Links have been made between the t214 "Mariko" chip mentioned in Switch Firmwares back in March 2018 and the new chip in the upcoming Nvidia Shield, suggesting they are the same chip. (TL/DR is at the bottom)
Highlights from article: https://www.eurogamer.net/articles/digitalfoundry-2019-switch-new-tegra-x1-silicon-comes-into-focus
"The Switch hardware upgrade story starts last year - specifically around March 2018 - when Nintendo released the 5.0 version of its system software, known internally as Horizon. Support for the standard 't210' version of the Tegra X1, codenamed Logan, was joined by a new, hitherto unknown revision: t214, also referenced as Mariko. Nvidia's Tegra codenames are based on the real names of superheroes (Parker being Tegra X2) but Mariko is something different. Back in the day, she was a love interest for Logan/Wolverine in the Marvel comics - the obvious inference being that this is a partner chip, not a new product."
"The Google Play Developer Console Device Catalogue is listing a new Shield, based on a t210b01 chip. So what's the connection between this and the t214 Mariko? For answers, we looked to ResetEra's Thraktor, whose short post managed to link the dots between the two chips, with the evidence all but confirming that both processors are actually one and the same. Github commits like this one and this one both replace t214 with t210b01 with the commit message, "Should use t210b01 in any code/comment". There are other smoking guns linking the two processors too - here, here and here, for example - where commit messages refer to t214 while the code or filenames reference t210b01 instead. Another piece of evidence Thraktor points out to us is that Nvidia uses the MAX77812 voltage regulator for the new t210b01, and the exact same component is used for t214 within the Switch firmware."
"DVFS tables for the t210b01 are available, and they can be compared to the standard t210. Operating voltages are indeed reduced, and while CPU and GPU frequencies supported by t210 are retained, faster clocks are also available. The GPU limit of the Tegra X1 - specced at 1GHz but with a max 921MHz on Switch - is increased to 1.267GHz on the new processor. It's worth noting that the frequency data for the new processor is rather old and may reference an incomplete engineering sample chip, and things may be different in the final production chip"
"Limiting clocks improves battery life and reduces heat, putting less strain on the active cooling assembly within the hybrid console. Regardless, the option is there for Nintendo to use the presumed process advantages not just for improved efficiency and battery life, but performance too. It's important to note is that the evidence does suggest that t210b01 is fully compatible with the original Tegra X1 - the DVFS table for the new chip lists all of the clock speed modes available, not just the maximums, and the modes used in existing Switch games are all in the line-up"
"standard Switch performance could likely be achieved with no cooling assembly required at all, meaning that a prospective Switch mini wouldn't just be smaller with more battery life, it could be silent too. Meanwhile, a decent performance uplift on the GPU side could obviously improve frame-rates and image quality in a range of games using dynamic resolution scaling. Based on the clocks in Nvidia's documents (though again, this may be outdated info) there wouldn't be a revelatory increase in system performance - nothing like a generational leap as such, or even anything as profound as the jump from 3DS to New 3DS"
"In short, a refresh at this point allows the Switch to potentially produce more performance, less heat, extending battery life while being cheaper to manufacture."
TLDR
Tegra X1 revision in the new shield console has been linked to a chip first referenced in Switch Firmware 5.0 (Mariko).
New chip could support GPU clocks of up to 1.276ghz compared to 1ghz of the current. (1ghz chip used at 921mhz on Switch).
New chip would have improved efficiency at same performance as old chip meaning higher battery life and could potentially be cooled without the need of a fan, meaning a silent switch.
If used for a performance increase it would be a minor improvement.
With the new chip the Switch would be cheaper to produce, have better battery life, be quieter and potentially have a modest increase in power.
Personally I'm not sure increased power will be leveraged with the new chip, and I'm ok with that. Longer Battery life and a fanless switch would be great.
