Отфильтровал на боте 
(отсюда) статистику батчей по Devils Canyon 4790K / 4690K и Pentium K с комментариями владельцев. 
Может кому пригодится...
Batch 4790K	Comments
L329C241 	5.0ghz 32m 1.325v 4c/8t - Mikecdm - mild scaling, -140 6.2 benchable / 6.5 validate
L331C501	4900, 4c/8t, 1,35v, SP 32M
L331C501 	5Ghz 4C/4T @1.317  Corsair H110
L331C516	5 ghz water 1.34 4/8 ht 5.7 ghz Dry Ice 1.65 4/8 ht "
L331C518 	4.7ghz wprime 1.2v 4c/8t superV - no scaling above 1.2v!
L336D105	5GHz 4/4 @ 1,250V SP32M @ water
L336D106 	5.0ghz 32m 1.35v 4c/4t - barti27
L336D106 	5.0 1.35v
L352C118 	5.0ghz 32m 1.3v 4/4t - dumo
L352C119	4.7ghz cinebench 1.2v 4c/8t - Gunslinger - no cold scaling
L352C120	fast test @ss (-20 evap with 200w load)
                Wprime1024 5ghz 4/8 - 1.200 in vista 32bit
                HyperPi32m 5ghz 4/8 - 1.222 in vista 32bit with bbse @2400 8-11-7-24
                Maximus VII hero 0401
L418C133	5ghz 1.238v possible...but with hard mems and 4.9ghz cache pretty hard...so 1.245 here.
                5111 with 1.297v
                both on air.
                ln2 coming weekend "
L418C133 	4.8ghz wprime 1.31v 4c/8t - splmann
L418C134	WC : Wprime 32M 5Ghz 1.31v (1.32v) 
                LN2 : super Pi1M 6.4GHz 1.65v (1.72v) // Vccin2.2
                LN2 : Wprime32M 6.2GHz 1.6v (1.66v) // Vccin 2.2
                Cache : superpi 1M 5.8GHz 1.45v (1.5v)
                Vid : forgot to check
                CB/CBB : -130. 
                Good cpu, easy to bench. Didn't test with lower Vccin. I tried 2.4v one time, but no imporvement ( step is too big 6.4 ok , 6.5 not ok)
L418C134	6500 MHz max valid 2c/2t 
http://valid.canardpc.com/hbpt3u 
                cb -120, 5 Ghz 32m air ca 1.33 volt IIRC "
L418C134	6200 GHz, 4c, 1,70v, 3DMark03
L418C134 	5.0ghz 32m 1.358v 4c/8t - zzolio
L418C134 	5G 32M 1.37 4/8 same for hwbot prime
L418C164 	5.0ghz 32m 1.25v 4c/8t - steponz - no cold scaling -105 CB
L418C169 	5.0ghz 32m 1.300v 4c/8t - Mikecdm - no cold scaling -105 CB
L418C213 	5.0ghz 32m 1.35v 4/8t - SHIMIZU - no cold scaling 
L418C213(2L)	VID 1.056V - 5GHz 32m 1.3V 4/4 - no cold scaling -106 CB
L418C213(2L) 	VID 1.056V - 5GHz 32m 1.325V 4/4 - no cold scaling -118 CB
L418C221 	5.0ghz 32m 1.4 v 4c/8t - NameGT - good cold scaling 6.7 validation.
L418C222	no document but assumed cold scaling
L418C223	5.0ghz 32m 1.35v 4c/8t - LoudSilence - good cold scaling! 6.3ghz 32m at 1.7v. This gives hope for 6.5ghz with a 1.25v chip
L418C230 	5ghz 1.35v boot win7 4c/4t - slamms - okay scaling 6.2ghz at -115 (bad cb)
L419B531	Super Pi 32M
                5ghz 2c 1.25v
                5150 Mhz 2c 1.4v
                5300 Mhz 2c 1.53v (30ºidle - 60º peak full)
                4600/4700 Mhz uncore 1.3v
                vid 1.1v
                vccin 1.6v
                very strong IMC, 1378 Mhz 8-12-9 1T with GSkill Pi PSC 1.88v
                all on air "
L419B533 	5.0ghz 32m 1.272v 4c/4t - SniperOZ - cold scaling! -130 6.4ghz 1.93v. not the best scaling but not bad! it likes volts
L419B535 	air - 4.7GHz XTU at 1.25V - 4.8GHz XTU/WPrime 1024 at 1.32V
L419B538	air - 4.7GHz XTU at 1.25V - 4.8GHz XTU/WPrime 1024 at 1.32V
L419B540	4.7ghz 32m 1.23v 4/c8t - stasio
L419B593	1. 5ghz 1.3v 32m, 6.2ghz 32m - decent imc
                2. 5ghz boot 1.4v, - terrible imc
                3. 4.8 max 1.32v, untested cold - also bad imc 
L419B593	5Ghz 4c8T 32m 1.36v water,imc seems really good
                Will test imc more and scaling on Ln2 next week 
L419B640	good for air bad for ln2 6.25 1m.
                vid 1.058
                4.8ghz - 1.268v - corsair H100i - white cheap thermal paste - auto fans
L419B722        best can't boot 5 GHz air
L420B218 & 220	ive tested a few L420B218 & 220 
                5ghz 1.28v no problems, but didnt scale with cold. max 6.2ghz 
L420B758	WC : Wprime 32M 5GHz 1.32v (1.33v) 
                LN2 : super Pi1M 6.4GHz 1.66v (1.73v) // Vccin2.2
                LN2 : Wprime32M 6.2GHz 1.6v (1.66v) // Vccin 2.2
                Cache : superpi 1M 5.7GHz 1.45v (1.5v)
                Vid : 1.05
                CB/CBB : -130. 
                Good cpu, easy to bench. Didn't test with lower or higher Vccin.
L420B758	WC : Wprime 32M 5GHz 1.3v (1.31v)
                LN2 : super Pi1M 6.3+ GHz 1.65v (1.72v) // Vccin2.2
                LN2 : Wprime32 6.2GHz 1.6v (1.65v) // Vccin 2.4
                LN2 : Wprime1024 6.1Ghz 1.6 (1.65v) // Vccin 2.4
                Cache : superpi 1M 5.7GHz 1.45v (1.5v)
                Vid : 0.99
                CB/CBB : -135. 
                Good cpu, hot chip (WC and LN2)."
L420B765 	all cpus boot 5 GHz air 4/8 so are only lemons, not tragic on air (best is 1.23v @ 4.7 GHz OCCT), but not scaling on cold, best can do 5.8 GHz Physx 3d11 but 6.1 GHz 
                CPU-Z is max with 2 cores...
L420B765 	stock vid 1.08v and turbo was only 4.2 ghz@1.13v, i enhanced turbo and goes 4.4ghz@1.2v
                wprime 2.10 /c states disabled/ only vcore and /vccin 2.0v
                4.5Ghz@1.19v 1024m temps max 61/61/63/58
                4.6Ghz@1.19v 1024m temps max 62/61/64/59
                4.7Ghz@1.23v 1024m temps max 67/67/69/64
                4.8Ghz@1.28v 1024m temps max 70/72/75/66
                4.9ghz@1.42v crashes around 30% with temps on 92 on highest core 
L420B770	4.5ghz@1.12v wprime 1024M temps max 55/52/55/50
                4.6ghz@1.12v wprime 1024m temps max 56/53/55/51
                4.7ghz@1.14v wprime 1024m temps max 58/56/58/52
                4.8ghz@1.18v wprime 1024m temps max 62/60/62/57
                4.9ghz@1.25v wprime 1024m temps max 71/69/71/64 with errors"
L420B772	5ghz air 1.25v 32m stable LN2 incoming. 
L420B831	does only 4.8@1.35 wprime 1024m on water...exact like my 2nd.
                bios vid at stock 1.064, stock under load cpuz shows 1.216v.
L420B831 	6.5ghz 32m ( i think -130 cb)
L420B832	1.3v 5ghz air 32m
                6.3 GHz max -120c
                Good IMC 2933 "
L421B985	1.32v 5ghz 32m = 6.2 1m and 6.5 validate. -105. 1.7v. Im sure if it was -130 cb we would see nicer results.
L421B985	1.42v 5ghz 32m. No ln2 left for this one.
L421B985:	
A3959 - 5ghz 32m 1.4v
A3820 - 5ghz 32m 1.32v
A3669 - 5ghz 32m 1.3v
A3629 - 5ghz 32m 1.34v
A3559 - 5ghz 32m 1.37v"
L421B985 	A3624 is 1.27v 5ghz 4c 4t 32m XP. 1.25v cpu did 6.4ghz 1m at -138 (totally mental CB)
L421B985 	A3624 is 1.27v 5ghz 4c 4t 32m XP. LN2 tonight on best 2. 1.27v cpu did 6.3ghz 1m at -130
L421B985 	A3555 is 1.35v 5ghhz 4c 4t 32m XP. 1.35v cpu didnt test cause dont care 
L421C111	6230 pifast 5616 XTU CB 125°C 
L422B687	6.4ghz 32m (think -130 cb)
L425B370 A2127	neither of those CPUs load OS with 4c/4t at 5000MHz (Tried until 1.4V)
L425B405 A1127	neither of those CPUs load OS with 4c/4t at 5000MHz (Tried until 1.4V)
	
	
Batch 4690K	Comments
L331C514 	epic potatoes, no boot to OS 5 ghz
L420B903 A2183	neither of those CPUs load OS with 4c/4t at 5000MHz (Tried until 1.4V)
L421C035 A0431	neither of those CPUs load OS with 4c/4t at 5000MHz (Tried until 1.4V)
L421C035 A1204	neither of those CPUs load OS with 4c/4t at 5000MHz (Tried until 1.4V)
	
	
Batch G3258	Comments
3418C000 	5 x hasfail with 1.3v on Z87M OCF H100 x 4.6Ghz, 2 x 4.7Ghz, 1 x 4.8Ghz, 1 x 4.9Ghz that boots 5Ghz with 1,4"
314C011	        6.2ghz max valid, strong imc over 4 GHz 
http://hwbot.org/sub....ram_2002.5_mhz
                5 GHz 32m air around 1.43 volt
3418B976	1.4v 4.8ghz boot but unstable in os
3418B977	Was able to do 6.1GHz Pi 1M but coldbug was only -105'C
3418B977 	best is 1.55v @ 5 GHz boot to OS
3418B977 x 4	Best 4.9 @1.4 4.7 @1.3 Worst 4.7@1.4 4.6@1.3 Other two 4.8 / 4.6 
3418B983 	4.6GHz@1.35 before beeedos
3418B983 	VID 1.061v, 4.9ghz 1.35v wprime32m
3418B986 	VID 1.056 --- Wprime 32 5.0Ghz 1.33v --- 5ghz Boot 1.33v
3418B986 	VID 1.032 --- Wprime 32 5.0Ghz 1.32v
3418B987	1.32v 5ghz air 32m and hwbot prime 65c load
3418B987 	5GHz@1.33 32m wprime ok
3418B993 	VID 1.024 --- WPrime 32 4.8ghz 1.35v --- 5Ghz Boot over 1.4v
3418B993 	VID 1.040 --- WPrime 32 4.8ghz 1.35v --- 5Ghz Boot over 1.4v
3418B993 	VID 1.024 --- WPrime 32 4.8ghz 1.35v --- 5Ghz Boot over 1.4v
3418B993 	VID 1.032 --- WPrime 32 4.8ghz 1.30v
3418B993 	Vid 1.056 4.5 Ghz 1.2v 4.7Ghz 1.25v 4.8Ghz 1.3v 4.9Ghz 1.36v 5.0ghz 1.425v
3418C001	5200, 1,5v
3418C001	VID 1.002v SuperPi 32M 5GHz @ 1.27v HWBot Prime 6.3GHz Liquid Nitrogen - Coldbug -135'C 
3418C008	#1 4500 32M 1,2v, not less tested
                #2 4500 wprime 1024 1,176v minimized"
3418C011	CB15 4500MHz at 1.17Vcore, so pretty great daily CPU, on air goes up to 4700MHz 1.33Vcore... 4800MHz doesn't post at all, not even with 1.5Vcore lol...
                CB15 5800MHz at 1.7Vcore at -80°C
3418C011 	on air needs 1.45Vcore and +2V VCCIN for a valditaon for 4.8 but at stock voltage its 3.9 stable
3418C011 	VID 1.012 Air 4.5G 32M @ 1.25V 5G 32M @ 1.4V SS 5.5G 32M @ 1.6V -45C"
3418C013	Stock cooled Cinebench R15 @4.5Ghz 1.272 volt
3418C015	1.025v/1.39v 32M @5.0
3418C015 	sucks.. Tried 5 chips and there all over the place, Definitely not as strong as the B98*.
3418C015 	VID 1.044 --- WPrime 32 4.9ghz 1.30v
3418C017 A1017	6455.78_mhz, 1,96v, 1c 
http://hwbot.org/sub....58_6455.78_mhz
3418C019 A3099	6330Mhz, -> LN2 at -120/-125° with 1.9/1.95v real vcore
3419B303	need to test more but seems no 5 GHz air and at least 1.3-1.35 volt for 4,6 GHz cinebench"
3420B626 	4.2GHz@1.35 stable, higher vcore yielded nothing
3420B753	"bunnyextraction chip, needs 1.45V to run cinebench r11.5 @ 4.7ghz on air (prolimatech armageddon). 
3420b754	1.064v in hwmonitor and cpu-z for default
L417C016	6017ghz @ -116c, 1.648vcore / 2.4vin : IMC tested to 5.7ghz, 2850 c9 tight
L418C276	"1.134 VID, if I'm not wrong 4.5GHz 1.27V 32M 5GHz 1.5V no boot, 1.55 can't load XP.5.7GHz 1.8V Cinebench R15, CB -135. 5.8 boots to windows, 
                but no stability with CB15 - 0x124 BSOD.      Tried higher and lower vcore and vccin.