Slashdot Mirror

TI's also starting up a new R&D lab in honor of the occasion. The sent out an email inviting employees to apply for positions, but I'm just a product engineer so I don't get to do research. :-(

The DSP instruction set is documented.

http://focus.ti.com/lit/ug/spru732g/spru732g.pdf

The PowerVR is still unfortunate, however your choices for mobile devices are pretty much the PowerVR or Mali (and I've not yet seen a Mali device), and ARM are pretty tight fisted with documentation as well

I remember that the team from India that won the TI-DSP design contest in 2006, had come up with a similar idea (cause I knew those guys), I think they called it wibonet or somthing equally inane, they wanted to use the idea for rural connectivity IIRC.

yeah here's the link . http://tiidesigncontest.ext.ti.com/results.html, but it doesn't seem to have much in the way of information.

Who on earth needs a skillset like that?

Then, he will scoop up all of the unemployable engineers at slave labor wages, laughing the entire time while sitting atop his throne made of golden skulls. In fact, this article was probably penned while he sat atop said throne.

And he would have gotten away with it, too, if it weren't for you meddling kids.

I got the wrong link for the OMAP1030

Actually, for Edge, the OMAP1030 is the current TI solution, but it has only a single ARM9 for the radio and application processing.

Actually, for Edge, the OMAP1030 is the current TI solution, but it has only a single ARM9 for the radio and application processing.

Actually, for Edge, the OMAP1030 is the current TI solution, but it has only a single ARM9 for the radio and application processing.

Well, NEC Electronics does, and they're even 90nm! (650nm and 350nm versions also available.) Also Freescale, and Texas Instruments, and Maxim (no, not the magazine) and ... lots of people, especially those who want to have decently-performing analog circuits.

You kids these days thinking everything is CMOS. Go ahead and try to make me a 10GHz RF circuit in CMOS.

Well you're right, I was looking at an older TI catalog, but they still list budgetary pricing for this dual-phy peripheral interface at $10 in quantity.

http://focus.ti.com/docs/prod/folders/print/tsb43aa82a.html

If you're looking at an expansion card with multiple ports it probably has only 1 ASCI (if it has 4 ports or fewer).

There are those that will tell you that C and asm jobs are much more rare than C++, C# and Java. They may be right. I can tell you, as director of Firmware Engineering of the company I'm with, that it is MUCH MUCH harder to find unemployed C and asm 8/16-bit micro engineers than it is to find MSWindows, Linux, big system, C#, java and etc programmers.

My recommendation would be to go to http://www.ti.com/ez430 and for $20 buy yourself a eZ430-F2013. That comes with a MSWindows GUI demo package of a C compiler, assembler, simulator and debugger, with the real TI CPU. With this you can get your hands dirty writing firmware for your own application. It's worth it. It will really enhance your resume going into many lines of work in the embedded/firmware engineer lines of work.

This idea was actually around before the integrated circuit.
We got ICs because Jack Kilby invented the Integrated Circuit in response to a requirement to create modular electronics.
See here or here

Archos is based on a TI DualCore DaVinci http://focus.ti.com/docs/prod/folders/print/tms320dm6446.html , the mobile version. The DSP side contains codecs that are proprietary to TI. The ARM side contains Linux and a bunch of supporting software to read media files and software for AV Sync. More info regarding the DSP side architecture here http://wiki.davincidsp.com/index.php?title=Codec_Engine_Roles. The codecs are combined into a single DSP image, TI provides Eval codecs that can be used to hack to get more functionality. One can register on TI's website and get the concerning tools from their webiste to compile the DSP side components.

Where have they gone wrong?

All TI demonstrated in 1967 was a prototype that weighed 3 pounds. TI's own website places the introduction of their first "consumer electronics product", the TI-2500 Datamath, at 1972.
this page lists portable calculators appearing in 1970, and pocket calculators in 1971. No TI firsts there.

http://www.xbitlabs.com/news/mobile/display/200608 28031933.html

old data...Motorola is a close second. However, I now consider Motorola a global company.

Thus, more recently:

http://www.itbusinessedge.com/item/?ci=29705

You missed the point, same as the last clown. America has historical significance in the computing world, and is still considered the biggest player in the internet from several fronts. In fact, many people still bitch b/c American retains control over key core Internet routers. China and India have sheer numbers to compensate, but on a per capita basis, U.S. expenditures on computers is significanly higher than most other countries. We were here first and we've shared with the world. World, using eqully impressive brains (if not equally impressive capital) now contribute significantly. My point was, exclude the U.S. from mail routing and you'll be cutting your legs out from underneath you. Sure, you can still pull yourself around on your arms or use a wheel chair, but it wouldn't be the same.

Mobile phones have been weak in the U.S. ever since inception. This is partly b/c we have such a strong land-line infrastructure. I've heard that there is more fibre in the U.S. than in the rest of the world, but lack citation to point it out as fact. Many other countries have used mobile phones b/c that's all they have. If they tried to get copper/DSL/Cable to all the residencies, their local telcos would go bankrupt. (Ours are now offering entry level DSL for $15/month. FIOS and other fibre offerings are now bringing 15MB+ to the home for less than $100/month) Now matter how cool a cell phone gets, it won't replace the beauty of raw data througput pulling down movies/ISOs/VLF (Very Large Files) in minutes instead of hours/days that most of the rest of the world currently enjoys. Plus, why can't they figure out how to simply billing in the rest of the world. Most mobile operators I've experienced South/Central America, Caribbean, UK, and Europe all have confusing rates which are constantly changing, all varying depending on if you're calling another mobile phone, a land line, and one of 3 time options. Nevermind that the U.S. enjoys free long distance to all of its massive continent while Europeans are paying different rates to 20+ countries all in near proximity. I think I just threw up a bit in my mouth thinking about it...

But, if we must, lets look at the heart of cell phones.

Qualcomm is 90% of CDMA phones. Qualcomm is US company.
Quite a few video components rely on TI components. (Texas Instruments...wonder where that was founded?)

I'd ask you to take a look at a vendor market share report, http://www.marketresearch.com/map/prod/1334374.htm l, but that would cost money...something I know linux lovers are loathe to hand out. Here's a free one for you though:
http://focus.ti.com/general/docs/wtbu/wtbugenconte nt.tsp?contentId=4605&navigationId=12046&templateI d=6123
TI has largest market share followed by Qualcomm in 2nd place. So, Nokia's first place hold on the cell phone market is dependent on the chipsets from 2 American companies, and wireless technologies, while collaborative, require significant technical innovations.

Suck it. If it makes you feel any better, you should realize that many of the "Americans" working for our companies here are 1st generation foreign nationals who came here for the superior education and stayed on for the superior pay and lifestyle. We import your best and brightest all the time. Its the American way.

Just because our politicians blow doesn't mean the rest of our hard work goes unnoticed or unrewarded.

TI already has your system on a chip. It's called the DM6442, DaVinci.

http://www.ti.com/corp/docs/landing/ez430tool/inde x.htm?DCMP=430Day&HQS=Other+OT+ez430

No first hand experience.

Texas Instruments has $20 USB-based development boards for the MSP430 series microcontrollers. The MSP430 is a nice, clean C-friendly 16-bit architecture. It's practically luxurious, by microcontroller standards.

Although perhaps, as others have said, microcontrollers are a bit too hardcore for students. A better idea might be to use full-fledged PCs running DOS (FreeDOS if there are licensing issues), either by scavenging old laptops or by running a cable to the robot from a desktop PC. DOS is nice and simple, well-documented and well-supported with development tools.

(Disclaimer: I am employed in the creation of MSP430 development tools.)

hp49g+ - Pros:
SD card reader
RPN AND Algebraic notation
Speaker
MUCH faster than the 89Ti
Powerful CAS library
Built-in assembly

hp49g+ - Cons:
Lower resolution
Older models have worthless keyboards
Harder to use
Smaller userbase

89Ti - Pros:
Much larger screen
A truely gigantic userbase; some INSANE things have been done with this calculator, and this is where the 89 really shines:
F-Zero for the calculator: http://www.ticalc.org/archives/news/articles/13/13 3/133789.html
Headphone support: http://www.ticalc.org/archives/files/fileinfo/386/ 38629.html
Gameboy emulation: http://www.ticalc.org/archives/files/fileinfo/369/ 36950.html
Radium overclocking: http://www.ticalc.org/archives/news/articles/1/15/ 15975.html (just kidding)
Geometer's Sketchpad: http://education.ti.com/educationportal/sites/US/p roductDetail/us_sketchpad_89.html
And much more.

89Ti - Cons:
Slow 3D graphing (a program allows a TI-83+ to do this faster than TI's built-in 3D grapher for the 89, even if you take resolution into account)
CAS sometimes behaves oddly (try the cube root of -27)
Much slower than hp49g+
No RPN (third party RPN programs aren't very good)

Well, that's my take on the situation. I've used both calcs and I've tried to be unbiased, and it's really up to you to decide. I personally have an 89 because I love my virtual globe, my tesseract grapher, ELIZA, my star chart, my All Your Base Are Belong To Us screensaver, and such. At the same time, I am envious of the hp49g+'s superior speed and graphing capabilities, and it (the hp49g+) is really fun to use. However, the title of ultimate calculator would have to go to one of the following:

Qonos: http://www.hpcalc.org/qonos.php (it looks so cool!)
TI-Nspire CAS: http://education.ti.com/educationportal/sites/US/n onProductMulti/nspire_cas.html (even if it is ugly)
Voyage 400: http://www.ticalc.org/archives/news/articles/8/85/ 85069.html (I wish)

hp49g+ - Pros:
SD card reader
RPN AND Algebraic notation
Speaker
MUCH faster than the 89Ti
Powerful CAS library
Built-in assembly

hp49g+ - Cons:
Lower resolution
Older models have worthless keyboards
Harder to use
Smaller userbase

89Ti - Pros:
Much larger screen
A truely gigantic userbase; some INSANE things have been done with this calculator, and this is where the 89 really shines:
F-Zero for the calculator: http://www.ticalc.org/archives/news/articles/13/13 3/133789.html
Headphone support: http://www.ticalc.org/archives/files/fileinfo/386/ 38629.html
Gameboy emulation: http://www.ticalc.org/archives/files/fileinfo/369/ 36950.html
Radium overclocking: http://www.ticalc.org/archives/news/articles/1/15/ 15975.html (just kidding)
Geometer's Sketchpad: http://education.ti.com/educationportal/sites/US/p roductDetail/us_sketchpad_89.html
And much more.

89Ti - Cons:
Slow 3D graphing (a program allows a TI-83+ to do this faster than TI's built-in 3D grapher for the 89, even if you take resolution into account)
CAS sometimes behaves oddly (try the cube root of -27)
Much slower than hp49g+
No RPN (third party RPN programs aren't very good)

Well, that's my take on the situation. I've used both calcs and I've tried to be unbiased, and it's really up to you to decide. I personally have an 89 because I love my virtual globe, my tesseract grapher, ELIZA, my star chart, my All Your Base Are Belong To Us screensaver, and such. At the same time, I am envious of the hp49g+'s superior speed and graphing capabilities, and it (the hp49g+) is really fun to use. However, the title of ultimate calculator would have to go to one of the following:

Qonos: http://www.hpcalc.org/qonos.php (it looks so cool!)
TI-Nspire CAS: http://education.ti.com/educationportal/sites/US/n onProductMulti/nspire_cas.html (even if it is ugly)
Voyage 400: http://www.ticalc.org/archives/news/articles/8/85/ 85069.html (I wish)

Everyone assumed that Apple's $499/$599 prices for the iPhone was subsidized by Cingular.

Who thought that? At that price, it had better be profitable.

Cell phone internals are becoming very cheap. Check out the Texas Instruments "LoCosto" two-chip cell phone. Manufacturing costs are approaching $20. This isn't being reflected in the prices seen at US carrier's retail outlets, though. The handset price is inflated there, then "discounted" to compel users to sign up for "plans".

The Voyage 200 is essentially the TI-92

http://education.ti.com/voyage200

But yeah it's basically just the TI-89 with a QUERTY keyboard plus a few extra apps and a USB-to-TI cable

There is a flash app that you can buy for the TI-89 called Equation Writer that REALLY helps inputting long or complicated equations. It's only $15, and well worth it. I've been using it for about a year and a half now, and haven't gotten a math problem wrong as a result of inputting it into the calculator wrong since I got it. I bought it after getting mad in physics class. I input an equation the teacher wrote on the board TWICE and got a different answer both times. Both answers were wrong! They looked identical on the screen after being input in the pretty-print area. It was slight nuances of the parentheses that was killing me, but those kinds of errors are hard to catch on a tiny input line. Anyways, I bought it off of http://education.ti.com/ but I don't see it listed anymore. I still have it in my downloads area since I bought it, but that isn't helping me locate it. Sadly, Google isn't being much help either. It's called Equation Writer (EQW) 1.01.

I was forced to get a TI-85 in college when the calculus courses started requiring calculators as part of the ciriculum. I still have it and use it, happily to this day. The newer TIs are much better than my antiquated old thing (more memory, better connectivity, and Flash RAM for long-term storage), so I would certainly recommend the TI-89 Titanium or the Voyage 200. The feature sets and prices are pretty similar, but I would be worried that some professors or grad students might balk at allowing the Voyage 200 in an exam.

Lots of folk pooh-pooh the TI calculators as being inferior to the HP RPN models, but I have, as I said, been very happy with mine. You can write fairly sophisticated programs on them, and the newer ones even allow you to do some of the programming from a PC and transfer the program to the calculator over a USB cable. My only real gripe with the TIs (or any of the graphing calculators, for that matter) is that the displays haven't gotten much better over the past 20 years: they all, pretty much, still low resolution monochrome LCDs, many without a backlight. The processors, memory capacity and interface options have all progressed, but the displays have stagnated.

Maybe Apple can give us an iCalc to update the graphing calculator for the twenty-first century, or something (actually, it would be a pretty neat third-party add-on for the video iPod: a dock that slips around the iPod and gives you a full scientific calculator keypad, using the iPod for display and storage).

Having grown up in the era preceding the rise of the graphing calculator, however, I can say that there is a serious downside to learning higher math with a sophisticated calculational crutch: you may not get as good an instinct for the math as you would have if you had been forced to do the graphs by hand. I suppose that, in an age when even elementary schoolers are being given calculators, it's something of a lost cause to lament the loss of manual mathematical skills.

I was forced to get a TI-85 in college when the calculus courses started requiring calculators as part of the ciriculum. I still have it and use it, happily to this day. The newer TIs are much better than my antiquated old thing (more memory, better connectivity, and Flash RAM for long-term storage), so I would certainly recommend the TI-89 Titanium or the Voyage 200. The feature sets and prices are pretty similar, but I would be worried that some professors or grad students might balk at allowing the Voyage 200 in an exam.

Lots of folk pooh-pooh the TI calculators as being inferior to the HP RPN models, but I have, as I said, been very happy with mine. You can write fairly sophisticated programs on them, and the newer ones even allow you to do some of the programming from a PC and transfer the program to the calculator over a USB cable. My only real gripe with the TIs (or any of the graphing calculators, for that matter) is that the displays haven't gotten much better over the past 20 years: they all, pretty much, still low resolution monochrome LCDs, many without a backlight. The processors, memory capacity and interface options have all progressed, but the displays have stagnated.

Maybe Apple can give us an iCalc to update the graphing calculator for the twenty-first century, or something (actually, it would be a pretty neat third-party add-on for the video iPod: a dock that slips around the iPod and gives you a full scientific calculator keypad, using the iPod for display and storage).

Having grown up in the era preceding the rise of the graphing calculator, however, I can say that there is a serious downside to learning higher math with a sophisticated calculational crutch: you may not get as good an instinct for the math as you would have if you had been forced to do the graphs by hand. I suppose that, in an age when even elementary schoolers are being given calculators, it's something of a lost cause to lament the loss of manual mathematical skills.

Although, I'm a little partial being a developer for TI and working on the next generation of calculators, I would have to say the TI nspire is the next big thing. It should be out next quarter. More to come.... http://education.ti.com/educationportal/sites/US/n onProductMulti/nspire_cas.html http://en.wikipedia.org/wiki/TI-Nspire_CAS

The TI Voyage 200 http://education.ti.com/educationportal/sites/US/p roductDetail/us_v200.html is the upgrade to the TI-92 Plus. Ten years ago, the local community college used both the HP48GX and the TI-92.

Try the TI-89. Hoever, I prefer my TI-58C.

For a factor of 100 difference, I care.

I used to care about the difference between symbol rate and bit rate—heck, I am an architect at a company that made its fortunes in part through DSP-driven modems—and even my coworkers don't really care whether you refer to baud or bps, since typically the meaning is clear from context, and depending on what portion of the system you're referring to, the terms are interchangeable.

Go look at the stty man page. Go ahead, I'll wait. Now point to where it says "bps" or "bits per second." What? Can't find it there? But it says "baud?" Must've been written by non-technical people who don't know their terminology. BTW, I tried this experiment on Linux, Solaris, and for good measure, AIX. For the record, AIX does refer to "bits per second," but those intellectually lazy folk[*] at Sun refer to baud. And you can forget those Linux hippies. ;-)

I used to fight "the good fight," correcting people on the difference between bps and baud, and correcting people for pronouncing GIF with a hard G. (Look! It says right here in the Compuserve spec that it's a SOFT G!) It's a great way to alienate people, including technical types. It's one thing to use the terminology correctly yourself, and to correct someone if they're using terminology incorrectly in a context where it matters, but to do it as a matter of course just makes you a PITA.

As for the "doss" vs. "dose," I had a number of teachers and other authority figures in the 1980s insist it was pronounced "dose," correcting me when I would pronounce it "doss." I believe they didn't realize it was an American acronym, and not derived from the Spanish word for "two."

--Joe

[*] Tongue planted firmly in cheek.

I wrote a short paper concerning RFID technology about a year ago, it mostly concerned the hardware and systems architecture. There was no shortage of reports and studies of RFID keys being cracked like the mobile speedpass http://www.jhu.edu/news_info/news/home05/jan05/rfi d.html.

http://www.ti.com/rfid/shtml/news-releases-rel02-1 0-05.shtml. Some of these passive rfid tags have no access control whatsoever. Meaning one take a small RFID programmer into their favorite store and start changing prices, or worse, write a virus to the RFID tag so the next time it's polled it'll get injected into their SQL DB. Possibly compromising their entire POS system. Ironically, this sort of stunt if done well enough could result in a jackpot of creditcard numbers so it wouldn't matter if you used an RFID enabled card or not at that point :).

Some random RFID links.
http://www.schneier.com/blog/archives/2005/03/rfid _security_a.html
http://www.rfidgazette.org/2004/06/rfid_101.html
http://www.rfidjournal.com/article/articleview/133 9/2/129/
http://www.technovelgy.com/ct/Technology-Article.a sp?ArtNum=20
http://www.enigmatic-consulting.com/Communications _articles/RFID/Link_budgets.html

A nice article on RFID virus attack
http://www.cbronline.com/article_news.asp?guid=B96 0208D-9ECF-4F0B-B964-4DD779BFF905

http://www.computerworld.com/securitytopics/securi ty/story/0,10801,100459p2,00.html

From which comes a nice quote, this is from 2005.
"The TI technology is vulnerable to attack because it uses a decade-old, 40-bit cryptographic key to encrypt communications between the RFID DST tags and readers, the researchers found. TI also used an unknown and proprietary encryption algorithm on its DST devices. But Rubin's team reverse-engineered the secret algorithm by observing how DST tags responded to specially crafted challenges. Once they guessed the algorithm, researchers created a software program that could be used in so-called brute-force attacks on DST devices to recover the secret cryptographic keys, Rubin said."

The site, http://rfidanalysis.org/ that hosted these findings no longer exists but you could probably find it cached on the net somewhere, wayback machine maybe.

Remember that RFID represents a system and not one piece of technology. The implementation of the system is dependent on the deployment plan. I could make an "RFID system" with 2 933Mhz radios and a pair of 8-bit microcontrollers from digikey for around $150. Sure, you could pull my data out of the air, but technically speaking I'm using RFID. I could also build my own RFID key system with 2048-bit encryption to act as the keys to my car. It's not that difficult to develop, really just assembling existing technologies. RFID can be done "right" and it is a promising technology. I wouldn't shun it for alot of commercial applications but for personal applications, well ask yourself the question. Is this thing a necessary part of your life?

Peter

By contrast, IBM is one of the 3 remaining American companies that still makes general-purpose, complex, and powerful cores for crunching scientific applications. The other two companies are AMD and Intel.

...build your own USB "converter". Companies like Texas Instruments have lots of devices like PCM2704, that allow access to an unprotected sound bitstream. It's pretty simple to build a fake digital speaker that just redirects the data to a fake digital line in. Some microcontrolled usb sound devices contain both input and output devices on the same IC, so you can software redirect the output (coming from the computer) to the input (going back to it).

So you don't even need an "Analog hole". You can use a digital hole and don't lose any quality at all. And this kind of device is perfectly accepted by any "content protection" driver schemes.

It's impossible to protect sound files.

These little things can record the temperature and humidity that the case or pallet experienced. That alone will help Wal-Mart keep quality up, as the rfid scanners will reveal pallets or cases exposed to extremes, outside of guidelines. I like Wal-Mart, and am sorry to see negative stories coming to light about them. I do save on groceries when I go to Wal-Mart, and get fresher items also.

Somebody has to do the distribution, not everyone can chase down their own cow and milk it. For those who can't, there's Wal-Mart, with milk in gallon jugs at a reasonable price.

Buttermilk, Chocolate Milk, Skim Milk, Fake Milk made from soybeans.

How can you go to bed at night without a nice glass of milk?

Don't forget the endless selection of Potato Chips, available at Wal-Mart, too!

No. RFID is radio frequency identification, which is vague and meaningless. Some RFID tags are RO, some RW, some more complex IO. Some have crypto/hash capabilities. However, they are all RFID.

Consider the one paragraph breif on the TI RFID Compact Series Digital Signature Wedge Transponder DST+

This new generation of secure RFID transponder provides additional levels of security. In addition to the proven TI encryption known from the DST transponder, mutual authentication increases security and sophisticated diagnostic features allow fraud prevention and after-theft diagnosis. It offers 50Byte of EEPROM memory from which 26Byte are free for user data. The DST+ can be operated in DST mode in which it is functional compatible to the DST.

I see this design could have some uses with the distribution of medication and supplies in hospitals. It could allow the medical staff greater security to the possibility of someone using their password to steal narcotics, etc.

I thought the distinction between passive and active RFID was that 'active' tags had a continuous power source. Passive RFID tags get their power, typically via induction, from the reader and therefore are relatively limited in transmit power and reading distance. This does not preclude them from having receivers and being read/write.

The little glass vial RFID tags made by TI come in both Read Only and Read/Write. http://www.ti.com/rfid/shtml/prod-trans.shtml#lowf req

Of course the HP device requires contact so it's not really an RFID tag at all. :-\

... in other news scientist are using RFID embedded socks with RFID enabled dryers to solve one of the greatest mysteries of our time. Where do all the missing socks go?

It remains to be seen what effect this will have on the GStreamer community, as cheap DaVinci hardware may never find its way into hackers' hands.

What do you mean by cheap hardware? Do you mean fully designed, implemented and commercially available set top boxes, HUDs, etc. or the DaVinci DSPs themselves. The processor is tentatively priced in the $40-50 range for quantities of 1,000. For less than a few hundred bucks you should be able to design a prototype using said processor. If the DaVinci modified GStreamer source code is released (as it should according to the GPL) then it may not be all that expensive money wise for the dedicated hacker. This sounds like a great project to bring open-source hardware (i.e., opencores.org) and open-source software communities together. Of course, the time investment will be expensive, but it shouldn't be overly difficult to develop open source codecs such as OGG-Vorbis to plug in to GStreamer for the DaVinci hardware.

Even the evaluation modules and the development kit aren't too pricey for serious development. For under $15,000 you have everything needed for development. Plus, both versions of the development kit come with MontaVista Linux Professional Edition packaged with them. While that may be too expensive for the average hacker, it's not completely unfeasible to see a development based on the architecture pop up in the open community.

Having an open source starting point such as GStreamer available definitely boosts the feasibility of an open-source hardware and software Tivo-killer, DVR or other set-top box.

It remains to be seen what effect this will have on the GStreamer community, as cheap DaVinci hardware may never find its way into hackers' hands.

What do you mean by cheap hardware? Do you mean fully designed, implemented and commercially available set top boxes, HUDs, etc. or the DaVinci DSPs themselves. The processor is tentatively priced in the $40-50 range for quantities of 1,000. For less than a few hundred bucks you should be able to design a prototype using said processor. If the DaVinci modified GStreamer source code is released (as it should according to the GPL) then it may not be all that expensive money wise for the dedicated hacker. This sounds like a great project to bring open-source hardware (i.e., opencores.org) and open-source software communities together. Of course, the time investment will be expensive, but it shouldn't be overly difficult to develop open source codecs such as OGG-Vorbis to plug in to GStreamer for the DaVinci hardware.

Even the evaluation modules and the development kit aren't too pricey for serious development. For under $15,000 you have everything needed for development. Plus, both versions of the development kit come with MontaVista Linux Professional Edition packaged with them. While that may be too expensive for the average hacker, it's not completely unfeasible to see a development based on the architecture pop up in the open community.

Having an open source starting point such as GStreamer available definitely boosts the feasibility of an open-source hardware and software Tivo-killer, DVR or other set-top box.

It remains to be seen what effect this will have on the GStreamer community, as cheap DaVinci hardware may never find its way into hackers' hands.

What do you mean by cheap hardware? Do you mean fully designed, implemented and commercially available set top boxes, HUDs, etc. or the DaVinci DSPs themselves. The processor is tentatively priced in the $40-50 range for quantities of 1,000. For less than a few hundred bucks you should be able to design a prototype using said processor. If the DaVinci modified GStreamer source code is released (as it should according to the GPL) then it may not be all that expensive money wise for the dedicated hacker. This sounds like a great project to bring open-source hardware (i.e., opencores.org) and open-source software communities together. Of course, the time investment will be expensive, but it shouldn't be overly difficult to develop open source codecs such as OGG-Vorbis to plug in to GStreamer for the DaVinci hardware.

Even the evaluation modules and the development kit aren't too pricey for serious development. For under $15,000 you have everything needed for development. Plus, both versions of the development kit come with MontaVista Linux Professional Edition packaged with them. While that may be too expensive for the average hacker, it's not completely unfeasible to see a development based on the architecture pop up in the open community.

Having an open source starting point such as GStreamer available definitely boosts the feasibility of an open-source hardware and software Tivo-killer, DVR or other set-top box.

By typing http://www.isuppli.com into his Web Browser, as many posters indicated already. Some say that iSuppli does not know about specific deals on specific parts, and that's true, but you can't expect Apple to open the kimono on such sensitive, private deals. In many cases parties are not even allowed to talk about specific prices that they agreed upon (typical for MS OEM deals, as an example.) Anyway, he would be within a few percent off at most, since even a most lucrative deal can't go that far below a volume price, and that is usually well known. It's very cold, hard data. For example, see here.

He's simply saying that there isn't enough real data from apple to judge the prospects of the company.

"Since January, Mr. Renck has been advising clients against owning Apple shares." (TFA)

I don't read it as "not enough data", I read it as "sell all you have, as fast as you can." How can anyone read it differently?

Furthermore, what is unwise about it? It may be conservative, but not investing in a company is not unwise.

It may be wise and conservative only if the investor does not know who and what Apple is, and considers it a shady, fly-by-night company. But a Wall St. analyst ought to know better, and he is specifically paid to know better. But in this case he behaved like a scared cat, as if Apple directors are about to grab the cash and run to Argentina. There is no reason for alarm, aside from being alarmed of the "conservatism" (if we call it this way) of certain analysts.

But Apple's secrecy about it's roadmap does give me pause. Nobody has a clue what it up Steve's black turtleneck sleeves.

Ok, imagine two armies about to meet in a decisive battle. One is led by a general who can't stop talking, and every grunt in his army knows all the general's plans a week in advance. Another army is led by a tight-lipped general, who keeps all the strategy in his head, and in heads of his closest assistants who aren't talking either. All other things being equal, who is more likely to have an advantage?

DaVinci is Texas Instruments single chip solution for mobile phones and multimedia rich embedded devices. They mixed a TI DSP chip in with the ARM core( anyone remember OMAP ) for a high performance single chip solution. Prior to this, smartphones used one processor for the radio and one processor for the GUI/applications. The holy grail here is one processor for everything significantly reduces cost. Intel DSPs are not near as popular as TI's and so it's a no-brainer to use TI's stuff in this case.

http://hardware.newsforge.com/article.pl?sid=06/01 /05/163242&from=rss
and
http://www.ti.com/corp/docs/landing/davinci/firstp roducts.html?DCMP=DSP_DaVinciCatalog&HQS=Other+PR+ thedavincieffectpr

LoB

Many of those games were on cutting-edge equipment at the time (ok, not the 2600). They used multiple processors, custom graphics chips, and custom sound chips. Yet, we've been able to emulate them pretty well.

Even the very advanced-for-its-time TMS34010 graphics processor chip has been emulated in software to play Hard Drivin'. That chip had bit-addressable memory, built-in clipping, overflow math, and 2D fill.

In 15 years, hardware and software will advance to emulate lots of the stuff in modern video games.... all in software.
In the past, video games were closely tied to their hardware (the atari 2600 had to count its opcodes to fit the NTSC video signal). Now the interface is cleaner and independent of the hardware -- just emulate DirectX or OpenGL well and you've got a good start at emulating the whole game.

Just FYI, Sun doesn't make its CPUs. Texas Instruments does. Sun is what you call a "fabless semiconductor company" (I just learned this term this morning.)

One of the most expensive parts of a new chip design is having a set of masks made. I've heard these can cost millions of dollars on a typical VLSI design, but let's be conservative and say you can get a simple set for $50000. If you make 1000 chips, your cost per chip is $50 in masks alone. This doesn't even include the cost of silicon, fab fees, and so forth.

If you want to go cheap, several companies make low-cost microcontrollers. A quick hop over to http://www.ti.com/msp430 shows that you can get a 16MHz, 16-bit microcontroller with 128b RAM, 1K flash, analog comparator, slope AD converter, watchdog timer, 16 bit timer with 3CCR's, 16 IO pins, and brownout reset for only 90 cents each. I don't know what you want to do, but I bet you could do it with a chip like that. Plus, TI isn't the only company making stuff like this. If you shop around a little, you can probably find something for as little as 50 cents.

They're a lot different than any other architecture...

Actually, they are similar to a number of DSPs and other discrete solutions from the past. For example:

The TMS 320DM64x series of DSP from TI which has an ARM9 and a number of DSPs on it.

The TMS 320DM54x and 55x series of DSP from TI which has an ARM7 and a number of DSPs on it.

And a descrete version in the CSPI MAP 1310/11 which had a PPC and multiple multi-core DSP chips on it as early as 1997.

They're a lot different than any other architecture...

Actually, they are similar to a number of DSPs and other discrete solutions from the past. For example:

The TMS 320DM64x series of DSP from TI which has an ARM9 and a number of DSPs on it.

The TMS 320DM54x and 55x series of DSP from TI which has an ARM7 and a number of DSPs on it.

And a descrete version in the CSPI MAP 1310/11 which had a PPC and multiple multi-core DSP chips on it as early as 1997.

embedded systems need time to evolve first. right now a solution would be full custom, pricing it very high. we need some commoditazation. i'd place that around 2007, but the bandwagons going to move fast once it gets here.

look around in three years; computers will have undergone a throughput revolution. pci-expres will achieve penetration into the embedded market so there'll be a low-trace-count high-bandwidth low-voltage interconnect. the major semiconductor people just released the first round of reasonable priced off the shelf H.264 encoding (ti's da vinci for example). Current solutions needed either high bandwidth raw storage or proprietary encoders, all of which are in the 'custom' domain.

I dont know where CCD's and the op-amps stand. I'd wager they're a bit of a limiting factor. I'd be the thermal issues become problematic when you're running a CCD at over 20 Mbps of raw data. Active cooling is power intensive.

The tech's getting here...