>> 15years ??
Viewing the future as in comparison to the past is worthless. Everything is
about to change. The next 25 years will be nothing like the past 25. So
worrying about your original MCU being around 15 years from now misses the
point entirely. Boards can be rev'd. Designs change. It's really no
big
deal. Plan to adapt in the future to survive.
We have no idea what semi companies will even be around 5 years from now,
much less 15 years. Massive contraction is underway. I talked to several
semi companys and their production is currently running about 50%. Many
fabs will be closing. Some companies will go under or merge. There is no
way to predict what parts will be around in the future. That is about the
only thing you can plan on with certainty.
Chris.
LPC2138
Started by ●December 17, 2008
Reply by ●December 17, 20082008-12-17
Reply by ●December 17, 20082008-12-17
On Wednesday 17 December 2008 16:12, s...@gmail.com wrote:
> Hi,
> Is it a wise decision to use LPC2138 for new design? I need to
> choose a device, which will have supply/support for another 15
> years.. Will NXP supply/support LPC2138 for another 15 years. Or
> will they vanish soon. Any known issues with LPC21238. Flash data
> retention for 15 years?
In fifteen yrs you could proly roll your whole board into a fpga.
--
Rgds
JTD
> Hi,
> Is it a wise decision to use LPC2138 for new design? I need to
> choose a device, which will have supply/support for another 15
> years.. Will NXP supply/support LPC2138 for another 15 years. Or
> will they vanish soon. Any known issues with LPC21238. Flash data
> retention for 15 years?
In fifteen yrs you could proly roll your whole board into a fpga.
--
Rgds
JTD
Reply by ●December 17, 20082008-12-17
Hi,
> We have no idea what semi companies will even be around 5 years from now,
> much less 15 years. Massive contraction is underway. I talked to several
> semi companys and their production is currently running about 50%. Many
> fabs will be closing. Some companies will go under or merge. There is no
> way to predict what parts will be around in the future. That is about the
> only thing you can plan on with certainty.
I agree. But the 8051 is the cockroach of the semiconductor industry.
--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
CrossWorks for ARM, MSP430, AVR, MAXQ, and now Cortex-M3 processors
> We have no idea what semi companies will even be around 5 years from now,
> much less 15 years. Massive contraction is underway. I talked to several
> semi companys and their production is currently running about 50%. Many
> fabs will be closing. Some companies will go under or merge. There is no
> way to predict what parts will be around in the future. That is about the
> only thing you can plan on with certainty.
I agree. But the 8051 is the cockroach of the semiconductor industry.
--
Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
CrossWorks for ARM, MSP430, AVR, MAXQ, and now Cortex-M3 processors
Reply by ●December 17, 20082008-12-17
I am still amazed that the 8051 is still alive and working so well
considering that its new variants has still has 256 bytes of internal
ram which includes the stack pointer.
I am still supporting code for this processor
Keil and iar C compilers are something else
Regards
Jean-Jacques
--- In l..., "Paul Curtis" wrote:
>
> Hi,
>
> > We have no idea what semi companies will even be around 5 years
from now,
> > much less 15 years. Massive contraction is underway. I talked
to several
> > semi companys and their production is currently running about
50%. Many
> > fabs will be closing. Some companies will go under or merge.
There is no
> > way to predict what parts will be around in the future. That is
about the
> > only thing you can plan on with certainty.
>
> I agree. But the 8051 is the cockroach of the semiconductor
industry.
>
> --
> Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
> CrossWorks for ARM, MSP430, AVR, MAXQ, and now Cortex-M3 processors
>
considering that its new variants has still has 256 bytes of internal
ram which includes the stack pointer.
I am still supporting code for this processor
Keil and iar C compilers are something else
Regards
Jean-Jacques
--- In l..., "Paul Curtis" wrote:
>
> Hi,
>
> > We have no idea what semi companies will even be around 5 years
from now,
> > much less 15 years. Massive contraction is underway. I talked
to several
> > semi companys and their production is currently running about
50%. Many
> > fabs will be closing. Some companies will go under or merge.
There is no
> > way to predict what parts will be around in the future. That is
about the
> > only thing you can plan on with certainty.
>
> I agree. But the 8051 is the cockroach of the semiconductor
industry.
>
> --
> Paul Curtis, Rowley Associates Ltd http://www.rowley.co.uk
> CrossWorks for ARM, MSP430, AVR, MAXQ, and now Cortex-M3 processors
>
Reply by ●December 17, 20082008-12-17
leon Heller wrote:
> I think that PICs have been around almost as long. GI originally designed it
> as a peripheral controller for their rather nice16-bit processor.
Yeah but you can't get an original PIC1650 anymore. Not that you'd
want to get one, really. The oldest type they still make is probably
the PIC16C54 (with no revision letters added) and that's late 1980's
vintage. No idea about the exact year though.
I'm just about to make a fresh batch of boards that were designed in
1993, pretty much exactly 15 years ago, and they have a Philips 80C552
processor in PLCC68 case, and that exact same part is still available
from NXP. But that has to be the weird exception because it's a 8051
after all.
If nothing goes seriously wrong, I'd expect processors like the LPC2138
last for about 5-10 years. I've had some nasty experiences with
Atmel and Freescale discontinuing parts in that age range, and having
no 100% replacement for them...
> I think that PICs have been around almost as long. GI originally designed it
> as a peripheral controller for their rather nice16-bit processor.
Yeah but you can't get an original PIC1650 anymore. Not that you'd
want to get one, really. The oldest type they still make is probably
the PIC16C54 (with no revision letters added) and that's late 1980's
vintage. No idea about the exact year though.
I'm just about to make a fresh batch of boards that were designed in
1993, pretty much exactly 15 years ago, and they have a Philips 80C552
processor in PLCC68 case, and that exact same part is still available
from NXP. But that has to be the weird exception because it's a 8051
after all.
If nothing goes seriously wrong, I'd expect processors like the LPC2138
last for about 5-10 years. I've had some nasty experiences with
Atmel and Freescale discontinuing parts in that age range, and having
no 100% replacement for them...
Reply by ●December 17, 20082008-12-17
Paul Curtis schrieb:
> I'd say 8051s will be around in 15 years. Those bloody things won't die.
I'd wonder what its size would be in 32nm technique :-)
--
42Bastian
Note: SPAM-only account, direct mail to bs42@...
> I'd say 8051s will be around in 15 years. Those bloody things won't die.
I'd wonder what its size would be in 32nm technique :-)
--
42Bastian
Note: SPAM-only account, direct mail to bs42@...
Reply by ●December 17, 20082008-12-17
On Wed, 17 Dec 2008 14:42:03 +0100, you wrote:
>Paul Curtis schrieb:
>
>> I'd say 8051s will be around in 15 years. Those bloody things won't die.
>
>I'd wonder what its size would be in 32nm technique :-)
..doubt you'd get a 5V version in 32nm though...
This is a nice graphic illustration of shrinking chip sizes - same EPROM 12 years on :
http://www.electricstuff.co.uk/shrinkingchip.html
>Paul Curtis schrieb:
>
>> I'd say 8051s will be around in 15 years. Those bloody things won't die.
>
>I'd wonder what its size would be in 32nm technique :-)
..doubt you'd get a 5V version in 32nm though...
This is a nice graphic illustration of shrinking chip sizes - same EPROM 12 years on :
http://www.electricstuff.co.uk/shrinkingchip.html
Reply by ●December 17, 20082008-12-17
----- Original Message -----
From: "Sam Laur"
To:
Sent: Wednesday, December 17, 2008 1:04 PM
Subject: Re: [lpc2000] LPC2138
> leon Heller wrote:
>> I think that PICs have been around almost as long. GI originally designed
>> it
>> as a peripheral controller for their rather nice16-bit processor.
>
> Yeah but you can't get an original PIC1650 anymore. Not that you'd
> want to get one, really. The oldest type they still make is probably
> the PIC16C54 (with no revision letters added) and that's late 1980's
> vintage. No idea about the exact year though.
>
> I'm just about to make a fresh batch of boards that were designed in
> 1993, pretty much exactly 15 years ago, and they have a Philips 80C552
> processor in PLCC68 case, and that exact same part is still available
> from NXP. But that has to be the weird exception because it's a 8051
> after all.
>
> If nothing goes seriously wrong, I'd expect processors like the LPC2138
> last for about 5-10 years. I've had some nasty experiences with
> Atmel and Freescale discontinuing parts in that age range, and having
> no 100% replacement for them...
Atmel does have a habit of doing that, several of the older AVRs are
unobtainable. One of the nice things about Microchip is that they seem to
have a policy of keeping their older chips in production.
Leon
From: "Sam Laur"
To:
Sent: Wednesday, December 17, 2008 1:04 PM
Subject: Re: [lpc2000] LPC2138
> leon Heller wrote:
>> I think that PICs have been around almost as long. GI originally designed
>> it
>> as a peripheral controller for their rather nice16-bit processor.
>
> Yeah but you can't get an original PIC1650 anymore. Not that you'd
> want to get one, really. The oldest type they still make is probably
> the PIC16C54 (with no revision letters added) and that's late 1980's
> vintage. No idea about the exact year though.
>
> I'm just about to make a fresh batch of boards that were designed in
> 1993, pretty much exactly 15 years ago, and they have a Philips 80C552
> processor in PLCC68 case, and that exact same part is still available
> from NXP. But that has to be the weird exception because it's a 8051
> after all.
>
> If nothing goes seriously wrong, I'd expect processors like the LPC2138
> last for about 5-10 years. I've had some nasty experiences with
> Atmel and Freescale discontinuing parts in that age range, and having
> no 100% replacement for them...
Atmel does have a habit of doing that, several of the older AVRs are
unobtainable. One of the nice things about Microchip is that they seem to
have a policy of keeping their older chips in production.
Leon
Reply by ●December 17, 20082008-12-17
----- Original Message -----
From: "42Bastian"
To:
Sent: Wednesday, December 17, 2008 1:42 PM
Subject: Re: [lpc2000] LPC2138
> Paul Curtis schrieb:
>
>> I'd say 8051s will be around in 15 years. Those bloody things won't die.
>
> I'd wonder what its size would be in 32nm technique :-)
The very small size is one reason why it is still very popular. It's often
combined with something else on the same chip, like a wireless transceiver.
Leon
From: "42Bastian"
To:
Sent: Wednesday, December 17, 2008 1:42 PM
Subject: Re: [lpc2000] LPC2138
> Paul Curtis schrieb:
>
>> I'd say 8051s will be around in 15 years. Those bloody things won't die.
>
> I'd wonder what its size would be in 32nm technique :-)
The very small size is one reason why it is still very popular. It's often
combined with something else on the same chip, like a wireless transceiver.
Leon
Reply by ●December 17, 20082008-12-17
Mike Harrison schrieb:
> On Wed, 17 Dec 2008 14:42:03 +0100, you wrote:
>
>> Paul Curtis schrieb:
>>
>>> I'd say 8051s will be around in 15 years. Those bloody things won't die.
>> I'd wonder what its size would be in 32nm technique :-)
>
> ..doubt you'd get a 5V version in 32nm though...
Yes. Maybe the 5V tolerant pad-logic would eat up the gain in size.
> http://www.electricstuff.co.uk/shrinkingchip.html
Amazing. See, the limiting factor seems to be the bonding.
--
42Bastian
Note: SPAM-only account, direct mail to bs42@...
> On Wed, 17 Dec 2008 14:42:03 +0100, you wrote:
>
>> Paul Curtis schrieb:
>>
>>> I'd say 8051s will be around in 15 years. Those bloody things won't die.
>> I'd wonder what its size would be in 32nm technique :-)
>
> ..doubt you'd get a 5V version in 32nm though...
Yes. Maybe the 5V tolerant pad-logic would eat up the gain in size.
> http://www.electricstuff.co.uk/shrinkingchip.html
Amazing. See, the limiting factor seems to be the bonding.
--
42Bastian
Note: SPAM-only account, direct mail to bs42@...
