#milkymist IRC log for Friday, 2012-05-04

wolfspraulbut I rather focus on m1 today and fix many many things on that, and then think about the next step. which is bigger/more powerful fpga, altera or -7 series, zynq (?)00:00
wolfspraulwe need customer demand, better marketing, easier product. otherwise it's just random lab tests, and we are better off with everybody using different devel boards :-)00:00
wolfspraulI think Sebastien extracted an end user classic out on the mailing list - did you see it?00:00
wolfspraulthe poor guy who cannot update his m1. "I did it about 10 times. Not sure about reproducible"00:01
wolfsprauldo we care about those things?00:01
wolfspraulor we just want to play with some cool features here and there?00:01
wolfspraulthere is no economic need or demand to make an lm32 asic, other than us learning how to tape-out and work with tsmc or whoever00:02
wolfspraulthere is hardly an economic need to upgrade the fpga, but I hope we can create that momentum...00:02
wpwrakthe problem with the lm32 core is that it's incredibly slow00:02
wpwrakin m1, we get away with this because we do very little in the core00:02
wolfspraulin the long run - of course most likely we have to "make chips", that is tape-out stuff. but where and when and which product? too early for me00:02
wolfspraulI fully understand [slow]00:03
wolfspraulwe are in this situation because that's what we wanted, no?00:03
wpwrakin a more software-centric device like a nanonote, the slowness would be considerably more noticeable00:03
wolfspraulwe look down on such lame companies as ARM and Intel...00:03
wpwrakit's what fate wants for us. for now :)00:03
wolfspraulI feel good about the direction btw, very good. and of course I like Sebastien's lm32 asic idea. but I try to put it in a realidad perspective.00:04
wolfspraulthe best result would be that we learn the process, tape-out, work with a foundry, etc. like an applied school.00:04
wolfspraulother than that there is no need, no demonstrable user demand for the result00:04
wolfspraulas you just said, I dismissed people saying "combine fpga with asic cpu" for years, well. maybe they are not so crazy after all, but it's us? :-)00:05
wpwraki think for M1 the FPGA-only approach is still tolerable00:06
wolfspraulif we make an asic cpu, the reason is? faster lm32? how much faster? it can then compete with what? who cares, who competes?00:06
wpwrakeven here the slowness makes itself felt sometimes00:06
wolfspraulthe approach with m1 is to sell it as a product where people don't compare mhz00:06
wolfsprauland I do believe in that approach00:06
wolfspraulyes I know00:06
wpwrakthe issue is whether the M1 can keep up with its application00:06
wolfspraullet's also hope that migen brings some speedup for the memory at least00:06
wpwraknow, M1 has basically a single application. this makes the task easier.00:07
wpwrakif NN-Ya-FPGA-only is to be marketed as a "universal" machine, like the Ben, then resource pressure from applications will be much larger00:07
wolfspraulmy point is: the 'make asic' idea and direction is right, and is one reason why the zynq chip would be a huge distraction. but I don't think we will have such asics running on our desks this year, next year.00:08
wolfsprauljust saying00:08
wolfspraulproove me wrong00:08
wpwrakdidn't you want to venture into making ASICs ? :)00:08
wolfspraulabsolutely, I just said I think most computing power comes from the chips anyway00:08
wolfspraulbut the discussion earlier was zynq or nozynq00:09
wolfspraulthat's a practical question, something we can most likely source in small quantities in early 201300:09
wolfspraulfor a few USD00:09
wpwrakwhat i like about sebastien's idea is that it would a) considerably reduce the complexity of the task, and b) preserve de facto all of the flexibility00:09
wolfsprauland solder together in our fabulous open process00:09
wolfspraulyeah, me too00:09
wolfspraulonly that it will not happen in reality :-)00:09
wolfspraultoo early, no demand, no marketing pressure or ideas behind it00:10
wolfsprauljust engineers wanting to try something new00:10
wolfspraula lab experiment00:10
wpwrakzynq seems to lead us back into more proprietary gateware. i do hope we'll have something like llhdl in the end. for me that's the most interesting result of all this.00:11
wolfspraulother than that *of course* milkymist has to be taped-out eventually to be sucessful00:11
wolfspraulwhat else? the end result of an IC design is to run in an fpga? no way00:11
wolfspraulyes agreed00:11
wpwrakwhy not ?00:11
wolfsprauland it gets us closer to asic00:11
wolfspraulthe "end result"00:11
wpwrakmaybe we could even synthesize part of a patch into the FPGA00:11
wolfspraulyes, of course00:12
wpwraki view the FPGA as a living thing. not just a development platform00:12
wolfspraulbut the whole point to work on a free IC design is to eventually make hardwired chips out of that design00:12
wolfspraulmust be, I think00:12
wpwrakif you can make use of the extra flexibility (which causes the extra cost) of the fpga, then this doesn't have to be so00:12
wolfspraulyes agree, the fpga reprogrammability is something we haven't even started using yet00:13
wpwrakit's like buying microcontrollers with mask ROM or with flash00:13
wolfspraulI agree with you, we just talk about a wide subject, different perspectives00:13
wolfspraulI am not against the fpga reprogrammability, still writing my thoughts on "lm32 asic idea"00:13
wolfspraulwpwrak: come on Werner. isn't that "I did it 10 times, not sure about reproducible" a classic?00:14
wolfspraulI love it!00:14
wolfspraulit reminds me who we really work for, or I at least00:14
wolfspraulusers are so cool, so free00:14
wolfspraulyou saw it?00:14
wolfspraulthat guy is so right00:14
wolfspraulI do care about that, want to make this stuff work for regular people. if it's just a devel board, there are many other/better devel boards that constantly come out with a stream of 'interesting' features.00:15
wolfsprauland, as lars knows, Xilinx is flooding the world of smart engineers with Zynq devel boards right now00:15
wolfspraulto come to a hacker desk near you... :-)00:15
wolfspraulI wanted to ask a Xilinx rep about artix-7 pricing and ISE webpack support the other day. the reply: "Don't you want to buy a Zynq devel board? we have them now!"00:16
wolfspraulhuh? :-)00:16
wolfspraulmisunderstood me? :-)00:16
wpwrakso far, we have a fairly spotless record when it comes to fatal flaws. the hard school of openmoko seems to be paying off ;-)00:16
wpwrakthey probably get a bonus if they sell you a zynq :)00:17
wolfsprauloh they are pushing it like crazy00:17
Action: roh still wants a system-mpu on the next board. and if its only sequencing the power rails ;)00:17
wolfspraulmonthly meeting "how many zynq boards did you sell?" etc. etc.00:17
wolfspraulroh: which chip do you have in mind?00:18
wolfspraulhow about multiple lm32 cores btw?00:19
wolfspraulin the fpga00:19
rohwolfspraul: it needs to be seperate of all other cores. something tiny, depending on the number of tasks. in case of only reset line and power rail sequencing even a attiny could to the job00:20
wolfspraulaside from optimizing the timing/speed of the core in the fpga (if that is possible), that looks like another way to get more CPU power out of there, no? don't know how hard it is though00:20
wolfspraulwhat's the advantage of that?00:20
rohi currently do not believe in any sense of lm32 and linux. the reachable speeds are too small by far00:20
rohwhen it comes to building something mobile i think there is currently no reasonable way not involving readymade cpu cores or soc.00:21
wolfspraulnot sure, depends on what you are building and how it is marketed00:22
rohthe advantage of having a system-mpu is that you can fix bugs later in firmware.00:22
wolfspraulthere are many ultra-low power fpgas, for example00:22
rohremember the nxp chip on the moko.. or the milkymist reset bug (fixed in hw by changing the reset chip and its powerrail)00:23
wolfspraulif you have the "iphone competitor" in mind, then your options are much less, best would be to just say 000:23
rohwolfspraul: i am thinking mostly about an successor to the nanonote we have. and i dont see it having less computing power00:23
wolfspraulok but adding such chips is again a huge distraction to our tiny community, it can easily stall and fragment the 3 folks we have hacking here and there00:23
rohbut that would be the result of using a purely fpga design from my pov00:23
wolfspraulthe natural reaction is "let's put this and that chip on it, then it's a great product"00:24
wolfsprauland that ignores entirely the dynamics of hardware and time it takes to stabilize a design etc.00:24
wolfspraulby the time all the bugs are fixed (if they ever are, most projects seem to just give up), the design is outdated and the next set of chips is glorified00:24
rohsure. i know the game now quite a time. thats why i basically dont believe a word most vendors say in any of their marketing00:25
wpwrakroh: you're basically proposing what i suggested for gta03 :) alas, then the other folks turned that little "system manager" into a monster. i think they had something with 40+ pins in the end :-(00:25
wolfspraulto speed up the lm32 we have now, how about the following options:00:25
wolfspraul1. faster clock, tigher timing constraints in the fpga? possible?00:26
wolfspraul2. more cores?00:26
rohany speeding up doesnt make it more power efficient in an fpga. and if, only by a small percentage00:26
wolfspraul3. any improvements from mmu, better caching, faster memory, other architectural improvements?00:26
wolfspraul4. better optimizing compiler? gcc or llvm?00:26
wpwrak1. i asked sebastien about that, encouraged by the high clock rates he got for the (non-trivial) memory controller. he sounded as if he didn't think there's much room for improvement.00:27
wolfspraulmaybe those angles are all exhausted/impossible/too-hard/others? don't know00:27
rohno multiplication of the performance in sight which makes it 2-4 times faster (get to the +700mhz arm9/11 range)00:27
wolfsprauloh I'm sure we cannot always just point to Sebastien and wait00:27
wpwrak2. that could be an option. but of course, limited. we already have multiple cores, with a pretty fast AVR dedicated to USB :)00:27
wolfspraulsomeone has to sit down and try independently, and then I'm sure Sebastien will be interested to discuss results, if there are any interesting ones to discuss00:27
wpwrak3. faster memory is coming. but that probably won't help the core's slowness much, if at all.00:27
wolfspraulI don't easily believe in "cannot be optimized anymore"00:28
wolfspraulroh: see you just compared to ARM again :-)00:28
wolfspraulthat was the main point I tried to get to above00:28
wolfspraulif we do that, we will always be behind00:28
wpwrak4. difficult :) and again, you can perhaps squeeze out 20-30% on average if you put a lot of work into it.00:28
wolfspraulthen it's better to just use ARM00:28
rohi think developing system blocks in fpga(s) to make a foss soc possible is a good thing. and also very flexible. but in the end it limits our system performance massivley if we stay at softcores.00:28
wolfspraulhence the make small dedicated asic idea00:29
wolfspraulthe problem with that is - you first have to beat the core running in a 45nm (or next year 28nm) fpga00:29
wolfspraulthat's why I as wasking about "are we done optimizing the lm32 core as-is"00:29
wolfspraulthe optimization problem will only get bigger00:30
wolfsprauland in the very end, our great asic will be even slower :-)00:30
wolfspraulwith our 3 hackers trying their best between social responsibilities :-)00:30
wpwrakyou reach a certain point where the thing is just fast enough for the usual tasks00:30
wpwrakyou still suffer aging (as software gets more bloated), but that's slower00:30
rohwolfspraul: my point is not arm or not arm. its "get it powerfull enough so we can sell it ourside of the foss enthusiasts's hole"00:32
wolfspraulyeah but we need to be honest about what we compare with, now or later00:32
wolfspraulthe m1 is on a marketing path where it tries to have no-one compare it to an ARM, iphone, macbook, etc.00:33
rohwolfspraul: sure. i am. comparing it to 500mhz arm cpus is comparing it to crap from 5 years ago.00:33
wolfspraulso far not successful, but any path that leads us to comparing with those is a guaranteed fail imo00:33
wolfspraulthat's why I am not comparing00:33
wolfspraulthe moment you compare you are far better off with other options, drop milkymist00:33
rohstuff i buy now is >400mhz, regardless of how cheap, as soon as i can connect serious amouts of ram/flash (non-microcontroller market)00:34
wolfspraulmoores law has more life in it, I fully expect chips to still get hundreds of times more powerful, cost is < 1 USD anyway per chip, and will stay there00:34
wolfspraulbut what's the point from the perspective of milkymist soc?00:35
wolfspraulyou could port flickernoise to the raspberry pi, and maybe even beat the m1 :-)00:35
rohwolfspraul: youve seen the carambola board?00:35
roheven by abusing a dsl-router soc we could make a better nanonote. maybe using a fpga to add a proper gpu00:35
wolfspraulno, googling. thanks! nothing better than new urls :-)00:35
wolfspraulI think we are best off tailor-making small and integrated logic boards that bring out the full power of Milkymist SoC today, and allow for healthy growth (i.e. adding new features)00:37
wolfsprauland that's it, and we take that into as many end-user products (sales) as possible00:37
wolfsprauland I say upfront: comparing megahertz or other performance with other chips will make you think this is prehistoric tech, so you may be better off comparing the latest qualcomm chips with the latest nvidia chips (for example), and can save your time about Milkymist00:38
wolfspraulwe need to send people who want to do these comparisons away upfront, to be honest with them. the only way to like the Milkymist SoC is if you free yourself from those comparisons and only think about how much you can do with the Milkymist, not how much you cannot do.00:39
wpwraklook at who compares MHz00:39
wolfspraulthen we grow the milkymist, hopefully, and of course that will include hardwired circuitry at some point, hopefully. because that would bring more performance, if enough time and money is invested, at the expense of loosing programmability/updatability for those parts.00:40
wpwrakcustomers of "universal" platforms do. you probably have at least a rough idea of how many GHz your PC's CPU has00:40
wpwraknow, how about your microwave ? :)00:40
wolfspraulI'm just explaining my thinking, anybody can compare whatever they like.00:40
wolfspraulbut people come in good faith, then they compare, then they may be disappointed. so we are better served explaining to them upfront that there's another "megahertz myth" community here.00:40
wpwrakmilkymist is by and large a system with a fixed feature set. you therefore don't have to have a lot of reserves to run all the stuff that _may_ come along. you still need a bit, but there's less pressure00:41
wpwrakso for the m1, marketing can solve the issue00:42
wpwraknow, with the hypothetical NN-Ya-FPGA-only, you're a lot closer to the universal platform00:42
wolfspraulno no00:42
wpwrakhence the MHz enter the spotlight00:42
wolfspraulwe define the marketing message. marketing is free, the point is to find and create happy customers.00:42
wolfspraulno, they don't00:43
wolfspraul"megahertz myth" wasn't that a great invention?00:43
wolfsprauleverybody knows that term nowadays00:43
wpwraki don't remember hearing that one before00:43
wolfspraulI think it was Jon Rubinstein who created it, or is known to have created it.00:43
wpwrakjust the GHz race :)00:43
wolfsprauloh no, that was big in the 90's00:43
wolfspraulat Apple, of course, since they were lagging in mhz00:44
wolfspraulso they thought about what they could do :-)00:44
wolfsprauland they started talking about 'megahertz myth' a lot! systematically00:44
wolfspraulno need to give up, learn from the pros :-)00:45
cladamwwpwrak, i saw backlog. thanks for your feedback, I'll take some time to modify. :)00:45
wolfspraul"The term came into use in the context of comparing PowerPC-based Apple Macintosh computers with Intel-based PCs."00:45
wolfsprauland yes it was good old Rubi00:46
wpwrakfunny. once upon a time, when i had that SGI at my disposition, the typical "modern" sun had a 68030 at some 20-30 MHz. the SGI has a MIPS at, i think, 12 MHz. yet it was much faster. even though it was RISC and didn't have the mighty instructions of CISC :)00:46
wolfspraullater at Palm, and now at Amazon behind the Kindle strategy00:46
wpwrakcladamw: ah great, you found it :)00:46
wpwrakcladamw: i have a few more comments. still have to type them, though00:46
wolfsprauleven if we go asic, our mhz will lag FAR FAR behind anything you could possibly compare with, anything contemporary00:47
wolfsprauland like I said, you actually first have to beat the soft-core in a 45 or 28nm gate structure00:47
wpwraki don't think we need to match apple and enemies MHz for MHz00:47
wolfspraulwith our minuscule engineering resources even the tape-out, testing and communication process with a foundry could get us stuck for 10 years :-)00:47
wolfspraulno my point is: we are free in how we market and position our tech00:47
wolfspraulif we compare, then of course others compare00:48
wolfspraulif we invent smart terms like "megahertz myth" and stick to them, there is not much reason actually for people to not listen00:48
wolfspraulmaybe we have a point, maybe not00:48
wpwrakyou'd still find it difficult to sell devices that aren't fit for the intended purpose. at least in the long run00:48
wolfspraulbut I am sure - comparing with arm/nvidia/qualcomm/samsung and others will go nowhere00:49
wpwrakso ... how would you market the NN-Ya-FPGA-only with its 80 MHZ (?) core ?00:49
wpwrakroh: seems that i found the right question (-:C00:54
wolfspraulnice http://geekculture.net/joyoftech/joyarchives/243.html00:54
wolfspraulmegahertz myth cartoon00:54
wolfspraulclick on 'next joy' to continue00:55
wpwrakheh, great minds think alike :)00:55
wpwrakthe page was loading when i read your message00:56
rohwpwrak: hrhr01:14
rohwolfspraul: mhz is partially a myth. but in generic computing its still a fact. sure. we dont need 2ghz. but 80mhz are just not enough.01:14
rohi just tried ssh-ing onto a dvb-set-top-box. it tool like 20 seconds before it responded again because it was doing fscking crypto math01:15
rohwhy? because its a dirt cheap mips cpu which is ok to do controlling of data streams which dsps do the work on, but not for something simple as responding to a incoming ssh connection nicely01:16
wolfspraulbut "80 mhz" is a number and unit, and "how long it takes to login with ssh" is a use case01:16
rohthere are 2 answers to that question: one would be 'then dont do what its not designed to handle' ... which would lead to 'then dont do stuff' transferred onto the nn01:17
wolfspraulI don't know whether your dvb settop is selling well or profitable for the manufacturer, but I would guess that an ssh login is nowhere anywhere on the typical use cases for that product01:17
wolfspraulon the list of ...01:18
rohalso its a question of 'computations per timeframe' . sure. if you make an usb stack which can do all the work which currently a cpu does on linux mostly in hardware' .. but it dont see that transferred to all subsystems01:18
rohwolfspraul: it was an example of 'whats in that stuff nowadays'01:18
wolfspraulso my thinking for m1 is quite clear now - one is to continue to focus on its current and intended use, and make that better. there are ample opportunities to do that.01:18
rohin that case.. a 5 year old soc. pre-h264-hd area01:18
wolfsprauland the other one is to build an integrated and simple logic board around the milkymist soc to bring out what it can do today, and create room for growth01:19
wolfsprauland in parallel, work on credible marketing messages and sell that box to users01:19
rohjust an example. low end dsl-routers like the tp-link or the carambola soc have 300-400mhz mips cpus. which is ok mostly. just dont do any crypto or computation intensive stuff there.01:19
rohso you can route and switch packets. but crypto they only do sanely with hardware support (most router socs do have that nowadays. even the brcm soc in the original linksys wrt had)01:20
rohbut compared to a 'user facing frontend node with display and keyboard' i must say that cpus need more punch. even if its only to drive a serious amount of pixels or decode a audio file for which you dont have a hardware codec01:21
rohive learnt recently that nowadays even the cheap mp3 players dont do dsps anymore. they have small arm cpus and well enough written code to punch that work through in sw. why? because its cheaper then binding yourself to some dsp/cpu vendor and their toolkit and you can reactor market changes like new codecs faster.01:22
wolfspraulinteresting little carambola board01:22
wolfspraulI have *no* idea how that company wants to turn this into a good business though01:23
rohyeah. i thought so.01:23
wolfspraulralink was acquired by mtk01:23
rohbusiness? well.. all the arduino crowd looking for more punch.01:23
wolfspraulthey are making a module for a niche market against a small window of opportunity01:23
rohor serious ethernet connection01:23
wolfspraulok but those are all tiny phenomena01:23
wolfspraulmaybe a small lithuanian company can survive on that, don't know01:24
wolfspraulhow much are those boards?01:24
rohsure. but they will get a lot of people like me who would atm do projects re-using routers like the tp-links01:24
wolfspraulpeople like you are a tiny phenomena :-)01:24
wolfspraulrounding error01:24
wolfspraulhow much does this board cost?01:24
rohnot because they are like 7E more expensive than the router, but because it dont need to hope for continoued supply of something which is a market accicent not a planned feat01:25
rohcarambola was like 22E01:25
wolfspraulwell like I said the main chip on that board is already gone as an independent business as well, acquired by mtk (similar to qualcomm's purchase of atheros)01:25
rohand a tp-link costs me 15-23E depending on where and how many01:25
wolfspraulbut tp-link has massive (read: tens of millions) scale behind it01:25
wolfsprauland even that was not enough to keep Atheros independent01:26
wolfspraulwhy do you say 'was' - carambola no more?01:26
rohsure. still they could change the product anytime in a way that it makes it unpractical for me.01:26
wolfspraulI find the pricing crazy aggressive, well. that makes me wonder even more about their business.01:26
rohso carambola is intended to be used in verticals, while abusing openwrt based routers isnt.01:26
wolfspraulbut carambola has zero influence over the chips either, for sure01:27
wolfspraulin fact, ralink is already history01:27
rohwho's pricing?01:27
rohthe chips they use are single digit money. the expensive part is sourcing i bet01:27
rohsame as tp-links business. there is nothing to make it cheaper anymore. its commodity hardware.01:28
wolfspraultrue, many tech companies are on a nice path to reduce prices to 001:28
wolfspraulnobody needs them01:28
wolfspraulbut... tp-link will increase volume and has a giant global distribution network which I would well imagine is growing01:29
wolfsprauland either you have that or not. imagin you want to get into the AA/AAA batteries business.01:29
rohi think they cannot get much more market share. their spectrum is low-end home cpe stuff.01:29
wolfspraulwhat do you need so that your batteries are flying around in millions of stores worldwide?01:29
wolfspraulsure, and?01:29
rohhigh end is with asus and companies like arcadyan or avm which do oem for the big telcos.01:30
rohfrom my pov the real numbers are done via the oem telco designs. not the independant sales.01:31
wolfsprauland you think those customers will not want to squeeze out every penny?01:31
rohcheck heise for the arcadyan scandal regarding broken wifi keys (default keys)01:31
wolfspraultheir certification cycles may be longer that's why they appear a few generations behind01:32
rohthey have 5-6 digits of active devices per model01:32
wolfspraulwhat you describe as 'high-end'01:32
wolfspraulwhy did you say carambola 'was'?01:32
rohhigh end is gigabit ports, integrated vdsl modems etc. many features etc. usually has more recent soc, more ram and flash01:32
rohcarambola uses a low-end soc01:33
rohwolfspraul: s/was/is/ . sorry to confuse01:33
wolfspraulyeah, impressive01:34
wolfspraulI have no idea how that can compete with the flood of cheap tp-link models though01:35
wolfspraulthe only thing that differs are the many pins you can connect to01:35
wolfsprauland the it becomes a 'development board' as they say. maybe sparkfun/adafruit would be the strongest sales channels for that, but even then I doubt you can sell even a few thousand of it.01:36
rohand that its intended to be used in verticals. and the site01:36
roheh size01:36
wolfspraulthen you do the math, and I have no idea, really none, how you can finance anything with those margins01:36
wolfspraulit's the old problem of most 'module' makers01:36
rohi think it has a real chance as kind of a high-end arduino/arduino companion01:36
rohsure. its a market nice01:37
roheh niece?01:37
rohniche. thats the order of chars i wanted01:37
wolfspraulat adafruit/sparkfun a product with that type of bom would probably cost 40 USD or more01:37
wolfsprauland even then it's a small hobbyist/nerd market01:38
GitHub190[board-m1] adamwang pushed 1 new commit to master: https://github.com/milkymist/board-m1/commit/9a3c2737e8e95452f27afbcc4bb7bca7202d93d901:38
GitHub190[board-m1/master] make D+, D- lines with more clearance - Adam Wang01:38
wolfspraulthe moment any actual vertical business would start using this, they would quickly optimize the board away as their volume goes up01:38
rohmaybe. who knows what that lithuanian company does the rest of the day. maybe they do oem dsl routers all day and thats just a 'additional nice product' they could do because they wanted01:38
wolfspraulsure, it's nice. development board with unbelievably low margin.01:39
rohwolfspraul: sure. but some verticals dont exist when there is enough knowledge to do bga boards etc.01:39
rohwolfspraul: often there is 'we need foo, but only 10-100 times'01:39
wolfspraulall module makers I have seen are transient businesses01:39
rohthink art projects.. they use a lot of arduinos for that01:39
wolfspraulafter 2 years they are either dead or are doing something else, higher margin01:39
rohi think only extremely few people who prototype with arduino or similar boards build their own boards in the end. most just use and adapt some existing board. there is only gain in optimizing that away (and having the risk to fuck it up) when you do real mass-market01:40
wolfspraulwhich arduinos have you seen growing into vertical industry use btw?01:41
wolfsprauland 2 other questions: how many carambolas have you bought so far? and how many cheap tp-link routers?01:41
rohnot industry. vertical sure. projects for artists01:42
GitHub8[board-m1] adamwang pushed 1 new commit to master: https://github.com/milkymist/board-m1/commit/dbc59e37ae1dce363b74b1a8fdeb016f9c5baf7f01:42
GitHub8[board-m1/master] make ground symbols downward if there's more space. otherwise upward. - Adam Wang01:42
rohcarambolas none (because of paypal and my lazyness) and tp-link 201:42
wolfsprauldoes it grow into commercial/consumer electronics use anywhere?01:42
wolfspraulso the small inconvenience of carambolas unusual sales channel has kept you away from them so far :-)01:43
rohbut i have a project where we will use 20 or so of the routers soon. but we need a switch and the case so it will be tplink i guess01:43
wolfsprauland above I was talking about tp-links giant global sales network with certainly hundreds of thousands of points-of-sale01:43
rohwolfspraul: you need to get away from the massmarket aspect a bit. sure, thats your pov.01:43
wolfsprauland... more tplink :-)01:43
wolfspraulno no01:43
wolfspraulI am just thinking about what you say01:43
wolfsprauland you admit: so far 0 carambola because the sales channel was too inconvenient01:44
wolfspraultp-link was easier01:44
wolfsprauland the upcoming project, also tp-link01:44
wolfspraulinteresting :-)01:44
wolfspraulmaybe you are not alone...01:44
rohbut there are many other products which would never work in a mass-market but only their niches. which is nice, because massmarket products are boring and mostly optimized to death and thus less flexible01:44
wolfspraulI absolutely don't just think mass-market, not at all01:44
wolfspraulbut we are talking about carambola01:44
wolfsprauland I think hacking tp-link boards looks more worthwhile doing01:44
rohi have one long term project where i will test carambola for, which i had planned with a pcengines board till recently01:45
wolfspraulalthough even that is such small volume that it will never even make it to the radar at tp-link01:45
wolfspraul(sorry btw this is all off-topic for milkymist)01:45
rohdepends on how small i think my userland will be what board i will choose in the end01:45
wolfsprauland then about the arduinos, the last years that I am following I don't see that growing into consumer use01:45
wolfspraulnot just 'mass market', that's not what I mean01:45
rohwolfspraul: i think their goal and businessplan never intended to compete with tp-link in the first place01:46
wolfspraulbut the tech stays with a user base of 'experts', 'hobbyists', 'artists', 'nerds'01:46
rohfuck customers. they are boring.01:46
wolfspraulthat's roughly what the guys at mtk, nvidia, qualcomm, arm, and and and would be saying, just about 'hackers' :-)01:46
rohi dont believe that. they know that those are the ones working for them and having all the creative ideas who make new products.01:47
wolfspraulI am not blaming carambola for anything. I think they know what they are doing. like you said, they may have needed this board for themselves, and then just sell a few as well. that's all.01:47
wolfspraulnah, too small01:47
wolfspraultoo few01:47
rohthe world IS small.01:47
wolfspraulnot bad at all, but just too few01:47
wolfspraulplease look at the margins01:48
wolfspraulif something like carambola is priced at 100 USD or more, and they can find customers, that's a different story01:48
rohatleats when it comes to where the tech-savyness starts and ends. the majority of that stuff we are talking about is developed and run by very few people.01:48
wolfspraulit's easy to complain about the big guys being ignorant, when you forget what margins they work with, and are forced to work with by us, the customers01:49
rohi am not talking about making the current nn or a future one a tool for the mass-market. but to find a way to make it more interrresting to the tech-savy customers we have now01:49
rohsure. there are no margins in commodity electronics for the mass market. but did we ever seriously try to target that? i dont think so (and i am glad. customer support is hell there)01:50
rohthere are much more gains in the 'knows more what he/she is doing and likes fancy toys' market.01:51
wolfspraulnot sure, hackers are cheap :-)01:51
wolfspraultheir hacking budgets get cut by their girlfriends01:51
wolfspraulbut the arduino still passes, thank god01:52
rohthats why people need to get a bang for their buck. not cheap, but good and fairly prices is the sweet spot.01:52
rohthe hackers i know have girlfrieds who also hack *sigh*01:53
rohmaybe we should make a nn with a minipci-e slot, and not wire pci-e there but only the usb. then atleast a 3g modem would work01:54
roh*sigh*. i think before concepting anything new there we need to fix our mechanical problems. or is using what the nn did still an option?01:56
wolfspraulI agree mechanical needs to be simplified01:57
wolfspraulman I feel bad about spamming the milky channel, taking it to qi-hw01:57
kristianpaulkeep up with its application, and the history of the less know pdf reader feature :)03:04
wpwraktouch\'e ;-)03:06
Fallenouwolfspra1l: I think the 80 MHz softcore is enough for the M1 product for now (it does it's job with no obvious lag). it's a pretty precise use case though13:30
Fallenouwolfspra1l: for something like a general purpose computer (like the Ben is AFAIK), it might be really slow13:30
wolfspra1lmight :-)13:30
Fallenouhigh probability that it will :/13:31
FallenouHaving read lm32 softcore verilog code a bit while working on the MMU, I don't see a lot of room for improvement *BUT* I don't have the verilog expertise of lekernel13:31
wolfspra1lis the cpu pipelined?13:32
wolfspra1lI would think so13:32
Fallenouit's a 6-stages pipeline13:32
wolfspra1lcould we go multicore?13:32
Fallenouaddress/fetch/decode/execute/memory/write back13:32
wolfspra1lhow about caching?13:32
wolfspra1lbranch prediction?13:32
Fallenoucaching is already enabled13:32
wolfspra1lout of order processing?13:32
Fallenoubranch prediction is there, but maybe it can be improved, it's a really really prehistoric one13:32
Fallenoumulticode : I have no idea if this could be easily done , but yes, why not :) we would have to calculate the FPGA resources to check if it fits inside one FPGA13:33
Fallenouout of order ... I think it's not implemented right now in LM32 (no sure), seems hard to implement. maybe lekernel out of order DRAM controller will be enough to gain more performances13:35
Fallenoukeep in mind that the more gates you add in the LM32 design, the more risk you take to see your "max working frequency" to actually decrease13:36
Fallenoubut it does not mean we cannot give it a try :) it just means it's risky if you don't pay attention13:36
FallenouHopefully, FPGA technology improvements will give us extra MHz without any LM32 code change, I don't know how much though ... could we reach 200 MHZ ? with which FPGA ?13:37
kristianpaulmay altera? may be takinng apart lm32 separate could clock higher than a SoC13:42
Fallenounot so sure which part is slowing down the other: the soc or the lm3213:43
kristianpaulsoc have some extra wiring remenber wishbone and FML13:43
kristianpauland i guess we could get a better timing report, just we/i need to try..13:44
Fallenoucould be interesting to synthetise a migen based (milkymist-ng) soc with just uart+sdram+lm32 to see the max working frequency13:45
kristianpauli would said that for old soc, yes is interesting13:46
kristianpaulOn Arria2 it reaches 175Mz13:46
kristianpaulBut those where most simulations i dont remenber well what  Terpstra said13:46
kristianpaulFallenou: also, once you pass out 100Mhz barrier, i bet interesting behavior pop up :D13:47
Fallenouthis demi runs @ 50 MHz : http://www.latticesemi.com/products/intellectualproperty/ipcores/mico32/mico32ddrsdramdemo.cfm13:49
Fallenouhttp://www.latticesemi.com/products/intellectualproperty/ipcores/mico32/index.cfm at the bottom of the page13:50
Fallenou115 MHz on LatticeECP313:50
Fallenouwhich is a basic Lattice FPGA, not a big big fpga :)13:50
FallenouI own a dev board with this fpga13:50
Fallenou$99 dev board13:50
kristianpaulso slow..13:50
kristianpaul:) and is lattice13:51
wolfspra1lwe should not compare what we know runs on our board with all sorts of unverified numbers from web sources13:57
wolfspra1lwith our experience, we should not :-)13:57
wolfspra1lthere are a hundred ways how you can reach "xxx mhz"13:57
wolfspra1la thousand13:57
wolfspra1lit's not just good/bad, it's that we don't know the details, then we just carry forward the number and forget all sorts of compromises that may have been made13:58
wolfspra1llike "ok, it froze after about 10 seconds, but until then it ran nicely at xxx mhz"13:58
wolfspra1lor hundreds like that13:58
Fallenouhum it's official Lattice demo, I bet they don't publish demos which freezes after 10 seconds :)13:59
wolfspra1la real company demo13:59
wolfspra1lthen that number must be clean like the pope's shirt13:59
wolfspra1lsorry I take back all I said13:59
wolfspra1lit's a real company demoe13:59
wolfspra1lI read a report the other day about the "megapixel" ads in smartphones14:00
wolfspra1lthe number that is published in advertisement, and any actually different 'pixels' you can determine in the outcome14:00
wolfspra1loh well14:00
wolfspra1llet's say those numbers have taken on life of their own14:00
wolfspra1lsome vendors (ahem, cough, Apple) are publishing mostly actual numbers that can be verified, sometimes what apple publishes is even less than the actual measured performance14:01
wolfspra1lbut a lot of the others, oh well14:01
wolfspra1lpublished number 50% above what can be measured14:01
wolfspra1lso back to reality, if this is an "official lattice demo", then FOR SURE you should not believe that number14:02
wolfspra1l"official lattice demo" means that their marketing decided that it had to be that number14:02
wolfspra1land then engineering had to somehow fudge it up, but marketing would say it anyway14:02
wpwrakat least if the result is great enough for the competition to hate it, it may get some vetting14:06
wpwraksome of the more outrageous "optimized" benchmarks have been exposed in such situations14:07
wolfspra1lwell, it's just the fact14:07
wolfspra1l"official company demo" means, to anyone experienced, that the number will not hold in real life14:07
wolfspra1lOF COURSE NOT14:07
wolfspra1llook, give you an example :-)14:07
wolfspra1lnvidia is a successful semiconductor company14:08
wpwrakbut anyway, none of these numbers sound too promising. i don't see anyone claiming more than a 100% increase, relative to what we have now14:08
wolfspra1lwe all agree14:08
wolfspra1lthe CEO has repeated many times his philosophy, in interviews, etc.14:08
wolfspra1l1. sell first14:08
wolfspra1l2. worry how to make it14:08
wolfspra1lhe is right!14:08
wolfspra1l*if* a customer comes and puts money on the table, anything can be made14:08
wolfspra1lso first claim something14:08
wolfspra1lsee whether money shows up14:08
wolfspra1l*then* talk to your engineers :-)14:08
wolfspra1lthat's not a secret or some fraud or anything14:09
wolfspra1lthat's how the system works, in the example of a very successful company (nvidia)14:09
wolfspra1ljust saying14:09
wolfspra1lFallenou can think about it if he wants :-)14:09
wolfspra1lsell first, then worry about making it...14:09
wpwrakas long as your engineers can always deliver enough on your promises, no problem14:09
wolfspra1lthat's their problem14:09
wolfspra1lthat's why they get paid :-)14:09
wpwraki somehow prefer the engineering-run companies ;)14:10
wolfspra1lthe nvidia ceo also believes in hiring the top talent from the whole world because he believes talent is so limited that if you hire them all, your competitors have a problem simply from the fact that all the good guys are gone already14:10
wolfspra1lthose things all go together14:10
wolfspra1lso "company demo" does not mean that it actually works14:10
wolfspra1ljust remember that14:10
wolfspra1lit does not14:10
wolfspra1lnot all companies would be as drastic as "sell first, then worry how to make it", but it's a good reminder of a valid strategy of a very successful semiconductor corp14:11
wolfspra1lcertainly far more successful than lattice :-)14:11
wpwrakwell, different resources, different customers, probably also requirements changing at a different pace14:12
wolfspra1lbut you would agree with me that "company demo" is a very good indicator that whatever they demo does not *actually* work :-)14:13
wolfspra1lthat's our bottom line14:13
wpwrakand the negotiators may also be from different branches14:13
wolfspra1lthat's why it's called a "demo"14:13
wolfspra1lmaybe we all want this to work, yeah :-)14:13
wolfspra1la dream14:13
wolfspra1lcompany dream14:13
wpwrakdemo can be pretty much anything. maybe it works, maybe it's a complete fake. (e.g., a video instead of the machine doing it in real time)14:14
wpwrakmaybe it's easy to reproduce (with some tolerance), maybe not.14:14
wpwrakmaybe the thing you see being demoed works on the inside like your application, maybe not14:15
wpwrakthe only somewhat reliable benchmark is your own application14:15
wpwrakergo, make sure it's easy to port :)14:16
Fallenou16:15 < wolfspra1l> sell first, then worry about making it... < I totally agree. Every company does this :)14:36
wolfspra1lok then, we are all on the same page about the 'company demo' and can focus on the demonstrable performance on m1 :-)14:50
wolfspra1lnot demonstrable, *usable* even :-)14:50
Fallenousure it's a demo, but they publish documentation and source code, they can't lie about the working frequency :p14:59
Fallenouit's too important a number for them to lie14:59
Fallenouanyway , their number is not crazy, so it is not really hard to trust15:00
Fallenouit must be right (115 MHz)15:00
Fallenouit's not like 200 MHz :)15:00
FallenouIIRC Milkymist was running at 90 or 100 MHz a few years back :)15:00
Fallenouon Virtex 5 *agreed*15:01
Action: Fallenou heading home, bye15:15
wpwrakmaybe it's more than a demo. more like an application example15:15
hypermodernAt LGM talking about Qi Hardware17:23
sb0and, as always, I have a good reason not to be at LGM... :(17:52
Action: Fallenou is back19:23
qi-botThe firmware build was successful, see images here: http://fidelio.qi-hardware.com/~xiangfu/build-milkymist/milkymist-firmware-20120504-2123/21:06
qi-botThe firmware (using branch) build was successful, checkout the VERSIONS for detail, see images here: http://fidelio.qi-hardware.com/~xiangfu/build-milkymist/milkymist-firmware-20120504-2306/22:51
--- Sat May 5 201200:00

Generated by irclog2html.py 2.9.2 by Marius Gedminas - find it at mg.pov.lt!