I attended the inaugural VR & AR World conference as well as the 16th annual Broadband World Forum (BBWF) at the ExCel Centre in London from October 18th to October 20th. VR (virtual reality) & AR (augmented reality) World was "virtually" thriving with excitement and potential, but was a bit short on current economic reality. In contrast, BBWF was a more mundane event, but with established carrier budgets in the billions of dollars behind the scenes making bets in areas like G.fast .
G.fast is a digital subscriber line [DSL] protocol for high-speed broadband transmission of around 300MB/second over copper lines. G.fast speed advantage runs out of steam beyond one kilometer. The cost is quite high, however, to rip out copper lines at the last kilometer of a network and replace it with super-fast optical fiber lines. British Telecom ( BT ) estimates it would cost about £29B ($35B) to replace its copper lines with fiber. BT plans to limit these optical fiber replacement costs, by spending only £6B ($7B) through 2020, retrofitting the last kilometer with G.fast processors and equipment. Beneficiaries of G.fast include companies like Broadcom, now part of Avago ( AVGO ), as well as ADTRAN ( ADTN ). While G.fast is slower than optical fiber, it is much faster than the 30MB-50MB/second speeds that most DSL consumers have today. G.fast, in fact, can be a conduit for millions of consumers to enjoy the new world of VR & AR.
BBWF represents a more mature carrier and network infrastructure industry, where carrier network spending is projected to decrease 3% in 2016. This carrier spending slowdown in part reflects the beginning of a shift to software defined networks (SDN) and network function virtualization (NFV), and away from traditional routing equipment from vendors such as Cisco ( CSCO ). Beneficiaries of this long-tailed transition to SDN and NFV include Red Hat ( RHT ) which provides open-source Linux integration for cloud infrastructure, spanning OpenStack and software-defined networking. For more color on SDN and NFV, please refer to my latest article on this subject.
In contrast, VR & AR is just at the beginning of a long journey, in a market that is expected to grow from a small base today to several billion dollars by 2020. FutureSource projected VR content alone to reach $6B by 2020. Ovum forecasts the total VR market to reach $285B by 2020. In my opinion, 2020 VR & AR revenues will be a fraction Ovum's estimate. Nonetheless, the dynamic growth potential of the VR & AR markets is evident.
Today, consumers are just becoming aware of virtual and augmented reality, mainly through games. This summer's Pokemon craze put AR on the map, albeit in a dumbed-down version. To increase awareness and make VR & AR more widespread, a few things need to happen.
First, the price-points need to come down to mass-market levels. Valve/HTC (2498.TW) VR headset Vive ($800 ASP) and the soon-to-be-released Meta 2 ($949 ASP) AR headset are moving closer to critical-mass price points. Microsoft (MSFT) Hololens ($3000 ASP) had some impressive features such as moving and changing virtual objects with your hands, but is still priced too high for most consumers. First-mover Facebook (FB) Oculus Rift ($800 ASP including Oculus Touch) is also helping to create a viable VR market. But $800 headsets are still out of reach for many consumers.
To be fair, there are some lower-priced, generally less-advanced headsets including Sony (SNE) PlayStation VR ($400 ASP plus PlayStation console), Samsung (SSNLF) Gear VR ($200 ASP including controller plus smartphone, co-designed with Oculus), Google (GOOG) (GOOGL) Daydream ($100 ASP plus smartphone), LG (LGEAF) 360 VR ($250 ASP plus smartphone), and Zeiss VR One ($150 ASP plus smartphone). I was impressed with HumanEyes Technologies' Vuze 360 degree immersive audio/video camera and advanced video-editing stitching technology. I would categorize these technologies as quasi-VR, although they have the most near-term potential. According to Digitimes, shipments of VR-enabled smartphones are expected to soar to 70 million units in 2016, accounting for 5% of global smartphone shipments.
Second, content needs to improve for wider-scale monetization. Vive has been on the market for six months. During that time, Vive partnered with various game companies, two of which have already achieved revenues in excess of $1M. As the breadth and quality of the games improve, the market will grow. Today, the VR & AR markets are still missing the killer app or game beyond Pokemon that will really spike the growth curve. Epic Games/Tencent Holdings' (TCEHY) Robo Recall to be launched in Q1 2017 potentially could be a killer game.
Third, the hardware needs to step up. The latency has to come down to the milliseconds, the power consumption needs to decline dramatically, field of vision needs to expand to closer to 180 degrees, head-set weight and size need to shrink, augmented display technology needs to improve, and telecom networks need to be able to transmit these VR & AR features to consumers. Encouragingly, progress on these fronts was demonstrated at the trade show. The Lighthouse laser tracking for virtual navigation inside Vive is especially impressive.
I chaired an Innovation Panel at VR & AR World, with Christian Brändli of Insightness , Vitaly Ponomarev of WayRay and Dominic Eskofier of Nvidia (NVDA). Insightness makes event-based visual positioning modules that process pixels like the human brain. The company's Silicon Eye vision sensors and proprietary processing algorithms lead to exponentially faster image processing/lower power consumption, while providing 3D spatial awareness of where you are. Insightness can enable next-generation AR and VR hardware to meet the low power consumption, fast millisecond response times, and focal positioning to create a better end-user experience. WayRay builds AR car windshield navigation systems. The company's unique holographic and inertial sensor technology, enable longer and wider fields of view than exist today. WayRay should be a boon to the budding driverless car market.
Nvidia GeForce, Quadro and Tegra graphic processor units (GPUs) are the engines for high-end gaming, including VR & AR. Nvidia, via its Tesla and GRID chips, also has a fast-growing artificial intelligence [AI] business. Nvidia faces competition from a number of GPU competitors like Advanced Micro Devices (AMD) in gaming. For AI markets like data centers, high-performance computing and driverless cars, Nvidia faces competition from GPU-based Nirvana/Intel (INTC), Intel from the computer processing field, CEVA (CEVA) from the digital signal processor angle, ARM/SoftBank (SFTBF) and ThinCI. While Nvidia chips' parallel processing capabilities are impressive, Nvidia's key competitive advantage may be providing software engineers well-documented platforms for building applications on top of its core processors, such as CUDA for parallel programming and G-Sync for displays.
Advances in processors and platforms (Nvidia), displays (WayRay) and image sensors (Insightness) can help VR & AR hardware reach its potential. Nvidia would be the publicly-quoted way to broadly access the future growth of VR, AR and AI, although the stock is not cheap.
Visionaries espoused the potential of VR & AR in the coming decade. Huawei's Richard Li saw three main VR & AR markets - video, gaming and online shopping. Video and gaming have the largest potential according to Mr. Li. For VR gaming, Mr. Li estimated 216M gamers generating $11.5B in revenues by 2025. Shopping has potential, but will require high-bandwidth real-time interaction with objects.
Meta's Ryan Pamplin provided a vivid view of an augmented and virtual reality future. He pictured hardware devices to be 3D (not flat), wide viewing, have interactive sculpting of virtual objects, with the ability to turn on and off application layers that are contextually relevant to the user. By 2020, the form factor will be small enough to put on glasses or lenses, and AR and VR will become more main-stream from on-line shopping and fashion designers to mobile phones. By 2030, Mr. Pamplin said that we will go beyond glasses to human brain-derived images, millisecond latency, multi-gigabit throughput, photo-realistic optics, all processed in the cloud.
See also Depomed: A Gift From The Indexers on seekingalpha.com