Prev: Tiny Bootloader
Next: Link&Locate 86?
From: Bill Giovino on 30 Aug 2006 23:52 "Ian Bell" wrote... > Bill Giovino wrote: > > > Except for companies where the 8-bit is their primary company strategy > > (like Microchip), there is a subtle defocusing of 8-bit from the major > > players in order to make room in the fab for higher margin parts. The real > > growth for embedded systems is in low-power 16-bit processors. > > > > But really, the answer to your question all depends on what you want the > > data for. Are you looking to serve a market with tools, compilers, > > services, etc. or are you planning a new microcontroller introduction? Are > > you looking at competitive or non-competitive cores and markets? You have > > to treat the 8051 separately because it is less an 8-bit product, rather, > > it is a separate market segment. These are all important issues. > > > > Regards, > > > > Bill Giovino > > Thanks for your input Bill. I want the information simply to satisfy my > curiosity. Are there really more PICs made a year than 8051 derivatives? Or > does Motorola still head the 8 bit pack. It is said that more 8 bitters > are used in Japan than anywhere else - so how many and what type? Just how > many of each are made? How many AVRs etc etc? What proportion (by number by > value) is the 8, 16 and 32 bit segments and how have these changed over the > years? and so on. The last publicly available data for rough market > breakdown is for 1990 to 2000 and is: > > WorldWide Microcontroller Shipments (in millions of dollars) > > '90 '91 '92 '93 '94 '95 '96 '97 '98 '99 '00 > 4-bit 1,393 1,597 1,596 1,698 1,761 1,826 1,849 1,881 1,856 1,816 1,757 > 8-bit 2,077 2,615 2,862 3,703 4,689 5,634 6,553 7,529 8,423 9,219 9,715 > 16-bit 192 303 340 484 810 1,170 1,628 2,191 2,969 3,678 4,405 > > > WorldWide Microcontroller Shipments (in Millions) > > '90 '91 '92 '93 '94 '95 '96 '97 '98 '99 '00 > 4-bit 778 906 979 1036 1063 1110 1100 1096 1064 1025 970 > 8-bit 588 753 843 1073 1449 1803 2123 2374 2556 2681 2700 > 16-bit 22 38 45 59 106 157 227 313 419 501 585 > > What has happened since? > > Ian Hi, Ian! First, let me tell you what I've been telling companies for the past, oh, eight years: "The death of the 8-bit microcontroller has been greatly exaggerated". Now, Microchip leads the pack in 8-bit, both literally and in spirit. Japanese manufacturers prefer Japanese uC suppliers. The 8051 market saw it's first hit ever with ARM's acquisition of Keil, as ARM's own public analyst statements clearly show that their strategy is to move 8051 users into ARM. I'm seeing very strong growth in embedded 16-bit because 16-bit has more processing power than 8-bit, and lower power than 32-bit. That's why TI's 16-bit MSP430 is such a strong performer in the marketplace. http://www.microcontroller.com/news/ti_msp430_50newdevices.asp (roadmap) I'm presently working in an accurate analysis of the present market to be posted on Microcontroller.com, but it won't be ready for another month. BTW, whose figures are those above? Bill Giovino Executive Editor http://Microcontroller.com
From: Jim Granville on 31 Aug 2006 01:52 Ian Bell wrote: > Robert_Teufel wrote: >>I work for a company that manufactures 8051 derivatives and we follow >>the market all the time and I try to keep this as unbiased as possible. >>The leader in units is Microchip, accounting for all the PIC1x devices. > > > I have heard this anecdote several times (on the web) but it is always > unqualified. Does it mean Microchip shipped more 8 bitters than any other > single 8 bit manufacturer? Yes - and it also bundles ALL those different PIC cores into one basket. > if it does it tells only part of the story - or > does it mean more PICs were shipped than any other type of 8 bit micro - a > totaly different statistic. I suspect there are still more 8051 derivatives > shipped than PICs for example but I need some way to get these figures. 80C51's are over 1 billion / year, and microchip is approaching that, but only if they fudge things by pretending all their PIC cores are the same :) PICs would own the low-pin-count, Low IQ, business, and the low ASP of PICs shows how many rice-grains they ship. - and also shows how slow takeup has been on PIC18 and dsPIC. However, the 80C51 is now moving into the low-pin-count territory, as is the AVR, and Zilog et al, so their base-segment is getting more crowded. eg at 20 pins, you now have 16K Flash 80C51 variants. Microchip's analog business is growing faster than their uC's. > >>The leader in Dollar volume is probably still Freescale (latest data I >>have is 2004), including ancient devices such as the HC05, all 08 >>devices, HC11 and may be some more. > > > They have been for some time; plenty of anecdotal evidence of that. Freescale wins, partly because many japanese vendors put thiers into the 16 bit basket. They also rake up all their 8 bit cores, to do this. > >>The 8051 claims to to have the widest coverage of applications or in >>other words, most designs of all 8-bits. > > > Does it? Where is this written? That would really be a no-contest, especially if you properly compared core-by-core, and not by brand name. > >>The AVR does not make it into this most... list, yet but it is >>definitely a another "force" in the 8-bit. > > > Interesting because I found a news item from just a few years ago that > claimed the AVR had 30% of the 8 bit market. That was carefull spin - it was not 30% of the 8 bit, but 30% of the flash 8 bit, at _that_ time, and by volume. Helped by a few things : Hitachi puts most of their mature flash into 16 bit stats; Atmel ship large volumes of sub 50c AVRs; and PICs were slower to move to Flash, and did so on the more complex devices first. You can reality check this for 2005/2006 by comparing Atmel/Microchips annual reports : total shipped (all) AVRs [all cores, mask & ram & smart cards] passed 500M a little while ago, whilst (all) PIC's are into their 3rd billion IIRC. I think PIC has also recently passed 50% revenue from flash threshold. Cumulative 80C51's are probably comfortably past 10 billion here in 2006. > >>Going with any of these architectures should be fairly save for years >>to come. > > > Now at least two people have misunderstood my question. Perhaps I did not > explain myself well enough. I am not trying to make a design decision. it's > just you hear all sorts of anecdotes about the relative sizes of 8,16 and > 32 bit markets, who has what share of which by volume, by value and so on > but I am trying to get a consolidated picture of this. Market research > companies create reports providing this info but the cost thousands of > dollars. and they are largely useless :) -jg
From: Chris Hills on 31 Aug 2006 03:34 In article <q6ydnfOLxM5rwWvZnZ2dnUVZ_sidnZ2d(a)comcast.com>, Bill Giovino <editor1(a)nospam-microcontroller.com> writes >Hi, Ian! > >First, let me tell you what I've been telling companies for the past, >oh, eight years: >"The death of the 8-bit microcontroller has been greatly exaggerated". Absolutely... There are still 4 bit systems out there (Japanese toy market?) > >Now, Microchip leads the pack in 8-bit, both literally and in spirit. Japanese >manufacturers prefer Japanese uC suppliers. The 8051 market saw it's >first hit ever with >ARM's acquisition of Keil, as ARM's own public analyst statements >clearly show that >their strategy is to move 8051 users into ARM. I'm seeing very strong >growth in embedded >16-bit because 16-bit has more processing power than 8-bit, and lower >power than 32-bit. >That's why TI's 16-bit MSP430 is such a strong performer in the marketplace. >http://www.microcontroller.com/news/ti_msp430_50newdevices.asp > (roadmap) I would say that the market will split into 8 bit and 32 bit. Apart from a few specialised parts like the MSP430 the 16 bit market will die out. Book mark this post and get me to eat my hat in about 5 years time :-) -- \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\ \/\/\/\/\ Chris Hills Staffs England /\/\/\/\/ /\/\/ chris(a)phaedsys.org www.phaedsys.org \/\/\ \/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/\/
From: Jim Granville on 31 Aug 2006 05:07 Chris Hills wrote: > > I would say that the market will split into 8 bit and 32 bit. Apart > from a few specialised parts like the MSP430 the 16 bit market will die > out. > > Book mark this post and get me to eat my hat in about 5 years time :-) Do we have to wait ~5 years ? Spin and marketing will conspire against such predictions. A good, very current example, is the ZNEO from Zilog. Zilog pitch this as a 16 bit uC, as the base opcode is 16 bits, with some larger opcodes. But it has 16 x 32 bit registers, and can do 64 bit operand maths. Compare that with the CortexM3 (another new core), it has 16 x 32 bit registers, and the base opcode is 16 bits, with some 32 bit ones. It can multiply to 64 bit result, but seems to lack a 64/32:32 - this is pitched as 32 bit controller. Who is 'right' ? This shows the flaw in trying to firstly pigenhole uC into boxes, and then moving pick winner(s) and looser(s). Freescale look set to somewhat abandon these 8/16/32 bit pigenholes, so maybe in 5 years time, we'll look back on attempts to quantify complex devices with a single number, as quaint ? :) -jg
From: Ian Bell on 31 Aug 2006 13:43
Jim Granville wrote: > > Google for Microcontroller market share, and you'll find many snippets. > Been doing that - got quite a few on file - but it is rather tedious and somewhat incomplete. > Usually from some company claiming they have gained on the pack.... > > eg August 2005 > "... 32-bit market -- a market Sanghi estimates is worth $3 billion a > year. The 8-bit microcontroller market, in contrast, is now hovering > around $5 billion a year." > I assume that's true so I'll add it to my snippet file. > Also, annual reports have good info, where you can work out ASPs and > run rates. > These can also "reality check" some of the boasting, and some of the > claims out there do not survive this audit trail! > That's a good idea - I think many are publicly available. Cheers Ian |