Talk:TI MSP430

Latest comment: 1 year ago by 84.140.192.133 in topic Number of transistors, die size?

Platform, Generation, Family nomenclature

edit

In TI's documentation, the MSP430 devices are all part of a common platform. Each of the MSP430X1XX, MSP430X2XX, etc are members of a generation, and there are families within each generation. I updated the "families" section to instead describe "generations" but the entire article needs to be updated. As is, family is used in multiple contexts. It is used to describe the entire MSP430 platform, and for generations. This should be rectified, but I don't have the time for it now. shoez 17:56, 20 April 2009 (UTC) —Preceding unsigned comment added by Shoez (talkcontribs)

Note for details see TI doc SLAB0340 page 3 --shoez 17:59, 20 April 2009 (UTC) —Preceding unsigned comment added by Shoez (talkcontribs)

Donnelt (talk) 06:59, 4 April 2019 (UTC)== Marketing, not encyclopedic ==Reply

In my view this whole entrey the way it is now is one giant marketing placement for TI, and should be drastically shortened to just a few lines.66.176.20.143 01:52, 31 January 2007 (UTC)Reply

I agree, marked the article with advert template. --Alsh 22:58, 3 February 2007 (UTC)Reply
How does it compare to the other microcontroller pages (Microchip's PIC, ATMEL, etc.)? --132.205.139.54 (talk) 15:56, 30 April 2008 (UTC)Reply
From the looks of it, it seems far less detailed than other microcontroller pages. I don't think it's anymore marketing than the PIC or AVR pages. 192.91.75.29 (talk) 21:32, 3 June 2009 (UTC)Reply

This is information about a microcontroller, if you shorten it to only a few lines then it will be useless as a reference.. —Preceding unsigned comment added by 24.68.154.165 (talk) 01:48, 13 October 2007 (UTC)Reply

Agree. I work for a competitor, and this page is full of useful information about the part. Can someone tell me which piece of information they think is useless? shoez 00:14, 17 April 2009 (UTC) —Preceding unsigned comment added by Shoez (talkcontribs)

Blatent marketing. Extolling 1uA idle current when AVRs will run on 100nA. Lacks objectivity. How does this pass muster?

Age?

edit

Does anyone know how old the MSP430 is (when it was invented)?

Well, approximately I do: I was a student of electronics in Munich, Germany, when I heared of a micro power controller being developed by TI in Freising. This was in 1993 or 94 and it turned out to be the MSP430 family. —Preceding unsigned comment added by 88.64.69.17 (talk) 12:58, 7 October 2008 (UTC)Reply

Hardware Development Tools

edit

I removed this section as it all seemed to be links to sales sites, which fail WP:EL. If you know of sites about the development tools that aren't primarily intended to sell them, then feel free to add it back with those links. Mark Grant 18:24, 26 June 2007 (UTC)Reply

  • But with respect - that's silly? Isn't the point of the external links section to point to resources that can be used that are outside of that which Wikipedia will enclose on the page. Surely if some of those links point to physical resources it is inevitable that they will have to be sold? Nothing is free

By removing this section you seem to be removing a valuable resource, with real world applications. What if people reading this article want to experiment?

  • And BTW the WP:EL does not forbid linking to sites, it just asks people to consider whether they are useful and relevant to the article. All of the sites in the harware section you unilaterally removed were!
It says such sites should generally be avoided. There's a difference between, say, linking to a site with a review of a book and linking to an amazon.com page selling the book. Or between linking an Intel page about one of their CPUs and linking to a site which sells Intel CPUs. Why do you think that sales sites should be included? Mark Grant 17:22, 28 June 2007 (UTC)Reply
  • What good is a book review, unless one can read the book? My point is that vendors fill a vital place in society, without which, virtually none of the products and services discussed on Wikipedia would exist. They offer the service of supplying things that are otherwise just concepts. Your front page features a piece on the Wonderbra currently. The piece rightly links to the Wonderbra site, where it offers the viewer the chance to find sales outlets. There is little practical information contained on the site, but it should be linked to because it offers content outside the scope of Wikipedia. Vendors do that too. Why should they not be allowed? Again - WP:EL does not explicitly say that they are not allowed.
Wikipedia is not an advertising site. It's not a site for people to find places to buy things from. If this was an article _about_ those specific hardware development tools then adding a link to the official site would be quite legitimate even if it was primarily intended to sell them, as allowed by WP:EL. If there's a site listing available development hardware then I'd say that linking to it would be a good idea, but linking to sites which just offer them for sale is not.
To reiterate, WP:EL states: 'Except for a link to a page that is the subject of the article or an official page of the article subject—and not prohibited by restrictions on linking—one should avoid'. So linking to the official Wonderbra site in an article about Wonderbras is perfectly legitimate. Linking to a site that sells PCs in an article about Pentium CPUs would not be.
More to the point, the reason why I removed that link in the first place was because someone was posting links to the same site on multiple articles. That may have been done with good intentions, but it looks very much like advertising spam and not an attempt to improve articles. Mark Grant 11:56, 1 July 2007 (UTC)Reply

CISC !

edit

It's a full cisc design comming from pdp-11. It's not a load&store architecture. It's not a fixed size instruction set (2 16 bits adresse could be embedded at the end of each instructions). There is absolutly nothing "riscy" here. —Preceding unsigned comment added by 213.30.139.86 (talk) 13:38, 4 September 2007 (UTC)Reply


PDP-11

edit

Do we really need to compare it to the DEC PDP-11 in the first paragraph? The two processors have very different applications, and their use do not overlap in time.

--132.205.139.54 (talk) 15:54, 30 April 2008 (UTC)Reply

MSPGCC - last chance for sourced criticism

edit

I've switched a cite tag to OR since it seems more appropriate. That line seems to have a number of issues - it speculates on a possible source (because a older version is used). This is OR and not particularly valid - It uses the GCC frontend but like all GCC compilers the backend is largely architecture-specific. Added to that, even the suggestion that gcc is inferior in the first place is entirely unsourced. Since that has been tagged since Feb 2007, if no reliable sources can be found for the assertions they are probably ready to be trimmed altogether. We are not in the business of pushing trade towards commerical companies based on unsourced speculation. CrispMuncher (talk) 16:06, 5 November 2009 (UTC)Reply

Well, I've taken it out in the absence of any information. CrispMuncher (talk) 21:02, 6 November 2009 (UTC)Reply

Plucking the peacock

edit

The current version of this article includes a pretty blatant marketing dump into the "generations" section. Lots of WP:PEACOCK words. E.g. all the adjectives in "a wide range of high-performance analog and intelligent digital peripherals". It needs a major plucking, but I'd like to discuss here if anyone thinks any of the information there is too marketing-oriented, or it's just the phrasing. (Yes, I'll do it, as soon as I accumulate more round tuits.) 71.41.210.146 (talk) 02:25, 20 January 2011 (UTC)Reply

MIPS at how many Mhz?

edit

for the MSP430x1xx the article talks about 8 MIPS, but at what clock speed? — Preceding unsigned comment added by Mihaigalos (talkcontribs) 12:32, 15 February 2011 (UTC)Reply

Presumably 8 MHz, the maximum clock speed of the '1xx family. (1xx, 3xx and 4xx go to 8 MHz, '2xx to 16 MHz, and '5xx to 25 MHz, all reducing at supply voltages below 3.3V.) Like the 6502, the processor is almost perfectly memory bound. The number of cycles an instruction takes is the sum of the number of instruction fetches, data fetches, and data stores required to execute it. For register-to-register operations, instructions are a single word and so can be executed 1/cycle. 71.41.210.146 (talk) 12:23, 19 February 2011 (UTC)Reply

"Guaranteed" writes to FRAM

edit

"FRAM is also capable of zero power state retention in all power modes, which means that writes are guaranteed..."

This does not follow. FRAM maintains its contents with a loss of power but that does not mean that writes are necessarily guaranteed. If power down occurs during a write the result is not guaranteed unless brown-out detection circuitry is used. — Preceding unsigned comment added by (talkcontribs) 13:30, 18 July 2012 (UTC)Reply

A recent document from TI (SLAA559 – October 2012) states that "The write-back mechanism is also protected from power loss and is guaranteed to complete safely under all power-fail events" ([1] - paragraph 2.2) — Preceding unsigned comment added by 82.61.86.28 (talk) 15:29, 2 December 2012 (UTC)Reply

Reading FRAM is like writing to core storage. Writeback is a special situation for ferromagnetic memory. In ferromagnetic memory reads are destructive. All the bits are shifted to a standard state and the flipping of the bit which induces a magnetically driven voltage pulse is what allows circuitry to discern the word content. Since read is destructive there needs to be a 'writeback' phase. That must be protected from power loss, since you dont want memory to be lost if there is a power loss during a read. So the referenced material about 'writeback' is in reference to reads. It does not directly address writes. However, I'd imagine that since a read/writeback cycle is longer than a write cycle that write integrity during power loss is likely an expected behavior. That would just be at the word level, though. Writing a series of values to memory probably wouldn't be protected. I would think that a programmer would have to include a checksum into multiword writes to detect incomplete transfers that are to blocks of memory. It is amusing to ferromagnetic storage again. Back in the 1970's core storage was the state of the art. First time being an old fart has given me some computer cred! ;-) Steve Kd4ttc (talk) 02:01, 19 July 2013 (UTC)Reply

References

edit

This processor was mentioned in podcast Security Now, episode 380, at 55 min 51 secs (MP3, 45 MB). Perhaps it can be used as a source? --Mortense (talk) 20:03, 30 November 2012 (UTC)Reply

Article title

edit

My understanding of naming conventions is that "Texas Instruments MSP430" should be written out in full, if we include the company name at all. Should the article be moved accordingly? --SoledadKabocha (talk) 03:29, 13 December 2013 (UTC)Reply

43oh URL broken

edit

For the record, the 43oh URL is broken today, it gives 403 Forbidden. So does the "Contact us" link on http://forum.43oh.com/. Sd4dfg2 (talk) 23:35, 7 July 2014 (UTC)Reply

Number of transistors, die size?

edit

How many transistors do the MSP430 chip models have? What is the die area size? 84.140.192.133 (talk) 12:46, 10 April 2023 (UTC)Reply