ESP32 & BT 5.0 roadmap

[dperez]

Hi to everybody,
Regarding to the latest news about Bluetooth and its new version BT5.0 that has interesting new features compared to actual 4.2 version:
• 2x Speed
• 4x Range
• 8x Data

I would like to know a bit more about ESP32 roadmap.
 Is ESP32 going to support any of this new features?
 Will become somehow compatible (not fully) with BT5.0?

Can somebody add some information about this subject? Thanks in advance.

[madscientist_42]

Hi to everybody,
Regarding to the latest news about Bluetooth and its new version BT5.0 that has interesting new features compared to actual 4.2 version:
• 2x Speed
• 4x Range
• 8x Data

I would like to know a bit more about ESP32 roadmap.
 Is ESP32 going to support any of this new features?
 Will become somehow compatible (not fully) with BT5.0?

Can somebody add some information about this subject? Thanks in advance.

(Not associated with the Espressif team, but someone who's knowledgeable about many of the details of this question... This is the state of affairs, as observed by a professional doing embedded Classic and LE work in the industry. It's intended to state the reality for everyone.)

It's BT 5 and it's not expected that there'll be a 5.1, etc.- next release will be "6" and any new optional features will be specified in the same manner that EDR is (e.g. "+EDR") when it doesn't merit a new version number.

As for those features... There's tradeoffs. If you're getting four times the range without doing rather nasty RF tricks, you're losing speed because you have to increase your signal to noise ratios to get that range. Half of the current LE rates. (2x LE speeds and so forth... ) So, you're talking about a MIXED bag even if the silicon can handle it because it means a differing modulation technique much in the same sort and manner as EDR is to BR on Classic. It's quite a bit different from everything, with a handful of the specific standards from LE enforced on the "new" stuff.

It's not like you can wave a magic wand and have 5 happen, sadly. For Espressif, they're going to have to eat their investment on the CEVA 4.2 core licenses they have for the BT support they do have (Which isn't cheap...that's a good 7-8 figures (more likely solid 8 from my understandings of things...) deep "oops!"), license the BT 5 core IP's from CEVA (As they're the only player right now with license-able macros for ASIC use as of a couple of days ago...), wire the new and different IP macros into their SoC design, schedule engineering sample runs with TSMC, do bring up on the samples and all that this entails, followed up with scheduling production runs with TSMC and then packaging the same. ESP32 took about a year from the whisperings to the first dribs and drabs of production, year and a half to where they're now plentiful. It'll take as long for that change. Considering the costs and the margins they're running, IF there's an ESP32 sampling before end of next year with BT5 it'd be because someone handed them a major cash infusion to offset the aforementioned issues.