8 bit microcontroller market
in [Embedded]
|
Prev: Tiny Bootloader
Next: Link&Locate 86?
From: Yuriy K. on 12 Sep 2006 23:36 John F wrote: >> Or you'll soon learn to choose bigger processor. Trust me. :P > No. I won't ;-p Well, you had you chance... >> You might need one more little feature, but there is no room for >> improvement. > > There is always room... just port one or two things to asm and see, > there is enough place for putting it in twice. Everything is already in asm. O-ops. Nothing to port and single unused byte in code memory. >>>>>> I write all my interrupt routines in C. >>> Less than 5 lines of ASM is hardly ever possible. Still 80% of time >>> is taken there. >> Inadequate processor or overall system architecture selection. >> Both 5 lines and 80% of time tell that. > Well... No. Non sequitur! Whatever. Without detailed description it's a pointless discussion. >>>> Unfortunately, I did quite a lot of x51 assembly programming. Just >>>> did not know better at that time. Fortunately, these days were >>>> long >>>> ago. >>> They are back! >> Not for me. :) > Then your field of operation will be very limited. I have more work in foreseeable future than I could handle I would say the job security is being way too good lately. ;) -- WBR, Yuriy. "Resistance is futile"
From: Yuriy K. on 12 Sep 2006 23:41 Joerg wrote: >>>>> OrCad SDT-III 3.22 >>>> Inferior to almost any current PCB design package. Very limited >>>> capability. >>> In which ways is it inferior and limited? >> In productivity. I.e. how long will it take you to get from idea to >> the routed PCB. I used it a little about 15 years ago. It was quite >> good at that time, but PCAD 4.5 happened to have more features and >> complete support from schematic to PCB and backward. AFAIR, SDT is >> just a schematic capture. One still had to use a separate PCB routing >> software. > It is but you could also get a layout package from them. It was bad. BTW, does SDT have more than one undo level? > And we still have the same incompatibilities > where the only way to get the data into the layout package is a netlist. Try to use the same package for both schematic capture and PCB routing. You will see a huge improvement in compatibility. -- WBR, Yuriy. "Resistance is futile"
From: John F on 13 Sep 2006 08:09 Yuriy K. wrote: > John F wrote: > >>> Or you'll soon learn to choose bigger processor. Trust me. :P >> No. I won't ;-p > > Well, you had you chance... I still have more than one, because there's still enough space left in code memory to control an elephant... :-) >>> You might need one more little feature, but there is no room for >>> improvement. >> >> There is always room... just port one or two things to asm and see, >> there is enough place for putting it in twice. > > Everything is already in asm. O-ops. Nothing to port and single > unused > byte in code memory. Well written asm is quite easy to port (sometimes I've done a script to do so)... That's especially true for porting accumulator machine code to another acc-machine. (8051, 80x86) >>>>>>> I write all my interrupt routines in C. >>>> Less than 5 lines of ASM is hardly ever possible. Still 80% of >>>> time >>>> is taken there. >>> Inadequate processor or overall system architecture selection. >>> Both 5 lines and 80% of time tell that. >> Well... No. Non sequitur! > > Whatever. Without detailed description it's a pointless discussion. Indeed. Processing Data in realtime using a minimum of resources. (Time/Power/Money) >>>>> Unfortunately, I did quite a lot of x51 assembly programming. >>>>> Just >>>>> did not know better at that time. Fortunately, these days were >>>>> long >>>>> ago. >>>> They are back! >>> Not for me. :) >> Then your field of operation will be very limited. > > I have more work in foreseeable future than I could handle > I would say the job security is being way too good lately. ;) That's great! But I guess there won't be any "multimillion" projects in there. This is the part where you try to save half a cent by working for a weak or two. That's where you are not able to switch the ?C just for convenience. -- Johannes You can have it: Quick, Accurate, Inexpensive. Pick two.
From: Vladimir Vassilevsky on 13 Sep 2006 10:47 John F wrote: >>Everything is already in asm. O-ops. Nothing to port and single >>unused byte in code memory. You take the code very personal, like the piece of art. The problem is nobody is going to appreciate this as the art for the sake of art :) > This is the part where you try to save half a cent by > working for a weak or two. That's where you are not able to switch the > ÞC just for convenience. So, the break even point (just to recoup the direct labor, not considering the hidden costs of the future problems and the learning in the past) will be somewhere at 1 million units produced. Will you please be so kind to give an example of such project? Vladimir Vassilevsky DSP and Mixed Signal Design Consultant http://www.abvolt.com
From: larwe on 13 Sep 2006 11:49
Vladimir Vassilevsky wrote: > So, the break even point (just to recoup the direct labor, not > considering the hidden costs of the future problems and the learning in > the past) will be somewhere at 1 million units produced. > > Will you please be so kind to give an example of such project? I am responsible for at least a couple of projects that sell more than this number of units per year. If we could save $0.25 per unit by putting in $250K of engineering and test time, we would do it without a second's hesitation. The criterion we use is that if it is expected to break even within a year, we can go ahead and do it on our own initiative. |