In mid-2014, Dell and Intel embarked on an ambitious, long-view product ecosystem known internally as Dell Stack. The project centers on a 6.4-inch, full-HD mini-tablet that would underlie the desktop, laptop, and tablet experiences: That is, one device would power your entire computing ecosystem.

The goal was to significantly streamline the transition that the average business user — and, eventually, consumer — makes between their multiple computing devices numerous times per day. Instead of constantly moving data and files from device to device (or device to cloud to device), Stack would allow users to transition the core handheld from screen to screen as use case dictated.

All of this hinged upon Continuum, a technology baked into Microsoft’s Windows 10. Since software written to be compliant with Continuum, called Universal Windows Platform (UWP) apps, will run equally well under Windows 10 or Windows 10 Mobile, it opens up a range of possibilities with respect to the very kind of use-case transitions that Stack attempts to address. That is, Continuum makes those possibilities a reality.

Flexible specs

With the underlying objective being a full computing ecosystem replacement, the typical ARM-compliant processors utilized in the vast majority of today’s tablets and smartphones would not be sufficient. Therefore, the Stack handheld was spec’ed out with an x86-based CPU from Intel’s Kaby Lake Y-series of low-power-consumption, dual-core, laptop-class processors. Specifically, the built-to-order system would be available with options from the m3, m5 vPro, or m7 vPro families.

As a tablet, the handheld would require a 3.5-watt power draw, but when docked as the heart of a desktop-replacement configuration, it would transition to a less-conservative 12 watts. Furthermore, it would be available in several memory configurations: either 4GB or 8GB of RAM, and 128GB or 256GB of internal solid state storage (supplemented by a microSD removable storage slot). Its main and front-facing cameras were initially envisioned as eight and five megapixels, respectively, with the latter to be paired with a biometric-enabling iris scanner.

The entire device was targeted to be under nine millimeters in thickness, putting it in the same league as many smartphones and tablets. Furthermore, a second generation was also on the drawing board, with an even more ambitious configuration: It would be a six-inch phablet with full telephony capabilities — a true x86 smartphone.

At some point, however, the handheld morphed from an ultra-thin 6.4-inch objet d’art to a seven-inch tablet which more closely resembles a modern Dell product. This was likely done in deference to the reality that, since it could not really be considered pocketable as a 6.4-inch mini-tab, it might as well be expanded a bit to improve its thermal properties.

Where is it now?

The first deliverables were not even targeted to reach the market until the spring of 2017, according to several documents reviewed by VentureBeat, but the current state of the project is unknown. It may have been shelved alongside Intel’s mobile ambitions when the chipmaker cancelled its Atom line of ultra-low-voltage chipsets, but since Stack envisioned utilizing laptop-class Kaby Lake silicon from the outset, it may have survived the strategy shift.

It’s also possible that Dell went into a holding pattern when HP announced its Elite X3, a Continuum-enabled Windows 10 Mobile phone, earlier this year. Although that handset employs more traditional mobile CPU architecture in the form of Qualcomm’s Snapdragon line, its use-case scenarios are strikingly similar.

One thing is quite clear: Based on the Twitter reaction to an early render of the handheld (also seen as the top image of this piece, next to a “dumb” tablet), along with its x86 pedigree, there is an incredible amount of interest in a project like this. And if Dell is not currently pursuing it, there’s a good chance that one of its rivals — or Microsoft itself, even — is. On paper, Dell has a winner here, but in the end execution of the shipped product(s) is what matters.