| 00:14.03 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 00:17.05 | *** join/#arm-netbook stefanro (~stefan@pD9FF9AF0.dip0.t-ipconnect.de) |
| 00:46.43 | *** join/#arm-netbook hramrach__ (~hramrach@gateway/tor-sasl/hramrach) |
| 01:27.35 | *** join/#arm-netbook DEAT (~heh@14-201-99-152.tpgi.com.au) |
| 01:30.26 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 01:38.12 | *** join/#arm-netbook discopig (~discopig@2001:5c0:1400:a::911) |
| 01:38.12 | *** join/#arm-netbook discopig (~discopig@unaffiliated/discopig) |
| 02:30.29 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 03:26.17 | *** join/#arm-netbook tuliom_ (~tuliom@177.157.253.252) |
| 03:27.58 | *** join/#arm-netbook Dessimat0r (~Dessimat0@cpc1-birk8-2-0-cust884.1-3.cable.virginmedia.com) |
| 03:30.48 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 03:40.29 | *** join/#arm-netbook winocm (~rms@opensn0w/developer/winocm) |
| 04:30.52 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 05:30.10 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 05:46.32 | *** join/#arm-netbook Quarx (~Quarx@46.233.205.180) |
| 06:05.29 | *** join/#arm-netbook herdingcat (~huli@221.208.13.214) |
| 06:12.54 | *** join/#arm-netbook herdingcat (~huli@221.208.13.214) |
| 06:30.37 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 06:39.55 | *** join/#arm-netbook hp__ (~kvirc@d5153E228.access.telenet.be) |
| 06:43.24 | *** join/#arm-netbook newbie1 (~kvirc@d5153E228.access.telenet.be) |
| 06:48.39 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
| 07:12.35 | *** join/#arm-netbook newbie (~kvirc@d5153E228.access.telenet.be) |
| 07:13.14 | *** join/#arm-netbook DEAT_ (~heh@14-201-99-152.tpgi.com.au) |
| 07:13.45 | *** join/#arm-netbook Quarx|2 (~Quarx@46.233.205.180) |
| 07:15.57 | *** join/#arm-netbook lerc_ (~quassel@121-74-233-23.telstraclear.net) |
| 07:17.31 | *** join/#arm-netbook hark_ (~hark@yavin.puscii.nl) |
| 07:18.08 | *** join/#arm-netbook wingrime (~wingrime@78.109.115.103) |
| 07:18.11 | *** join/#arm-netbook ausxxh_ (~xxiao@li41-126.members.linode.com) |
| 07:18.21 | *** join/#arm-netbook arete74_ (~arete74@net-93-64-241-175.cust.dsl.vodafone.it) |
| 07:18.38 | *** join/#arm-netbook nashpa_ (~nashpa@dliviu.plus.com) |
| 07:19.36 | *** join/#arm-netbook CaCtus491 (~CaCtus491@eth59-167-75-233.static.internode.on.net) |
| 07:20.05 | *** join/#arm-netbook Kebianizao|work (~Kebianiza@1.37.219.87.dynamic.jazztel.es) |
| 07:21.29 | *** join/#arm-netbook pcat (~cat@gprs-internet-bceea6-96.dhcp.inet.fi) |
| 07:30.42 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 08:04.57 | *** join/#arm-netbook rellla (~rellla@p4FE56742.dip0.t-ipconnect.de) |
| 08:27.59 | *** join/#arm-netbook rz2k (~rz2k@37-145-3-213.broadband.corbina.ru) |
| 08:30.55 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 08:59.04 | *** join/#arm-netbook Dessimat0r (~Dessimat0@cpc1-birk8-2-0-cust884.1-3.cable.virginmedia.com) |
| 09:23.15 | *** join/#arm-netbook madmalkav (~Hamlet@188.84.32.206) |
| 09:25.00 | *** join/#arm-netbook cnxsoft (~cnxsoft@node-4pz.pool-118-172.dynamic.totbb.net) |
| 09:28.36 | *** join/#arm-netbook roric_ (~roric@c-c47ce455.213-3-64736c14.cust.bredbandsbolaget.se) |
| 09:28.51 | *** join/#arm-netbook wingrime_ (~wingrime@78.109.115.103) |
| 09:30.56 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 09:48.34 | *** join/#arm-netbook stamina (~stamina@140-074-128-083.dynamic.caiway.nl) |
| 10:30.40 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 10:40.28 | *** join/#arm-netbook penguin42 (~dg@tu006.demon.co.uk) |
| 11:30.18 | *** join/#arm-netbook tinti_ (~tinti@pdpc/supporter/student/tinti) |
| 11:30.26 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 11:31.55 | *** join/#arm-netbook herdingcat (~huli@218.10.62.222) |
| 11:49.34 | *** join/#arm-netbook tuliom (~tuliom@177.157.253.252) |
| 11:56.12 | mnemoc | hi, do you happen to know any decent chinese "galaxy mini" (clone... size-wise) |
| 11:58.49 | mnemoc | to replace my 9yo's phone without pissing off my 12yo's :p |
| 12:01.53 | *** join/#arm-netbook datagutt_ (~datagutt@unaffiliated/datagutt) |
| 12:07.45 | *** join/#arm-netbook leowt (~leowt@a89-153-146-62.cpe.netcabo.pt) |
| 12:30.34 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 12:30.55 | *** join/#arm-netbook gimli (~gimli@xbmc/staff/gimli) |
| 12:38.15 | *** join/#arm-netbook Dessimat0r (~Dessimat0@94.116.205.161) |
| 12:40.08 | *** join/#arm-netbook eFfeM (~frans@c73189.upc-c.chello.nl) |
| 12:41.06 | *** join/#arm-netbook leviathanch (~quassel@c-82-192-226-27.customer.ggaweb.ch) |
| 12:42.49 | *** join/#arm-netbook herdingcat (~huli@218.10.62.222) |
| 12:52.32 | *** join/#arm-netbook datagutt__ (quassel@2a01:7e00::f03c:91ff:fe96:9323) |
| 12:54.00 | *** join/#arm-netbook datagutt__ (quassel@2a01:7e00::f03c:91ff:fe96:9323) |
| 12:56.25 | *** join/#arm-netbook datagutt__ (quassel@2a01:7e00::f03c:91ff:fe96:9323) |
| 13:30.49 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 13:43.24 | *** join/#arm-netbook ksdncv (~ksdncv@unaffiliated/ksdncv) |
| 14:30.45 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 15:29.55 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 15:38.04 | *** join/#arm-netbook wingrime (~wingrime@78.109.115.103) |
| 15:58.33 | *** join/#arm-netbook herdingcat (~huli@218.10.62.222) |
| 16:30.52 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 17:19.14 | *** join/#arm-netbook cjoe (~customerj@fibhost-66-7-177.fibernet.hu) |
| 17:31.00 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 17:36.56 | *** join/#arm-netbook penguin42 (~dg@tu006.demon.co.uk) |
| 18:31.33 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 19:19.22 | buZz | anyone still looking for a nice 5V supply? |
| 19:19.23 | buZz | http://www.ebay.nl/itm/5V-4A-20W-Single-Output-Switching-Power-Supply-/280767420997 |
| 19:22.41 | mnemoc | that's cool if you'll put your board inside a cabinet |
| 19:22.41 | mnemoc | for a bench .... it's pretty ugly :p |
| 19:24.04 | buZz | mnemoc: i will stuff it inside http://www.thingiverse.com/make:40907 |
| 19:24.06 | mnemoc | iirc i've seen 12A variants too |
| 19:24.34 | buZz | probably make another section to the case, to hold the PSU, a C13 plug, maybe add some female usb just for charging |
| 19:24.56 | buZz | ok, 120W sounds WAY too much to power a cubieboard ;) |
| 19:31.39 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 20:01.34 | *** join/#arm-netbook cjoe (~customerj@fibhost-66-7-177.fibernet.hu) |
| 20:20.52 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 20:27.58 | penguin42 | http://ssvb.github.io/2013/06/27/fullhd-x11-desktop-performance-of-the-allwinner-a10.html is facinating - I don't get why backward memset has a lower problem with screen bandwidth |
| 20:30.10 | ssvb | penguin42: one guy here asked the same question - https://plus.google.com/u/0/105007947798310229700/posts/1MExqLRm8cF |
| 20:30.52 | ssvb | does not know the right answer and can only guess... |
| 20:31.06 | penguin42 | ssvb: Odd that - it's me :-) |
| 20:31.15 | ssvb | ah, ok :) |
| 20:31.17 | penguin42 | but hmph G+ hasn't told me about the response |
| 20:32.55 | penguin42 | ok, that's fascinating - G+ still doesn't tell me about the response - it's not in my notification thing - hmph! |
| 20:33.26 | penguin42 | ssvb: I agree it might be interference - something like if both the raster and the memset are going forward they might conflict multiple times |
| 20:36.25 | penguin42 | ssvb: hmm so you're doing 1MB fills, so OK that's way bigger than the caches - which memsets did you use for those figures? I see you have some NEON ones |
| 20:38.51 | ssvb | penguin42: the test is using 32MB buffers, that should be way bigger than the CPU caches (except maybe on hi-end x86 servers), all types of memset show the same behavior |
| 20:39.42 | penguin42 | ssvb: ok, but it makes an interesting question of whether it's worth flipping the ARM memsets to go backwards |
| 20:40.41 | ssvb | penguin42: that would be slower on Exynos5 (Cortex-A15) - https://github.com/ssvb/tinymembench/wiki/Samsung-Chromebook-XE303C12-%28Exynos5250%29 |
| 20:41.13 | ssvb | 'NEON fill backwards' vs. 'NEON fill' and 'standard memset' |
| 20:41.53 | penguin42 | wow yeh it stinks backwards doesn't it |
| 20:42.36 | ssvb | ARM hardware is weird and sometimes needs some tricky gymnastics to make it run fast :) |
| 20:42.38 | penguin42 | ssvb: It's interesting all the cores (and sometimes individual chips) are noticeably different |
| 20:42.51 | penguin42 | ssvb: Yeh I spent about a year doing optimised memcpy/memchr etc |
| 20:44.10 | penguin42 | my memchr is the current eglibc one |
| 20:44.19 | ssvb | desktop x86 processors have clever automatic prefetches and other optimizations which make them rather 'boring' (the end result is mostly the same no matter what you do) |
| 20:44.41 | penguin42 | yeh you just run into bus bandwidths |
| 20:45.18 | penguin42 | ssvb: Although there are questions of not just throughput but power usage while hitting that throughput |
| 20:45.46 | penguin42 | ssvb: Like people were quite resistent to using Neon because it apparently uses a lot more power |
| 20:46.54 | ssvb | yes, it does :) |
| 20:46.58 | ssvb | https://github.com/ssvb/cpuburn |
| 20:48.12 | penguin42 | oh that's cruel :-) |
| 20:49.01 | penguin42 | ssvb: I just remember there were some oddments like having something where Neon was faster from A8 cache than A9 |
| 20:49.29 | ssvb | I would also add an a15 version, but don't want to dismantle my chromebook just to connect the multimeter for the measurements |
| 20:50.06 | penguin42 | haha - I guess those odroid boards would be good for that |
| 20:51.30 | ssvb | A8 also can dual issue some NEON instructions, which makes it up to ~20% faster than A9 on properly optimized code |
| 20:51.56 | penguin42 | yeh https://wiki.linaro.org/WorkingGroups/ToolChain/Benchmarks/InitialMemset if you look at the beagle graph shows the neon memset peak |
| 20:52.11 | penguin42 | that was fun stuff |
| 20:52.36 | penguin42 | ssvb: You could use ifunc to select the memset and go forward on A15 |
| 20:53.27 | ssvb | that's too much hassle and not worth the efforts :) |
| 20:53.57 | penguin42 | ssvb: Might be on the high refresh cases |
| 20:54.33 | penguin42 | you've got a factor of 2 nearly forward/backwards on the highest refresh |
| 20:55.24 | penguin42 | wonders if you can program the raster to scan backwards through RAM :-) |
| 20:59.12 | *** join/#arm-netbook wm (~wm@p549DCCF8.dip0.t-ipconnect.de) |
| 20:59.43 | penguin42 | well DEFE_LINESTRD0_REG is 32bit :-) |
| 21:13.20 | lkcl_ | http://elinux.org/Embedded_Open_Modular_Architecture/EOMA-26 - RFC! |
| 21:30.59 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 21:56.43 | *** join/#arm-netbook DEAT (~heh@14-201-99-152.tpgi.com.au) |
| 21:59.30 | *** join/#arm-netbook stamina (~stamina@140-074-128-083.dynamic.caiway.nl) |
| 22:08.25 | penguin42 | lkcl_: eDP instead of lvds you'd get a decent res then |
| 22:09.28 | penguin42 | lkcl_: http://www.vesa.org/wp-content/uploads/2010/12/DisplayPort-DevCon-Presentation-eDP-Dec-2010-v3.pdf says you can do 1680x1050 on 5 signal wires |
| 22:25.17 | lkcl_ | penguin42: how many ultra-low-cost SoCs have eDP? |
| 22:25.36 | lkcl_ | how many $5 SoCs have eDP? |
| 22:25.46 | penguin42 | lkcl_: I don't honestly know the answer to that, but how many will have them in 18months time? |
| 22:30.17 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |
| 22:30.39 | penguin42 | I guess eDP is only tending to be for laptop panels not phone panels, so it might stay that way |
| 22:34.32 | *** join/#arm-netbook Dessimat0r (~Dessimat0@cpc1-birk8-2-0-cust884.1-3.cable.virginmedia.com) |
| 22:36.14 | penguin42 | hehe someone has got the panels used on Apple Retina products working on PCs - just by wiring the eDP out |
| 22:39.32 | lkcl_ | cool! |
| 22:41.37 | penguin42 | http://emerythacks.blogspot.co.uk/2013/04/connecting-ipad-retina-lcd-to-pc.html says he got it for $55 which is rather nice as well |
| 22:45.01 | lkcl | penguin42: honestly couldn't tell you. |
| 22:46.34 | lkcl | penguin42: but i do know that there are hundreds of LVDS panels out there, and that LVDS-to-eDP ICs are likely to be 40 to 100% the cost of the target SoCs. |
| 22:46.51 | lkcl | initially anyway |
| 22:47.02 | lkcl | ... if you can get hold of them at all |
| 22:47.54 | penguin42 | hmm kind of depressing |
| 22:51.19 | *** join/#arm-netbook tuliom (~tuliom@177.157.253.252) |
| 22:52.42 | *** join/#arm-netbook Dessimat0r (~Dessimat0@cpc1-birk8-2-0-cust884.1-3.cable.virginmedia.com) |
| 22:53.37 | lkcl | TI has eDP-to-HDMI converter ICs @ $1.50 so maybe it's not so bad |
| 22:55.05 | lkcl | MIPI-to-LVDS converter IC: $1.70 |
| 22:56.01 | penguin42 | I guess the question is what type of panels do you expect to be used |
| 22:58.50 | lkcl | http://www.panelook.com/modelsearch.php?op=advancedsearch&order=panel_id&panel_type_main=&product_composition=&brand_family=&application=&rohs_compliance=&inch_low=&inch_high=&resolution_pixels=&aspect_ratio=&touch_screen_main=&depth_typ_low=-1&depth_typ_high=-1&depth_max_low=-1&depth_max_high=-1&mount_position=&display_mode_lc_family=&display_mode_color=&display_mode_polarizer=&surface_glare=&brightness_low=-1&brightness_ |
| 22:58.51 | lkcl | high=-1&transmissive_cr=-1&gray_depth_main=&display_colors_amount=&color_saturation=-1&viewing_angle_function=-1&viewing_direction=&3d_display=&sunlight_readable=&signal_type_category=eDP&vertical_frequency=&reverse_scan=&lamp_position=&lamp_type=&lamp_unit_replaceable=&lamp_life_time=-1&embeded_lamp_driver_board=&operating_temperature_low=-1000&operating_temperature_high=-1000&button=Search |
| 22:58.53 | lkcl | 127 models |
| 22:59.00 | lkcl | (sorry about the insane size of the link!) |
| 22:59.30 | lkcl | the only bit on there that's relevant is the signal_type_category=eDP |
| 22:59.34 | lkcl | changed that to LVDS. |
| 22:59.41 | penguin42 | yeh, are eDP and LVDS costs similar? |
| 22:59.58 | lkcl | 9408 models, 2086 of which are actively in production |
| 23:00.27 | lkcl | penguin42: that's gonna take a long time to work out. |
| 23:00.40 | penguin42 | nod - anyway, just a thought |
| 23:00.59 | lkcl | wow. only 78 of the eDP models in production. *49* already discontinued |
| 23:01.12 | penguin42 | if it saves you a pin and lets you get higher res |
| 23:02.26 | lkcl | penguin42: not really. you can get away with 8 lines for LVDS - that gets you 18-bit colour @ up to appx 1280x800 or maybe 1280x1024 if you're really lucky. |
| 23:02.51 | lkcl | i did once find a single-channel 1400x900 LVDS |
| 23:03.07 | lkcl | i think... ah well |
| 23:03.27 | lkcl | but this standard isn't targetted at anything above 1280x800 anyway. |
| 23:04.43 | penguin42 | but really think it is worth thinking in terms of what the world is going to be like in say 12-18months when you might have made one |
| 23:31.11 | *** join/#arm-netbook pwhalen (~paul@CPE78cd8ec9e408-CM78cd8ec9e405.cpe.net.cable.rogers.com) |