has investors still buzzing about social networking. Twitter, Facebook (FB), and LinkedIn (LNKD) represent today's big investment trend in consumer and business social networking. But the user growth rates of these social networks are easing. Top-line growth today for Twitter and Facebook depends on increasing monetization of users via targeted advertising, including mobile devices. This can work over the short term, but over time, users will become disenchanted and privacy concerns will increase. Meanwhile, these companies are trading at stratospheric valuations.
I discussed the Big Data and related Machine-to-Machine/Internet of Things investment theme in March of this year. The stocks that I highlighted - Attunity (ATTU) and Datawatch (DWCH) in Big Data, and Telit (OTC:TTCNF or TLM.L) and CalAmp (CAMP) in M2M/IoT, have more than doubled (except Attunity which rose 14%) since the March publication. This theme is valid, but the valuations of leaders like Splunk (SPLK) and followers like Datawatch or privately-held Recommind are pricing in much of the good news.
So where can investors turn to find the next Big Thing? In my opinion, the answer is semiconductor-biotech crossbreeds (SBCs for short). This is a vast area with many different categories ranging from neural networks (see this interview that discusses neural networks among other topics), to labs-on-chips for health monitoring on the skin (see MC10 as one example) or internally via biodegradable chips, to therapeutic delivery via MEMS (micro-electro-mechanical systems)-based sensor devices to cure diseases. SBC's also include genetic mutation screening for detecting and curing diseases, which is enabled by today's super-fast semiconductors and networks that have drastically reduced the cost and time to decode DNA. SBC's can even crossover into some IoT categories in basic health monitoring via RFID (radio frequency identification) or even more advanced sensors as described above.
I am not alone in recognizing this opportunity. In neural networking, Qualcomm's (QCOM) neural processing units are planned to be made available for developers in 2014. One image processing application can supposedly categorize objects from a moving car in real-time. IBM's (IBM) neuro-synaptic chip also mimics the brain with similar applications. Samsung (OTC:SSNLF) is also targeting neural chip design, in collaboration with leading research groups. The common thread is in-memory processing. The typical semiconductor design separates memory from the processor, and wastes a lot of energy and time switching back and forth for each computational task. The brain, however, combines memory and processing in one location (i.e. in-memory processing), within a greater neural mesh network. To some degree, Qualcomm, IBM and others are trying to replicate the brain's processing scheme on a chip. If successful, these chips can process thousands of inputs in parallel at a fraction of the power of today's chips. This would put today's power-hungry, 1x to 8x multi-core chips in the Stone Age, and even leapfrog ARM's (ARMH) low-power processors.
Today, SBCs are still at an early stage. I visited Aerotel Medical Systems at the Mobile World Congress trade show in Barcelona this year. Aerotel's rudimentary wristband-like devices such as a pulse oximeter, blood-pressure monitor, or glucometer connect into a mobile application that can then be transmitted to a monitoring call center. The form factors are bit large and clumsy, and the cost may be too high for mass adoption, but at least it's a start. There is a crying need for improving the health of the masses while reducing health care costs, and Aerotel's telemedicine approach is a step in the right direction.
The big kahuna for medical applications are therapeutics, as opposed to just monitoring. It is true that monitoring and detection of large-scale diseases like Alzheimer's and cancer are each several billion dollar opportunities. But the opportunity for a low-cost, more effective treatment for cancer could be 50x greater.
As a backdrop, back in 1998, I was one of the four founders of a company called Gene Control. Gene Control employed an internal vector-carrying heat-shock promoter, that when switched on, could destroy tumors or blood clots. Unfortunately, we were before our time, and never were able to perfect the system enough to pass human clinical trials. Today, the medical community is starting to look at an alternative delivery system using MEMS sensors to carry a tumor-destroying therapeutic cocktail (radiation, thermal or otherwise). If perfected, this chip-based delivery of a cancer therapy would enable the targeted destruction of cancer cells without destroying surrounding healthy cells and tissue. The cancer treatment time and cost would be less, the nasty side-effects would be reduced, and efficacy would increase. The market size of this opportunity could be over $580B. The math is 7.1B people in the world today x 8.2% who have cancer in some form x $1000 per treatment. Even if after subsidies, certain people cannot obtain this low-cost $1000 treatment, the addressable market is still at least $400B. On the other hand, the price can be raised based on affordability making the addressable market much higher. The point is this is a huge opportunity.
The showstopper issue today is the sensors. Their critical dimensions are too large. Non-planar surface patterning is not flexible enough. One potential solution to this sensor issue, thereby unlocking this enormous therapeutic cancer opportunity, is holographic lithography. Several holographic companies exist today, mainly competing in security and identification. But I know of only one holographic lithography company - Russia-based Nanotech SWHL (SWHL stands for sub-wavelength holographic lithography). Holographic lithography is a simple process that uses one mask, and one pass with perfect replication onto the wafer. SWHL excels at non-planar surface patterning, and can do critical dimensions that are 2½x to 4x smaller than today's existing MEMS patterning systems. In the past few months, Nanotech SWHL has drawn the attention of IBM, STMicroelectronics (STM), and others, but still has more product development work to do.
The SBC investment theme is a bit early for most investors, as there are not any public company pure-plays, and the SBC market has yet to blossom. If I were to choose one public company as a diluted SBC play, it would probably be Qualcomm. In addition to neural networks, Qualcomm is very active in seeding mobile healthcare with its Tricorder XPRIZE competition. Meanwhile, Qualcomm is the technological leader in the mobile semiconductor sector, with solid long-term growth prospects and an attractive current valuation.