Future of uCs

This form is for general electronics discussions, but, not PSoC.

Moderator: ericb

Forum rules
As you would expect spam, vulgar language, or discussions not related to electronics are banned from this forum.

Future of uCs

Postby toalan2001 » Mon Sep 28, 2009 8:10 pm

The status quo for a long time has been; fast core + lots of digital peripherials + minimal Analog (mostly limited to ADC). ARM, AVR, PIC, etcc all share that same successful approach. I remember a decade ago when the in vogue thing was FPGA with soft processor, but that did not have much of an impact on uCs. I also recalled excitement over uCs with integrated CPLDs, but again that did not have much of an impact.

PSOC is a different approach that has the ability to effect the future of the uC industry. Customizability of Analog and digital peripherials, but with some rails to guide the user. I can not say the Cypress has managed to find the perfect balance between ease of use and customizability, but they are getting there. With the PSOC3/PSOC5 the future is very very bright.

I also noticed another radically different trend. I was looking at propeller and XMOS uCs, they have no/minimal analog functionality and no/minimal digital peripherials. What they do have in abundance is cores, the idea is to have lots of cores available to emulate peripherials using firmware. This allows for customization, if you do not need many peripherials then you can dedicate more cores to crunching numbers and other tasks. It seems to me that these multicore uCs are easier to use than PSOC/uC + FPGA/ FPGA approaches. I like this approach very much since it leverages scales of economies behind it, billions each year are already spent shrinking transistors and ramming more cores into smaller silicon.

So that is my opinion of the future of uCs, either PSOC approach or multicore with no/minimal digital/analog circuitry. What is your opinion?
toalan2001
The Big Cheese
The Big Cheese
 
Posts: 334
Joined: Thu May 24, 2007 7:01 am

Re: Future of uCs

Postby jmg » Fri Jan 15, 2010 1:09 am

With the latest uC like the NXP M0 coming for easily sub $1, the key factor in the SoC equation is PRICE.
FPGA's struggle for growth, because as soon as a market matures, and the specs firm up, and the volumes grow,
poof, a focused silicon solution is released.

So the success of PSoC is going to be very price dependent, and history has not been kind.

The Atmel FPSlic pretty much sunk, because the price adders pushed it outside the mainstream
(the very small code memory choice did not help either.. )

Triscend went the same way...

Has anyone seen volume prices on the PsoC3 yet ?
jmg
The Big Cheese
The Big Cheese
 
Posts: 173
Joined: Mon Sep 14, 2009 6:42 pm

Re: Future of uCs

Postby toalan2001 » Fri Jan 15, 2010 11:24 pm

jmg wrote:With the latest uC like the NXP M0 coming for easily sub $1, the key factor in the SoC equation is PRICE.
FPGA's struggle for growth, because as soon as a market matures, and the specs firm up, and the volumes grow,
poof, a focused silicon solution is released.

So the success of PSoC is going to be very price dependent, and history has not been kind.

The Atmel FPSlic pretty much sunk, because the price adders pushed it outside the mainstream
(the very small code memory choice did not help either.. )

Triscend went the same way...

Has anyone seen volume prices on the PsoC3 yet ?


Yes, the M0 will have a big impact. It is a perfect example how mature the standard uC approach is.

PSOC is still in its infancy so it can not offer the same MIPS/$ as mainstream uCs.

Price of the PSOC itself is expensive, but it has the tendency to reduce the overall system costs by integrating external ICs into programmable blocks. But that can only be said for projects that make good use of the available blocks on the PSOC uC. Personally in my project, I am able to build a system much cheaper than my competitors while offering more features, better performance, and smaller size. I never felt that the price of the PSOC silicon was the big hurdle, it was always the long development time and high cost of tools that was the killer.

PSOC3 tools in terms of tool price and ease of use is much superior than PSOC1. I would hazard to say that if PSOC3/5 is 2x the price of a typical uC with the same MIPS/$, it would see better success than PSOC1. If it can drop well below the 2x price, then cypress has a sure fire hit.
toalan2001
The Big Cheese
The Big Cheese
 
Posts: 334
Joined: Thu May 24, 2007 7:01 am

Re: Future of uCs

Postby DavidCary » Sat Sep 10, 2011 2:57 pm

What would you do with 100 op-amps and a microcontroller on a single chip?

I've heard such a chip has recently been released: a chip with "programmable analog with a microcontroller ... the VCA-6 has more than 100 op-amps with each op-amp surrounded by configurable collections of resistors, capacitors, transistors, and switches."

Peter Clarke. "Triad offers 1 million gates plus analog"
EE Times -- EDN, July 25, 2011 http://www.edn.com/article/518896-Triad ... analog.php
DavidCary
Cheese Wheel
Cheese Wheel
 
Posts: 73
Joined: Sat Jun 17, 2006 8:51 pm
Website: http://opencircuits.com/User:DavidCary
Location: TULSA OK

Re: Future of uCs

Postby moxbox » Fri Sep 16, 2011 2:36 pm

I think the general discussion here has been pretty accurate,
it is largely going to be driven by price and features.

The Cortex-M0 with a small programmable mixed signal device is a potent combination.
Companies like Silego are making that possible with their GreenPak/GreenSak devices.

PSOC solutions are competive where you can integrate a lot of discrete/glue logic,
and they do well at that. However if you just need a dirt cheap micro,
they really don't compete well there, and they don't try to.

Personally, I would love to see Cypress do a "mixed signal peripheral" that
could be used with any microprocessor. But I doubt they'll go that direction.

BTW, Trisend didn't fail - they were too successful at stealing market share,
and were bought - then killed off...

;-) Tom
If we knew what we were doing, it wouldn't be called research, would it?
-- Albert Einstein
moxbox
Cheese Cube
Cheese Cube
 
Posts: 41
Joined: Thu Oct 15, 2009 9:59 am
Website: http://www.moxon.com
Location: Portland, OR

Re: Future of uCs

Postby DavidCary » Thu Jun 13, 2013 6:55 pm

toalan2001 wrote:I also recalled excitement over uCs with integrated CPLDs, but again that did not have much of an impact.

I think it's too early to rule out uCs with integrated CPLDs.

I see that yet another manufacturer has recently (*) introduced a microcontroller with some configurable logic.
They call this logic "configurable logic cell" (CLC) -- as far as I can tell, each CLC looks more or less equivalent to a CPLD "macrocell".
Still not quite as flexible as a PSoC.

(*) its datasheet mentions "2011-2013", which seems recent to me. overview.
DavidCary
Cheese Wheel
Cheese Wheel
 
Posts: 73
Joined: Sat Jun 17, 2006 8:51 pm
Website: http://opencircuits.com/User:DavidCary
Location: TULSA OK


Return to “%s” General Electronics Discussion

Who is online

Users browsing this forum: No registered users and 1 guest