Looking for Microcontroller Recommendations
in [Design]
|
From: Nico Coesel on 15 Jan 2010 16:14 Jan Panteltje <pNaonStpealmtje(a)yahoo.com> wrote: >On a sunny day (Fri, 15 Jan 2010 03:35:11 -0800) it happened John Larkin ><jjlarkin(a)highNOTlandTHIStechnologyPART.com> wrote in ><n9k0l59pcncno9800o9c9h6n6ee9cnoh4s(a)4ax.com>: > >>>>The future seems to be the ARM architecture. >>>> >>>>John >>> >>>I bought a book on ARM in the eighties? I was also the future then ;-) >>>But somehow x86 grabbed it. >> >>x86 has - for now - won on the desktop, but there are around 100 >>embedded uPs for every PC, and the embedded chips aren't x86. Some >>netbooks are ARM+Linux, and that may be an increasing trend. >> >>There's no justification for putting a hundred dollars (or more) worth >>of power-hogging x86 CPU into a web browser or a cell phone. Some of >>the low-end ARM chips cost well under a dollar now, and there is some >>awesome stuff for $5 or so. >> >>I'm sorta hoping that Intel will be the next DEC, except that I miss >>DEC. >> >>John > >I dunno, the ARM netbooks had a lot of trouble few month ago running some >applications (in Linux), I think some applications still use extensive asm, >a simple example present on all those netbooks is the mpeg2 / H264 / mpeg4 ... long list... drivers. >That would all have to be re-written for 'RISC' architecture is asm, and as you probably know >x86 has many many special instructions specifically for multimedia, Nope. Those drivers are already included in every major OS for ARM. See below. >So, mmx, 3dnow, give the x86 an incredible speed advantage over ARM, as ARM will have to to a zillion asm instructions >for every high level x86 instruction. >It is possible (I did not keep up after that book in the eighties) that ARM also added some special multimedia >capabilities, but then again there is the overhead of re-encoding. >So I do not expect ARM to make big inroads into personal computing and especially multimedia soon. >Why do you think Apple left their old architecture for ix86? Multimedia likely! That's where you're wrong. The ARM system-on-chips all have mpeg accellerators. Just look closely to your CPU usage while playing an mpeg movie. I have a pretty fast Intel based PC but it can't decode a 1080p movie. I need a video card with a mpeg accellerator to play such movies. You really should catch up with ARM. There is a lot of exciting stuff going on. From the ultra low power small '16 bit' LPC1100 series by NXP to the Cortex A8/A9 stuff from TI, Freescale and Samsung. -- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... nico(a)nctdevpuntnl (punt=.) --------------------------------------------------------------
From: Nico Coesel on 15 Jan 2010 16:19 MooseFET <kensmith(a)rahul.net> wrote: >On Jan 15, 3:35=A0am, John Larkin ><jjlar...(a)highNOTlandTHIStechnologyPART.com> wrote: >> On Thu, 14 Jan 2010 23:38:45 GMT, Jan Panteltje >> >> >> >> <pNaonStpealm...(a)yahoo.com> wrote: >> >On a sunny day (Thu, 14 Jan 2010 14:39:11 -0800) it happened John Larkin >> ><jjlar...(a)highNOTlandTHIStechnologyPART.com> wrote in >> ><777vk5tsp8r9616jnjcoaemmve2tc9j...(a)4ax.com>: >> >> >>On Wed, 13 Jan 2010 20:22:56 -0600, "RogerN" <re...(a)midwest.net> >> >>wrote: >> >> >>>Years back I played some with PIC microcontrollers but I've heard that >> >>>manufacturers are making better microcontrollers for less money. >> >> >>>Not looking for professional ICE or anything but maybe something compe= >titive >> >>>with Microchips in circuit programmer/debugger. =A0I was considering b= >uying >> >>>Microchips ICD3 with PICDEM 2 board for $230, but thought maybe someth= >ing >> >>>would be a better choice. =A0I'm considering Atmels line but wanted so= >me input >> >>>on others worth checking into. >> >> >>>Any recommendations on favorite microcontrollers that I can get up an >> >>>running with for a reasonable amount of dollars? >> >> >>>Thanks! >> >> >>>RogerN >> >> >>The future seems to be the ARM architecture. >> >> >>John >> >> >I bought a book on ARM in the eighties? I was also the future then ;-) >> >But somehow x86 grabbed it. >> >> x86 has - for now - won on the desktop, but there are around 100 >> embedded uPs for every PC, and the embedded chips aren't x86. Some >> netbooks are ARM+Linux, and that may be an increasing trend. >> >> There's no justification for putting a hundred dollars (or more) worth >> of power-hogging x86 CPU into a web browser or a cell phone. Some of >> the low-end ARM chips cost well under a dollar now, and there is some >> awesome stuff for $5 or so. >> >> I'm sorta hoping that Intel will be the next DEC, except that I miss >> DEC. > >As I have pointed out before the 8051 in fact out performed the 8088 >on many operations. > >A simple 32 bit Arm like processor would be just about the ideal >processor >for viewing web pages and working on documents etc. It can address >enough >memory and fling bytes around efficiently. For viewing a video a >little >hardware help could make it ok for that job. When something like >LTspice >is being run, a little more is needed: > >Programs like LTspice need to be able to do long floating point >operations >and move the floats to and from memory quickly. For this it seems >that >having a second processor that does this would be the way to go. The That is exactly why almost every ARM based system on chip has a floating point unit. -- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... nico(a)nctdevpuntnl (punt=.) --------------------------------------------------------------
From: Jan Panteltje on 15 Jan 2010 17:23 On a sunny day (Fri, 15 Jan 2010 21:14:55 GMT) it happened nico(a)puntnl.niks (Nico Coesel) wrote in <4b50d744.516134265(a)news.planet.nl>: >That's where you're wrong. The ARM system-on-chips all have mpeg >accellerators. Just look closely to your CPU usage while playing an >mpeg movie. I have a pretty fast Intel based PC but it can't decode a >1080p movie. I need a video card with a mpeg accellerator to play such >movies. So, even if they do, the trend is indeed towards integrating the graphics controller with the CPU, either in the same housing or on the same silicon. The latest Intel Atoms do this, and AMD knows about it too. Why then would you release a netbook where none of the current x86 software binaries run? Everything needs to be recompiled, and all codecs have to be re-written. For ***what*** gain? ARM is not _really_ using that less power. I would never buy one, too much incompatibility. And vendors of such high speed media applications really are not longing to write yet an other driver or codec in asm either. Not even for Linux, and most certainly not for just a few ARM based notebooks. And MS windows would likely not run on it either, unless they wrote in 7 in C.... All it's applications are closed source binaries, so forget that market! >You really should catch up with ARM. There is a lot of exciting stuff >going on. From the ultra low power small '16 bit' LPC1100 series by >NXP to the Cortex A8/A9 stuff from TI, Freescale and Samsung. Well, a simple x86 motherboard will do for now, if it has a par port, then you can even test 2x40 character LCDs with it :-) >Failure does not prove something is impossible, >failure simply >indicates you are not using the right tools... Right I just ripped the pins of a W3100A, I know I should not have tried to bend those back and get some solder wick, but alas, I was out of solder wick sooo. hehe
From: Nico Coesel on 15 Jan 2010 18:56 Jan Panteltje <pNaonStpealmtje(a)yahoo.com> wrote: >On a sunny day (Fri, 15 Jan 2010 21:14:55 GMT) it happened nico(a)puntnl.niks >(Nico Coesel) wrote in <4b50d744.516134265(a)news.planet.nl>: > >>That's where you're wrong. The ARM system-on-chips all have mpeg >>accellerators. Just look closely to your CPU usage while playing an >>mpeg movie. I have a pretty fast Intel based PC but it can't decode a >>1080p movie. I need a video card with a mpeg accellerator to play such >>movies. > >So, even if they do, the trend is indeed towards integrating the graphics controller with the CPU, >either in the same housing or on the same silicon. >The latest Intel Atoms do this, and AMD knows about it too. >Why then would you release a netbook where none of the current x86 software binaries run? This sounds a bit strange from a Linux user. Ever installed Debian on an SGI Indy? Its the same as installing it on a PC. Thats the beauty of Linux. It just runs on almost anything. Netbooks included. >Everything needs to be recompiled, and all codecs have to be re-written. Nope. That work is already done. You should get your facts right before commenting. I'm using Linux almost daily a non x86 embedded platform so I do have some real experience in that field. A x86 platform is still very inefficient. We looked into the Intel Atom and a Cortex A8 devices for a new embedded platform. The Atom just isn't suitable for real low power and small size due to the chipset. Also the price versus performance isn't that good compared to Cortex. Together with the commonly found hardware openGL and hardware video codecs (not included in the Atom!) a Cortex device has some serious muscle. -- Failure does not prove something is impossible, failure simply indicates you are not using the right tools... nico(a)nctdevpuntnl (punt=.) --------------------------------------------------------------
From: Nobody on 16 Jan 2010 03:35
On Thu, 14 Jan 2010 23:38:45 +0000, Jan Panteltje wrote: > I bought a book on ARM in the eighties? I was also the future then ;-) > But somehow x86 grabbed it. I'm fairly sure that there are more ARMs in use than x86's. Practically evey mobile phone has one, the Nintendo DS has two, iPods have one or two, and they're also common in other high-end embedded systems (e.g. GPS navigation) and in low-end computer networking devices (routers, NAS, etc). The x86 probably wins in terms of total revenue, though. |