Luminus has a line of flat 2 mm² emitters with 3535 package, SFT-20.
So far they are all colour ones.
But I found several references to SFT-20-W i.e. white.
https://www.aperza.com/catalog/page/10007922/28848/
So…it may be coming.
I asked them about it but didn’t get any response.
Holy crap.
I wanna have it in the 95CRI config.
This would be truly awesome in 95cri.
Only thing it, unless someone makes a custom DTP star, you’re going to have a lot of fun with the offset LES…
They have special pads layout so they need special MCPCB.
No DTP but 4.54 mm² cathode…
This document is from 2016:
I suppose the cathode acts as the thermal path… Also then the PCB will have to be electrically isolated in a typical flashlight.
I’m wondering if you flowed this on a 5050 star with the cathode on the DTP pedestal and soldered the DTP and cathode connection on the star - would center the LES?
No. It’s just 0.17 mm off center. So even with stuff like a regular centring ring this should have a larger corona and smaller spot but the same cd as perfectly focused, assuming the reflector is quite precise.
Thanks a lot for that!
So it seems like a never-mass-produced rather than not-yet-announced product. 
Maybe in a smaller OP reflector. I’m not so sure I’d get a great beam in a larger, precision reflector, without some fidgeting. The specs call for the width dimension +/-.15mm so the emitter could be as much as .24mm off if the SMD package was symmetrical, and maybe up to .32mm if it wasn’t (or even just .02mm off… 
)
is this happening?
No
Kaidomain now also lists a WHITE SFT-20 LED:
And the specific MCPCBs:
Apparently the white could also handle up to 10A according to the listing.
Very interesting! It’s got a smaller die than SFT25R and rated for more current, which means it should be able to handle 40% more current density. I wonder how much of that translates to throw.
Not a fan of the proprietary footprint though.
This is exactly what stands out to me too!
It feels really weird that a LED (specificly the white variant) that was already spoken about in 2019, only now starts to pop up and potentially (by the metrics you mentioned) could perform ‘better’ than the ‘newer’/more recent SFT-25R.
It makes me really wonder why the monochrome/R,G,B variants were released so many years ago, while it took so much longer for this one to show up.
I tried my best, but wasn’t able to find an official datasheet for the SFT-20-W.
I wonder if this “new” LED also was already noticed by @Simon_Mao and that maybe he has already managed to get a hold of some sort of (preliminary) datasheet and has more information about it. (I know Simon probably won’t reply here on BLF, so I might try to ask about it via another platform or email.)
I absolutely agree! Fortunately Kaidomain also offers the correct MCPCB’s for them, so I’m thinking of just ordering a few, just to give them a try, although I would like to see them tested/being compared to comparable counterparts in an accurate/scientific way, to really being able to determine if they differ in performance compared to what’s available currently.
I do agree that it’s a strange situation. I’ve definitely seen discussions of SFT20 in the past, but don’t remember seeing a white variant. I would be very surprised if there is an 2019-era datasheet claiming the same 10A current–that’s more current density than the rated current density of the Osram W1, the throw king of that time.
But also, Luminus seems to have a track record of doing stuff completely ad hoc, like those mysterious SFT40 4000K batches or unreleased SFT70 high CRI datasheets. And unannounced changes in production process for the same emitter.
It would be extremely nice if we could get a sample to @koef3 to test, though I can imagine he may be swamped with work or with other projects.
I had already sent a DM/PM to him to ask if he was potentially interested and if he had the time/energy to test one or two, just before you suggested it here in this topic. 
Regardless of his answer, I’ve ordered 2 SFT-20-W from KD’s Aliexpress store. It is a bit more expensive than from their own website, but the few times I’ve ordered in the past from AE store, shipping time was quite reasonable at least.
The other thing is that I currently can’t add any items to my shopping cart on their own website; I’ve tried it on my laptop, my iPad and on my android phone and neither would work. Last week I had no issues with that.
I’m excited to hear your impressions of this emitter!
I’ve refrained from ordering on KD’s website because of some shipping situations I’ve heard about. Also, the site itself seems poorly maintained. Some features don’t work at times (as you’ve encountered), and the site is currently under attack.
Good to know that their aliexpress store has good shipping times. I order random stuff from aliexpress pretty regularly and am generally impressed with their shipping.
Those are for a large part the same reasons that I’ve also refrained from ordering from/via KD’s own website to this date.
There are 2 (BIG) downsides between ordering from their Aliexpress store compared to their own website:
Fortunately the SFT-20-W on their AE store were not extremely more expensive than on their own site, partially helped by the fact that I had an AE coupon that knocked an extra 2 bucks off, closing the gap a bit further.
The first downside is somewhat inevitable–the platform has to extract some profit somehow. The second is a deal-breaker because everything I’ve ever wanted to order on KD is not on their aliexpress shop. At this point I’m just using KD to monitor the state of the art, and hoping that other suppliers like Convoy catch on.
Yesterday I recieved the 2 SFT-20-W I ordered from KD’s Aliexpress store. Ordered on the 20th of may, recieved on the 2nd of june, so just within 2 weeks and that is including the 4-5 days it took KD to actually ship them out.
I took a few quick shots with my phone, so not the best close up quality, but for now it will hopefully give a quick impression:
And a quick test to see if they did light up and confirm the white color:
The most noticeable is that the LES isn’t exactly centered on the package, which might give some focus/centering issues.
