PICvsAVRS Again(SORRY!)

[Wond3rboy]

i know this question has been asked again and again but i want to ask with respect to performance that are PICs beter then AVRs?Cause AVRs are leass expensive then PICs generally and if everybody is going for PICs then there has to be a good reason so i want to ask that reason.I mean if for e.g i am an AVR user then is there any reason for me to shift to PICs or vice versa or choose to to either of them if i am a newbie.Thank you

 

[oomchu]

hmm..being a newbie to this myself, I can't give very many technical specifications why people go with PICs. I can say, microchip is one of the larger manufacturers of microcontrollers and from what I understand there's a lot more software support and resources in general for the PICs. Those might be reasons enough for a lot of people to start using PICs vs AVRs. Also, you can get free samples from microchip of certain PICs. At least you could until they suspended it.

 

[colin mac]

**broken link removed**

 

[philba]

Soon, it could all be moot - they will both be Microchip products. That is if Microchip keeps AVRs around.

I have to admit, I don't see the sense in the acquisition but haven't dug all that deep.

 

[Papabravo]

Soon, it could all be moot - they will both be Microchip products. That is if Microchip keeps AVRs around.

I have to admit, I don't see the sense in the acquisition but haven't dug all that deep.

Buying a fab is way cheaper than building a new one, especially when the acquiree's stock is worth a fraction of its former self. At $5.00 a share you can buy General Motors for USD 4 Billion. I can't imagine why anyone would want to, but there it is. It used to be about 50 Billion or so give or take a few pennies.

 

[Wond3rboy]

Buying a fab is way cheaper than building a new one, especially when the acquiree's stock is worth a fraction of its former self. At $5.00 a share you can buy General Motors for USD 4 Billion. I can't imagine why anyone would want to, but there it is. It used to be about 50 Billion or so give or take a few pennies.

Pennies dont matter! so i am to take that there is no difference b.w using a PIC or AVR except the better price given by AVRs and its availability.Although i am going for PICs this vacation i intend to do AVRs also by using one of their ISP Programmers. they are easy to build and simple!

 

[Nigel Goodwin]

Pennies dont matter! so i am to take that there is no difference b.w using a PIC or AVR except the better price given by AVRs and its availability.Although i am going for PICs this vacation i intend to do AVRs also by using one of their ISP Programmers. they are easy to build and simple!

They are different devices, both will do the job - but PIC's are generally considered easier to learn, and are far more popular with much more support on the net.

AVR's are cheaper because PIC's are the market leader, Atmel are trying to compete, and have to price lower accordingly.

From recent reports it looks like Atmel and the AVR haven't been doing too well, it will be a shame is MicroChip take them over though

 

[Sceadwian]

That would be a crime against humanity unless microchip took the good from the AVR line and applied it to the PICs. PIC wins in the public eye due to market penetration, no other reason. AVR's and PIC's are equally easy to learn how to program period, AVR's have their strong points and PIC's have their strong points as well, but support, refrence, or development kits for either line is not lacking in ANY way shape or form. While there are statistically more pages related to pics out there again it's market penetration ONLY and has NOTHING to do with the technical aspects of the various architectures or the quality of the code and support for them. Also Atmel as a company struggling is more related to PICs massive market share and viewpoints from persons whom defacto side with PIC without considering the tehcnical aspects of both lines first, and from an industry at large which fights change with tooth and nail.

 

[CCD]

As it goes, AVRs are normally more robust than PICs! They don't easily blow up when something goes wrong in the circuit.
AVRs are certainly faster than PICs. An AVR running at 16MHz gives 16MIPS throughput while a PIC running at even 20MHz gives 5MIPS throughput! Certainly, speed counts here!!

----------------
Design with Microcontrollers

 

[3v0]

Higher and Deeper !

You can buy PICs in most any speed range. The fact that the processor takes 4 clock cycles to execute one instruction is not a drawback. Just use a 4X clock source. Need more speed move up a family or two.

I have not used AVRs but I have never noticed that PICs are not robust. In my book they are just the opposite.

3v0

As it goes, AVRs are normally more robust than PICs! They don't easily blow up when something goes wrong in the circuit.
AVRs are certainly faster than PICs. An AVR running at 16MHz gives 16MIPS throughput while a PIC running at even 20MHz gives 5MIPS throughput! Certainly, speed counts here!!

----------------
Design with Microcontrollers

 

[Ubergeek63]

You can buy PICs in most any speed range. The fact that the processor takes 4 clock cycles to execute one instruction is not a drawback. Just use a 4X clock source. Need more speed move up a family or two.

I have not used AVRs but I have never noticed that PICs are not robust. In my book they are just the opposite.

3v0

On a performance for dollar basis AVRs blow PICs out of the water and ARMS blow AVRs out of the water.

As for Microchip buying AVR the question has nothing to do with them being the same, it has to do with which lines will be dropped. AVRs might simply dissapear. Microchip will now have the AVRs and ARMs. I suspect they will drop the low end line that competes with PICs and keep the newer ARMs...while dropping the older ARM parts like the AT91R400008s.

 

[Nigel Goodwin]

I have not used AVRs but I have never noticed that PICs are not robust. In my book they are just the opposite.

Yes, I've never used AVR's either, but PIC's are incredibly robust, I've never managed to damage one during many years of use - including one which got so hot it's got my fingerprint burnt in the top! (don't put them in the wrong way round ).

 

[3v0]

I never understood the uC pissing contest. Choose whatever fits the project, if there is more then one pick the one that will allow you to get the job done the fastest.

Unless you work for yourself chances are that you will be told what uC to use. The few times I had a choice in the matter I picked the uC with the best tool-chain, primarily the C compiler. Time to market is everything and every year it seems more product content is in the software.

In general, embedded C compilers have been improving. If there are several good ones choice mostly comes down to what you are used to using.

3v0

 

[Sceadwian]

It's not so much a pissing contest as it is attempting to get facts straight. Nigel stated that PICs are easier to learn and better supported, both statements which are absolutely false. Bigger market share does not mean better supported, nor does a larger number of web pages with PIC content vs AVR as there is still more information out there for BOTH lines than anyone could possibly read. PICs and AVRs are both approximatly equal in how difficult they are to learn and there are fully developed free IDEs for both I believe, though I'm not familiar with the free software available for PICs, you can chose ASM, Basic, or C for both PIC and AVR.

Also 3V0 you can not say that the 4-1 clock rate is not a drawback. EVERY AVR even the lowest dirt cheap 30 cent tiny11 (though they're hard to find glad I stocked up) runs at 1 MIPS per MHZ. Only higher end later model PICS have PLLS, and even then you still need 4X the clock speed, which means to equal the raw processing ability of an AVR running at 20Mhz you'd have to have a PIC With a PLL that will run at 80MHZ, and you can get 20Mhz AVRs for dollars, that is a HUGE bonus for hobbyists.

Are there PICS that cost a dollar that run at 80mhz?

Though PICS having bigger market penetration can't be overlooked entirley if you intend on doing this professionally, as PICs are so pervasive you have no choice but to learn them. For a hobbyist or experimentor the practical choice is 100% in AVRs favor for price/performance. For people involved in industry though learning PICs is a requirement.

 

[3v0]

There is plenty of misinformation on all sides.

In my mind most of the uC's I use are a means to an end. (There are a few that rise above that) I agree that the 20MHz AVRs are good to have around when you need the speed at a very low price.

The processor instruction rate vrs clock speed stupidity was around prior to both AVR and Microchip. Regarding speed the ONLY things that count are how fast the chip executes instructions and what can be done with each instruction. The 4-1 clock rate is not a drawback. But people need to compare instruction rate rather then clock speed.

Having said that Microchip is footing the bill for the uC's I need to run my class. They have a few people on staff who's job it is to enable education. In my book that counts. To be fair I have not talked to other chip vendors about similar programs.

3v0

 

[AtomSoft]

You can buy PICs in most any speed range. The fact that the processor takes 4 clock cycles to execute one instruction is not a drawback. Just use a 4X clock source. Need more speed move up a family or two.

Hey 3v0 are you telling me that i can put a 80MHz crystal/oscillators on a 20Mhz pic and have 20MIPS?

 

[Ubergeek63]

Though PICS having bigger market penetration can't be overlooked entirley if you intend on doing this professionally, as PICs are so pervasive you have no choice but to learn them. For a hobbyist or experimentor the practical choice is 100% in AVRs favor for price/performance. For people involved in industry though learning PICs is a requirement.

But PICs don't have better market penetration...Microchip has good sales figures, but no matter how much Nigel would like that to be equal to better market penetration, it simply isn't.

They might be better than AVR's but they simply are not the best market penetration. Maybe 5 or 10 YEARS ago they had better yearly numbers but that was then. The market uses what makes the most sense economically in million piece volumes, and PICs simply do not make economic sense.

 

[blueroomelectronics]

But PICs don't have better market penetration...Microchip has good sales figures, but no matter how much Nigel would like that to be equal to better market penetration, it simply isn't.

They might be better than AVR's but they simply are not the best market penetration. Maybe 5 or 10 YEARS ago they had better yearly numbers but that was then. The market uses what makes the most sense economically in million piece volumes, and PICs simply do not make economic sense.

Do you have data to back up your claims? Microchip hit the 5billion PICs sold last year, they are very very cheap in quantity.
If you want some serious power far beyond any 8bit MCU Microchip offer the very fast and powerful 16bit PIC24F and the math capable dsPICs many can hit 40MIPs. Some PICs include USB & Ethernet built in.
Inexpensive tools like the PICkit2 or ICD series can program & debug can be had for less than $50 for a PICkit2

 

[blueroomelectronics]

Hey 3v0 are you telling me that i can put a 80MHz crystal/oscillators on a 20Mhz pic and have 20MIPS?

Not with the 16F or 18F, but the 16bit 24F (OSC/2) and most dsPICs go 30MIPs upto 40MIPs (PLL16x OSC / 4) so 7.5MHz crystal * 16 / 4 = 30MIPs

 

[AtomSoft]

oh with PLL ok.