Voice-enabled devices and AI were the innovations that everyone embraced over the past two years. In 2019, we see a broader landscape of changes come to the forefront. Some of these advancements are right around the corner, and others will take longer to develop into commercial products. However, each of these advancements provides insight into how our professional and personal lives will change.
1. Foldable devices hit the market. It's been years since we've had a new shape or material in the smartphone space, but that changed in the past two months. At the Samsung Unpacked event in February, Samsung announced the Galaxy Fold. The new device combines the functionality of a small tablet and a smartphone. The folded size is reminiscent of the Nokia communicator but opens into a 7.3-inch tablet. In addition to the foldable material, Samsung highlighted software usability innovations around the seamlessness of switching between phone and tablet modes as well as supporting 3 apps at a time in tablet mode. Given that Samsung’s a leader in the smartphone space, this bodes well for the new foldable category.
However, it takes more than one vendor to make a market. At the Mobile World Congress show, Huawei unveiled the Huawei Mate X. While the audience gasped at the near $2,000 price of the Galaxy Fold at Unpacked, Huawei broke a new record with a smartphone that exceeds $2,500 in pricing.
Make no mistake; these new smartphones are luxury, not mass market, products. A foldable phone will be the edgy technology status symbol of 2019. While only a select group will spring for the expense and first generation device, it’s the first time in nearly a decade that we’ve seen such radical innovation in the smartphone space and that’s exciting.
2. New packaging methods change the dynamics and trajectory of Moore's law. Just when we think we're at the end of Moore's Law, Intel announced a new way of obtaining improved performance in a smaller package. At CES the company showcased its new 3D packaging technology, called "Foveros," which brings the benefits of 3D stacking to enable logic-on-logic integration. Why does it matter? Foveros offers high-performance, high-density and low-power silicon in a tiny package allowing device manufacturers to build a wide range of products.
According to Intel, Foveros is expected to extend die stacking beyond traditional passive interposers and stacked memory to high-performance logic, such as CPU, graphics and AI processors for the first time. Effectively it creates smaller "chiplets," where I/O, SRAM and power delivery circuits that are fabricated in a base die with high-performance logic chiplets stacked on top. One other area that differs is that it offers the flexibility for product makers to "mix and match" technology IP blocks with various memory and I/O elements in new device form factors.
Expect to see the first Foveros products launch in the second half of 2019. It will be interesting to see what device manufacturers do with the technology since it minimizes the tradeoff between size and power. 3D stacking combined with new foldable material could open up a new wave of wearable technology innovation and enable innovation in enterprise IoT devices.
3. AI infrastructure stacks and model management gets easier with new tools. Companies across the globe are evaluating how to use machine learning and deep learning to improve how they analyze data and solve problems. However, the process of standing up the infrastructure for AI and managing a model's lifecycle must get easier. Cloud companies are creating new machine learning services, but you'll also see startups fill gaps. For example, Paperspace offers a suite of tools for exploring data, training neural networks, and running GPU compute jobs. It's Gradient products includes 1-click jupyter notebooks, a powerful job runner, and a python module to run any code on the Paperspace GPU cloud.
Just as the software industry created a set of processes and tools for the software development lifecycle, companies should expect to see a new category of AI software vendors come to the forefront with tools to improve model creation and management. For example, CognitiveScale offers an AI development platform that manages the entire lifecycle of AI application development, testing, and production. It includes capabilities for AI orchestration and composition as well as providing pre-built "AI skills" and industry AI application accelerators to speed up development.
Meanwhile, vendors such as H2O.ai offer a fully open source, distributed in-memory machine learning platform with linear scalability. It offers AutoML and and Driverless AI functionality that automatically runs through all the algorithms and their hyperparameters to produce a leaderboard of the best models.
4. Researchers focus on creating spatially-aware robotic solutions. In January Nvidia opened a new robotics lab in Seattle to tap university researchers and other types of technology talent. Today, robots are in use across the globe but typically work in environments without humans. The goal of the lab is to develop labs that can work alongside people. While this may sound simple, it's incredibly complicated to create a set of models and software that can operate robots in vastly unpredictable environments. The robotics lab is working on around a dozen research projects, including a kitchen project where the robot is learning how to move items, put items away.
5. Quantum computing gets closer to deployable technology. The recent consumer electronics show (CES), in Las Vegas, had more than a few surprises up its sleeve. One of them was a keynote from enterprise industry maven IBM where it unveiled IBM Q System One, an integrated quantum computing system designed for scientific and commercial use. A commercial grade quantum computing system is essential for two reasons. First, a quantum computer can't use a traditional data center. Why? Quantum computers require facilities with cryogenic engineering that deliver a continuous cold and isolated quantum environment.
Second, by delivering an integrated system, IBM offers auto-calibration to provide stable, repeatable and predictable high-quality qubits. Qubits only offer the processing power if they are stable. In the past, the dialogue focused on the amount of processing power (qubits) a computer could have. Today, the discussion should focus on how to make qubits stable and scalable. Having said this, the design of the system also marries function and beauty.
IBM announced partnerships and Proof of Concepts with large customers across a variety of industries. Exxon Mobil took the stage with IBM at CES to discuss how quantum computing can address computationally challenging problems across a range of applications. Potential applications for Exxon include optimizing a country's power grid as well as enabling more predictive environmental and highly accurate quantum chemistry calculations for materials discovery and more efficient carbon capture.
JP Morgan Chase, a global financial company, is working with the IBM Q Network. It's focusing on use cases for quantum computing that apply to the financial industry including trading strategies, portfolio optimization, asset pricing and risk analysis. Daimler AG is working with IBM to advance the potential use cases of quantum computing for the automotive and transportation industry.
Some areas of research include finding and developing new materials for automotive application through quantum chemistry, complex optimization problems such as for manufacturing processes or vehicle routing for fleet logistics or autonomous/self-driving cars, and the intersection of quantum and machine learning to enhance the capabilities of artificial intelligence. While we won't have a quantum computer on everyone's desks next year, quantum computing is definitely on the horizon.
IBM's not alone in its endeavors. Microsoft, Google and startups like Rigetti are researching and designing various aspects of quantum computing. Intel, not to be outdone, has applied its skillsets in computing to quantum as well. During the American Physical Society (APS) March Meeting in Boston, researchers from Intel Labs demonstrated the progress they are making toward developing a commercially viable quantum computing system, including breakthroughs in hardware and software development for the quantum computing stack and development of a tool that will rapidly speed the research and optimization of silicon spin qubit technologies.
These are just a few of the trends you'll see technology vendors working to advance in 2019 and early 2020. Other trends to watch include areas such as robotic process automation, revenue optimization and intelligence and mixed reality.
