Time and time again, the tech world has a way of reminding us that the development of groundbreaking technologies can be a very gradual process, rather than something for which a switch is flipped overnight.
Bloomberg delivered one more reminder of this kind on Tuesday, when it reported that Samsung (SSNLF) is working on a second-generation foldable phone that will have a 6.7-inch main display "that shrinks to a pocketable square when it's folded inward like a clamshell." The report follows several months of delays for Samsung's first-gen foldable, the Galaxy Fold, after numerous reviewers reported that their screens became damaged following a period of regular use.
Huawei, it should be noted, has also delayed the launch of its first foldable, the Mate X. Aside from screen reliability, issues such as price, device weight and thickness and software/UI optimization will need to be addressed before foldable phones are ready to go mainstream. Though only time will tell for sure, there's a good chance that Samsung's second-gen won't fully address all of these challenges either, but it could make meaningful progress towards addressing one or more of them.
Nvidia (NVDA - Get Report) , meanwhile, recently provided fresh examples of how real-time ray tracing and deep learning super sampling (DLSS) -- two technologies that it unveiled to much fanfare a year ago -- continue to gradually improve thanks to both its own efforts and those of game developers. The company disclosed that a new DLSS image-processing algorithm that's based on an Nvidia AI research model can boost frame rates by up to 75% when playing action-adventure game Control, while cautioning that the algorithm for now produces artifacts that it's working to eliminate. Nvidia also discussed how Control is able to deliver lifelike imagery for scenes by simultaneously using multiple ray-tracing effects.
Ray-tracing and DLSS are enabled within games via dedicated processing cores -- they're respectively known as RT cores and Tensor cores -- that are found within many of Nvidia's Turing-architecture GPUs, which first launched last year. Ray-tracing, which historically was too computationally demanding to perform within games, can deliver photorealistic lighting, shadows and reflections for game scenes. DLSS improves game performance by using AI/deep learning algorithms to fill in image details, and by doing so helps make it possible to run demanding scenes relying on ray-tracing.
But as Nvidia, which in January partly blamed a sales warning on consumers needing more time to be sold on Turing's ray-tracing and DLSS capabilities, would itself admit, fully realizing ray-tracing and DLSS' potential isn't something that will happen overnight. Ray-tracing will have a bigger impact on game imagery as developers become more experienced working with the technology, and as Nvidia launches new GPUs that pack more RT cores. Likewise, DLSS' value will grow with greater developer experience and the launch of GPUs packing more Tensor cores, and as the trained AI models it relies on improve.
Research firm Gartner's view of where emerging technologies are within their hype cycles. Source: Gartner.
There are plenty of other examples of technologies that have been hyped in recent years for which the technology's evolution is very much a gradual process. Fully autonomous vehicles, which in the near-to-intermediate term might only be available via robotaxi services available in a limited number of geo-fenced locales, are a good case in point. Augmented reality headsets, which will likely have to address challenges related to issues such as display quality, battery life and software/services support before they're ready to take off, are another.
Research firm Gartner likes to argue that new technologies go through a hype cycle in which a "peak of inflated expectations" is followed by a "trough of disillusionment" as the technology fails to live up to initial expectations. And that afterwards, as the technology evolves and more would-be adopters appreciate its value, it goes through a "slope of enlightenment" before ultimately reaching a "plateau of productivity."
Though one can quibble with how relevant Gartner's model is to understanding one particular technology or another -- for example, some technologies might not have a big "peak of inflated expectations," or a particularly large "trough of disillusionment" -- there is some value to the broader framework it provides. Consumers, investors and journalists alike have a time-tested habit of expecting too much from innovative new technologies shortly after they've launched, and becoming too cynical about them following subsequent setbacks.
But for an investor who keeps expectations in check during the early stages of the hype cycle, and also doesn't get too pessimistic when a trough of disillusionment arrives, these shifts in public sentiment can present some pretty good opportunities.