LPC1000 roadmap

lpcware · ‎06-15-2016

Content originally posted in LPCWare by NXP_Europe on Tue Apr 13 01:24:19 MST 2010
Hi All,
In the forum there are quite some questions on the LPC1000 roadmap. So an official statement of NXP then.
We are working very hard in extending the LPC1000 roadmap, and you'll see the portfolio growing very rapidly.
A quick overview to expect the coming months / quarters:

Cortex-M0:

[LIST]
[*] LPC1100 with lower power enhancements, so we are low but going even lower!!
[*]LPC1100 with CAN
[*]LPC1100 with USB
[*]And expect much more in the future!!
[/LIST]
Cortex-M3:

[LIST]
[*]LPC1300 with lower power enhancements, as with the LPC1100 we'll decrease the power consumption even more!
[*]LPC1700 with performance enhancements, like e.g. the LPC1769 running at 120MHz!!
[*]LPC1700 with LCD and EMI
[*]And expect much more in the future!!
[/LIST]
Cortex-M4:

[LIST]
[*]We're working very very hard getting the first Cortex-M4 device ready, as you all might know we announced to be the first licensee of this core on Feb 22nd. So bare a bit with us getting it on silicon!!
[*]Very nice products will come in this area!!
[/LIST]
So as you see, we'll come with much more in the Cortex-M family.

We really don't forget you guys, and certainly appreciate all inputs given. So as soon as there are announcements to make, it will be placed in this thread!

lpcware · ‎06-15-2016

Content originally posted in LPCWare by Chucky on Wed Sep 29 09:25:58 MST 2010

Quote: domen
Do you have anything tangible to support your claims?
Or does one example of longer asm code automatically make it worse for you?

Hi Domen,,

I have played around a bit and found out, that almost all of my 32-bit ARM code assembles to Cortex-M3 without any problems, including conditional execution, barrel shifter and use of all registers.

I was not aware of that as I started with Cortex M0, where only fractions of my old ARM code worked.

So I have to limit my criticism and initial performance comparison to Cortex-M0.

Wow, and I can spare the extra 50 cents to upgrade from LPC11xx to LPC13xx and use all of my existing code.

Sometimes it really feels good to have been wrong, what a delight ...

Chucky :)

lpcware · ‎06-15-2016

Content originally posted in LPCWare by domen on Tue Sep 28 23:10:40 MST 2010
Do you have anything tangible to support your claims?
Or does one example of longer asm code automatically make it worse for you?

lpcware · ‎06-15-2016

Content originally posted in LPCWare by Chucky on Tue Sep 28 15:23:59 MST 2010

Quote: TheFallGuy
You might think that, but that is not how it turns out in practice. At least according to:
http://www.arm.com/files/pdf/IntroToCortex-M3.pdf
It makes interesting reading

This paper is a nice example about marketing MIPS.

However, see my example -> Cortex-M3 -> only half the speed compared to real ARM.

Yes, they are cheap, and hey ... I appreciate it, but the way they are marketed ... I think it is misleading ... when it comes down to business a real ARM (7TDMI) stomps all over ...

Chucky :)

lpcware · ‎06-15-2016

Content originally posted in LPCWare by TheFallGuy on Tue Sep 28 11:50:09 MST 2010

Quote: Chucky

It works on Cortex-M3, but I think compared to real ARM assembler (due to insertion of it ne) it takes twice the time to execute -> only half speed ...

You might think that, but that is not how it turns out in practice. At least according to:
http://www.arm.com/files/pdf/IntroToCortex-M3.pdf
It makes interesting reading

lpcware · ‎06-15-2016

Content originally posted in LPCWare by Chucky on Tue Sep 28 08:44:40 MST 2010
Thanks for the tip.

It works on Cortex-M3, but I think compared to real ARM assembler (due to insertion of it ne) it takes twice the time to execute -> only half speed ...

Chucky

Quote: igorsk
It seems that GAS does not automatically insert IT instructions, so you just need to add "IT NE" before that instruction and it will assemble.

.thumb
.syntax unified
  it ne
  addne r1, r2, r3, lsl #12
...
arm-none-eabi-gcc.exe -c -mcpu=cortex-m3 test.s
arm-none-eabi-objdump.exe -d test.o
...
Disassembly of section .text:

00000000 <.text>:
   0:   bf18            it      ne
   2:   eb02 3103       addne.w r1, r2, r3, lsl #12

lpcware · ‎06-15-2016

Content originally posted in LPCWare by igorsk on Tue Sep 28 08:16:22 MST 2010
It seems that GAS does not automatically insert IT instructions, so you just need to add "IT NE" before that instruction and it will assemble.

.thumb
.syntax unified
  it ne
  addne r1, r2, r3, lsl #12
...
arm-none-eabi-gcc.exe -c -mcpu=cortex-m3 test.s
arm-none-eabi-objdump.exe -d test.o
...
Disassembly of section .text:

00000000 <.text>:
   0:   bf18            it      ne
   2:   eb02 3103       addne.w r1, r2, r3, lsl #12

lpcware · ‎06-15-2016

Content originally posted in LPCWare by Chucky on Tue Sep 28 07:23:03 MST 2010

Quote: TheFallGuy
I think it highly unlikely that NXP will be producing any new ARM7-based parts, as they are busy replacing them with Cortex-M based parts. The Cortex-M is smaller, faster, and better....

Perhaps you just don't understand how to program in Thumb2 (note the '2' on the end) assembler. Did you know that if you use '.syntax_unified' assembler directive, you can program in ARM assembler, and the assembler will automatically generate the correct Thumb2 instructions? This gives you access to all the registers. However, I really would recommend programming in C. When you generate Release (optimised) code, the code is really quite good, and so much faster to write.

The Cortex-M NVIC can achieve pretty much everything that you might want with a FIQ.

You mean I can write:
addne r1, r2, r3, lsl #12

and the assembler will convert that to thumb ?

That would be something ... however it does not work. Unified Assembler does not include it.

With real ARM assembler you could write the fastest and densest code, with Thumb you are back in the 90's, with C you are back in the 80's

Chucky :)

lpcware · ‎06-15-2016

Content originally posted in LPCWare by TheFallGuy on Tue Sep 28 06:06:51 MST 2010
I think it highly unlikely that NXP will be producing any new ARM7-based parts, as they are busy replacing them with Cortex-M based parts. The Cortex-M is smaller, faster, and better....

Perhaps you just don't understand how to program in Thumb2 (note the '2' on the end) assembler. Did you know that if you use '.syntax_unified' assembler directive, you can program in ARM assembler, and the assembler will automatically generate the correct Thumb2 instructions? This gives you access to all the registers. However, I really would recommend programming in C. When you generate Release (optimised) code, the code is really quite good, and so much faster to write.

The Cortex-M NVIC can achieve pretty much everything that you might want with a FIQ.

lpcware · ‎06-15-2016

Content originally posted in LPCWare by Chucky on Tue Sep 28 04:04:58 MST 2010

Quote: NXP_Europe
Hi All,
In the forum there are quite some questions on the LPC1000 roadmap. So an official statement of NXP then.
We are working very hard in extending the LPC1000 roadmap, and you'll see the portfolio growing very rapidly.
A quick overview to expect the coming months / quarters:

Cortex-M0:

[LIST]
[*] LPC1100 with lower power enhancements, so we are low but going even lower!!
[*]LPC1100 with CAN
[*]LPC1100 with USB
[*]And expect much more in the future!!
[/LIST]
Cortex-M3:

[LIST]
[*]LPC1300 with lower power enhancements, as with the LPC1100 we'll decrease the power consumption even more!
[*]LPC1700 with performance enhancements, like e.g. the LPC1769 running at 120MHz!!
[*]LPC1700 with LCD and EMI
[*]And expect much more in the future!!
[/LIST]
Cortex-M4:

[LIST]
[*]We're working very very hard getting the first Cortex-M4 device ready, as you all might know we announced to be the first licensee of this core on Feb 22nd. So bare a bit with us getting it on silicon!!
[*]Very nice products will come in this area!!
[/LIST]
So as you see, we'll come with much more in the Cortex-M family.

We really don't forget you guys, and certainly appreciate all inputs given. So as soon as there are announcements to make, it will be placed in this thread!

I would like to see small chips like the LPC11xx LPC13xx but with ARM7TDMI core.

The Thumb language and limited registers, lack of register mapping (FIQs) on the Cortex-M devices suck so much ...

Chucky :)

lpcware · ‎06-15-2016

Content originally posted in LPCWare by domen on Wed Sep 22 23:15:15 MST 2010

Quote: NXP_Europe

[LIST]
[*]LPC1100 with USB
[/LIST]

Any more info on this one?
Didn't see it mentioned anywhere else than here :(

lpcware · ‎06-15-2016

Content originally posted in LPCWare by igorsk on Tue Sep 21 03:33:06 MST 2010
Actualluy, LPC1800 series was announced just yesterday. And there are hints about upcoming Cortex-M4 parts.

lpcware · ‎06-15-2016

Content originally posted in LPCWare by mik84 on Tue Sep 21 03:16:16 MST 2010
What happened to this thread. The last time I looked at it was a sticky, and now it's gone... And where are the promised new microcontrollers? :)
http://www.arrow-israel.co.il/_Uploads/125NXP_M3_Roadmap.pdf

lpcware · ‎06-15-2016

Content originally posted in LPCWare by domen on Sat May 29 13:09:44 MST 2010
tqfp48 with CAN and USB would be nice... (like ST's stm32f103)

Edit: while it has both peripherals, you can't use them both at the same time, for that you need 105 or 107.

lpcware · ‎06-15-2016

Content originally posted in LPCWare by mik84 on Fri May 28 03:50:45 MST 2010

Quote: NXP_Europe
Hi mik84,

Can tell you that we'll have 64-pins M3 devices soon.
We'll do some announcements soon on this, we'll place it on this thread!

Kind regards,

I definetly need CAN bus. Will these 64 pin devices have CAN bus?

lpcware · ‎06-15-2016

Content originally posted in LPCWare by mik84 on Fri May 28 02:19:59 MST 2010

Quote: NXP_Europe
Hi mik84,
If it just was that easy..
I won't keep you from taking a LQFP80 and cut the pins you don't need. This will result in a very low-pin count device ;):rolleyes:.. just kidding..
Can tell you that we'll have 64-pins M3 devices soon.
We'll do some announcements soon on this, we'll place it on this thread!

Kind regards,

Thanks for the answer! Does "soon" mean, that it will be available for mass production within 6 months?
Another question: Can I expect that LPC17xx with lower pincount (64 TQFP,48TQFP) will be almost identical in terms of peripherials with the existing 100 and 80pin devices? (So no new feature added, or whatsoever)

And does NXP have the plan to fix the existing errata on the LPC17xx series?
I'm asking this questioin basically only out of curiosity, because the current errata list is the shortest what I have ever seen. Did NXP really put so much effort to deliver perfect silicon?

lpcware · ‎06-15-2016

Content originally posted in LPCWare by LarryP on Thu May 27 03:58:26 MST 2010
Greetings all,

The following is cross-posted from http://tech.groups.yahoo.com/group/lpc1000/
regarding the (alleged, since I have no direct knowledge) LPC177x series, with external memory access.

Given that this is an *official* NXP website, on the topic of NXP's cortex offerings/roadmap, it seems to me that the NXP folks reading could perhaps keep us more timely and directly apprised of upcoming products, rather than requiring random users to hunt for the relevant news and post it. While we're on this subject, is there any other news you can share with us?
:confused:

==============================================================

1.
Cortex-M3 with external memory
Posted by: "Pawel" [email]elektrycerz@yahoo.com[/email] elektrycerz
Wed May 26, 2010 10:09 pm (PDT)

Re: Cortex-M3 with external memory

Hello Jean-Jacques,

NXP is currently sampling the LPC177x series to lead customers. It is
pin-compatible with the LPC246x and LPC245x series and features a 32-bit or
16-bit parallel external memory controller. It will be on distributor shelves
at the end of September 2010.

It will have all of the features of the LPC176x series, but add/improve features
like EEPROM, faster ADCs, large amounts of RAM and more. Stay tuned!

--- In [email]lpc2000@yahoogroups.com[/email], "jdauchot" <tycho@...> wrote:
>
> Hi
>
> Does anyone knows of a Cortex-M3 CPU that has EMC for access of 1MBytes of
SRAM or more?
>
> Will NXP produce a versuon of LC17XX with EMC?
>
> Regards
>
> Jean-Jacques
>
Reposted

lpcware · ‎06-15-2016

Content originally posted in LPCWare by igorsk on Wed May 26 12:55:18 MST 2010
Posted by "nxp_marketing_usa" (I wish these people used actual human names) to the LPC2000 group:

Quote:
NXP is currently sampling the LPC177x series to lead customers. It is pin-compatible with the LPC246x and LPC245x series and features a 32-bit or 16-bit parallel external memory controller. It will be on distributor shelves at the end of September 2010.

It will have all of the features of the LPC176x series, but add/improve features like EEPROM, faster ADCs, large amounts of RAM and more. Stay tuned!

lpcware · ‎06-15-2016

Content originally posted in LPCWare by NXP_Europe on Wed May 26 01:14:13 MST 2010
Hi All,
I agree with David, at the end you'll save costs using a much smaller LQFP, HVQFN or even the CSP package in the final product.

Quote: LPCVictoria
I think a converter to dip would only be for bread boarding a circuit. Anything after that should be on a surface mount pcb. They are inexpensive enough to do test runs even for a few hundred units in a final run.
David

Though we recognize this request.
Some counter arguments of doing DIP packages.

[LIST]
[*] The DIP package is a really large amount of plastic, which needs to be payed. So are you willing to pay x time the device cost for prototyping?
[/LIST]

[LIST]
[*] It will be very expensive since only low volumes are requested if I read the forum correctly. The competitors runs in high volumes which make their production cost much and therefore also the device cost lower.
[/LIST]

[LIST]
[*] All the plastic affects (negatively) the analog performance.
[/LIST]

[LIST]
[*] And the one I'm most in favor of, it is just not environmental friendly anymore creating such large plastic packages .. It's 2010, the 'green era' has started!!
[/LIST]
So my message to all of you, BE GREEN, use the LPC1100!! ;)

Kind regards,

lpcware · ‎06-15-2016

Content originally posted in LPCWare by NXP_Europe on Wed May 26 00:50:50 MST 2010

Quote: mik84
I think also, that the LPC17xx family is just begging for LQFP64 and LQFP48 packages. Just cut off some pin, and put it on the market. Thats all, job done. :)

Hi mik84,
If it just was that easy..
I won't keep you from taking a LQFP80 and cut the pins you don't need. This will result in a very low-pin count device ;):rolleyes:.. just kidding..
Can tell you that we'll have 64-pins M3 devices soon.
We'll do some announcements soon on this, we'll place it on this thread!

Kind regards,

lpcware · ‎06-15-2016

Content originally posted in LPCWare by NXP_Europe on Wed May 26 00:41:20 MST 2010

Quote: jmg
speaking of announcements, this one too :)

http://www.nuvoton.com/NuvotonMOSS/Community/ProductInfo.aspx?tp_GUID=6d78fb24-4f43-4806-9318-ebf3be...

http://www.nuvoton.com/hq/enu/ProductAndSales/ProductLines/ConsumerElectronicsIC/ARMMicrocontroller/...

Bigger family of Cortex M0's 32KF/4KR to 128KF/16KR
LQFP48/64/100 and also now QFN36

but their PLLs are not _nearly_ as smart as the LPC17xx one.

Hi jmg,
Obviously we don't like that you guys advertise competitor products on our forum, please do that elsewhere. Though this time I appreciate it since you recommend our product.

Kind regards,