agreed. ill take a 800-900 (hopeful) lumen K1 with > 500kcd range and 5000k tint. sign me up. 
plus, 3 amps sounds to me like good battery life?
yup! or 4 of them in a KR4 or D4. I just sold my beautiful brass d4v2 with W2 because i couldnāt justify the 6500k tint any moreā¦ Itād be awesome to have similar range in a 5000k tint!
Even with the highest bin, 800lm at 3A wonāt give much more than 900lm at the peak of about 4-4.5A
Still quite short of the 1000-1100 of the CULNM1
CULNM1?
90% of cslnmn and 5k is good for me. 3A makes driver options pretty easy and cheap too.
Yeah sorry, that one 
Seems like the XP-P is āhereā. As in itās on Creeās website. XLampĀ® XP/XT LEDs - Cree LED
1.05*1.05 mm die though, so itās going to need some miracles to compete with the white flats on brightness alone. Likely lower cost, better availability, 5000K cct, and 3535 footprint are a big win though. I like how the surrounding silicone pour is below the plane of the phosphor. This should make for cleaner beams / less artifacts. 3A vs 3A, it should hold itās own. Looks to be not much room for overdriving. Letās hope weāre wrong about that somehow.
It has a pretty high forward voltage. The 219b/sst-20/LH351D have a forward voltage under 3.3V at 3A. The Xpp has a forward voltage of over 3.6v at 3A on the datasheet. This should be a fet capable emitter but I donāt see it going much over 800 Lumens even when overdriven.
As soon as one is available at reasonable cost Iāll test it and post results. Latest datasheet only shows U5 bin (360lm) available but in all CCTs. Hopefully 5000K rolls out soon and isnāt puke green.
Edit: The Vf looks to be nearly identical to the Osram Black Flat S (HHL532.TK). Running DD from an LG H26 I see a max of about 4.95A in a super low resistance build. So these should run nicely off of 3500mAh 18650s and 5Ah 21700s
Oh, weāll throw the strongest cell we have at them and see if we can turn them blue.
I got 3 XP-P emitters in the mail today. Will post test results soon, but initial eyeballing says that our predictions are right. Not much headroom here. 3A lux reading was 1350, peak was 1500 around 4.25A (roughly). I ordered the 5700K (lowest in stock). Tint is purplish from the bare emitter. Check back laterā¦
Yeah Iām more excited about the XHP50.? HI version.
The Vf on this thing is crazy high. Hereās a test comparing to all the Osrams that I have.
Put one in my Jacob A60. Beam and color are very, very nice. Will get kcd numbers shortly and edit this post. XPP yielded 224kcd. CSLNM1 (MTNe) did 233kcd, HHL532.TK (mouser) did 278kcd. Guessing this will slot in between, but nearer the NM1. For having 10% larger LES, this is a nice number.
Thanks for testing.
Difficult to say what this led adds to the CSLNM1, it may be a bit more efficient at lower currents, but less than seems because of the higher Vf. But if it comes in a range of attractive tints that could convince me to use it.
I think it wouldāve been a hit about a year or two ago before 3030 and 4040 mcpcbs and gaskets were widely available and when DD FET drivers were king. Nowadays itās not really doing anything special. If the 5000K comes in a bin or two higher flux, then itās a different story.
Measured that A60ā¦ got 224kcd. So not quite hitting the NM1, but close enough. I feel like the hotspot is noticeably larger, but I canāt be sure since I only have one of these lights.
Could be useful for lights that have a weird reflector hole size with a non standard centering ring. Or a multi emitter light like a convoy S12 that use a non standard MCPCB that doesnāt come in 3030. I know itās possible to put osrams on a 3535 but It seems like a pain.
Thanks for testing JaredM. This emitter is exciting to me. I, just thinking multi emitter light like a Noctigon KR4 or even a 7 emitter or 18 emitter light, in 5000k with throw 90% of a cslnm1? Amazing.
now, the real interest, when are xhp50 HI coming out, and is there a data sheet?
Is the only way to subscribe to a thread in blf, is to post?
For myself, I am not interested in high currents. Looking at relative flux, it looks like around 144 ish at 350 ma. Might be good with 1 watt, single amc7135 small lights and small optics.
The rest of this post is my personal use for wearable lighting, daily use, inspection etc, 10 hour work daysj, with crews and family (hoping this emitter might crack an old problem): I typically look for 4000 lux at 1 meter for as close to 10 hours (grudgingly, I will take 8 hours) flat output. The 2011 4aa Fenix hp 10, could do this with a 139 lpw xpg1, some good heat sinking, a 1 watt ish cc buck driver, ar lens, and a 26 mm aluminum reflector. it would be nice to see this same performance in a single 18650 and a 20 mm lens. So far, all I see is 6 amc7135 and pwm down to useless efficiency and runtimes, made to impress with the lumen numbers, rather than efficiency numbers. Leds are about efficiency and what we can do with the limitations we place on ourselves for practical work a day applications. Not one off experiments.
This is why I make my own lights and have lost interest in just about anything sold today. I need lights that meet specs, since I loan them out to people who do not, and should not, be changing batteries, but need 10 or even 20 hours at 4000 lux at one meter (1000 lux at 2 meters). I am pretty happy with 2s18650 buck efficiency, but I see a big difference going down to linear drivers. However, it is 2021 and the problems remains: The smaller the light, the fewer the cells, the smaller the optic, the less intensity per ma, the less juice can be fed to the led if you need runtime, the less efficient the driverā¦ I struggle to see with my smaller lights. The best was the imalent hp 10, which was an xplhi and dial, I bought 3. They offered best intensity and runtime, but only offered me barely 4 hours of useful light (barely enough time to finish a day after sundown, but maybe not if used in morning during loading time) which is about 2 hours more than convoys, etc. The only reason I bought them is that I can wear them on the belt when not using them on my head. I canāt do this with 2p18650 lights, which means I must quit what I am doing to find those lights, if I brought them to the job.
Subscribe button at the bottom of the 1st post.
Thatās 4000 candela (4kcd) or 126m throw.
The ThruNite Catapult Mini does 21 lm on low for 12 hours with a 18350 cell. Itās about 200 cd/lm, so thatās over 4000cd. 40mm head though, which is probably not what you want.
The 27mm bezel Convoy S21B with W2 LED is 9000cd on the 10% mode, which should last about 7 hours. There isnāt a 5% mode unfortunately but that almost sounds like exactly what you want. A smaller 18650 host with W1 might even do it too.