Slashdot Mirror

The article mostly mentions L5 and the specsheet doesn't mention L1C: https://www.broadcom.com/produ...
Of course, they will need something on L1: L1 C/A might suffice but L1C+L5 would be better, so who knows, maybe they just didn't mention L1C.

The iPhone has a BCM4774 chip to handle the GPS. It's a low power GPS chip.

There has been a specific GPS chip since the iPhone 3G.

Microsoft leads in innovation and others just follow suit.

Next you Linux folks will be unlocking secret CPUs on the Raspberry Pi.

In typically Linux fashion, Raspi was unlocked first.
http://www.broadcom.com/blog/c...

According to this teardown of the Apple AirPort Extreme A1521, it does in fact have a Broadcom BCM4360 chip rated for 1.3Gbps. So the parent's original question is a good one. Why wasn't this included in the test??

>> and deploy networking topologies such as Ethernet, with proven security.
Ethernet is already widely deployed in cars for data hungry applications ( infotainment) For other uses, ethernet is absolutely not suitable ( price, power, wiring constraints, EMC, safety, .....)

That's why they are using 2-wire ethernet.
https://www.broadcom.com/press...

I'm not sure where you got that information about Ethernet widely deployed in cars for Infotainment. If you can send me an article about that I'd really like to read it.

- The GPU documentation, but I've never seen the SGX documentation in any SoC TRM, or for any other GPU

Some documentation is available for the GPU in the Raspberry Pi.

The RaspberryPi is a pretty much never something you should actually choose. Its certainly not a $35 general purpose computer

Running great as my £25:
- Webserver
- Ftp server
- Bittorrent server

Also runs great as a external device controller. You know, robotics and real world implementation that school kids can learn stuff from?

You don't want to use a Raspberry Pi, its crappy hardware with a broken USB stack, shitty graphics stack due to Broadcom not making a driver or release the specs.

Broadcom released specs and driver source this year http://blog.broadcom.com/chip-...

Yes there are better devices out there, yes the Pi is now "slow", but its also old in a fast paced ARM race.
But the Pi is designed for exploration and learning. Which is does not only as a device, but as a foundation and community.

£25 for your kids to learn from a hands on device, count me in.

Sure it is. I don't see you bitching about your phone, pc, car, tv, microwave oven though. You do realise that after this announcement, videocore is the most open core on an ARM chip ever, right?

btw, http://www.broadcom.com/docs/support/videocore/VideoCoreIV-AG100-R.pdf here you go...hack away

Some Broadcom chips include features that seem to work like a TDR: http://ja.broadcom.com/products/BCM5397

I first noticed them in HP G4 servers, there was a diagnostic page that would test all four pairs of wires and show a graph in addition to pass/fail. It also measured the cable length, so maybe it used TDR. For some reason I seem to think they removed that feature from newer driver versions though.

Who, besides you, calls it a microcontroller? It's called an application processor, just like all other processors used in smartphones.

From http://www.broadcom.com/products/BCM2835

The BCM2835 is a cost-optimized, full HD, multimedia applications processor for advanced mobile and embedded applications that require the highest levels of multimedia performance. Designed and optimized for power efficiency, BCM2835 uses Broadcom's VideoCore® IV technology to enable applications in media playback, imaging, camcorder, streaming media, graphics and 3D gaming.

As a result, there's a whole lot of subtly different variants out there. Not all of them are 100% binary compatible with each other. I haven't been able to find out exactly which variant is used in the raspberry pi.

FTFF, The SoC is a Broadcom BCM2835. This contains an ARM1176JZFS, with floating point, running at 700Mhz, and a Videocore 4 GPU. (links mine)

So this guy is going to send a 20 inch long model boat across the Atlantic. Right.

The Liquid Robotics Wave Gliders already travel around the world's oceans autonomously. Liquid Robotics sent Wave Gliders from Hawaii to California, then up to Alaska and back. The Wave Glider looks like a surfboard, and trails an underwater "glider". As wave action moves the surfboard up and down, the gliders's spring-loaded vanes pull it forward. The glider has a powered rudder, the only moving part. The surfboard has solar panels, a computer, a GPS, a compass, and an Iridium satellite phone. Wave gliders have been through major storms without problems. Control is good enough that they generally stay within 50 meters of the programmed track. The U.S.Coast Guard classifies them as "floating debris", so they don't have to show lights. They're no more of a threat to ships than a loose surfboard.

The "Rasberry PI", after all, is simply a board which takes a quite good IC and brings out the pins to connectors. It's not like the Rasberry PI people developed the Broadcom BCM2835.

Intel was the top contributor to Linux 3.0 (by lines) (source)

IBM is in there, too at #8

Google pushed the Linux kernel and WebKit into an uncountable number of handhelds

Apple deploys Webkit, too, on a smaller number of handhelds

Amazon deploys Android, too (just without Market support), and they use Linux in their cloud offerings.

If you hate Microsoft, give in to your anger and join Oracle (there are a lot of angry JCP and OpenSolaris fans but hey, they made that Linux list, too!)

Remember those handhelds that run the Linux kernel and/or WebKit?

• Broadcom
• Atheros (are they are part of Qualcomm now? You can check out Qualcomm, part of "Qualdroid")
• Marvell

all made the top Linux contributor list, too.

I'll assume that other posters will cover the Red Hat and Novell bases.

http://www.ti.com/product/am3359 - Note the 20MB TRM that is the third PDF linked on this page. This is the sort of documentation TI provides for their products. This is the sort of documentation you need when hardware hacking.

Compare to:
http://www.broadcom.com/products/BCM2835 - nothing but a marketing blurb

Kernel source code is not sufficient documentation - especially given Broadcom's tendency to put no useful comments into their code, making it impossible to answer questions such as "why the fuck is my CPU getting a 1 Hz wakeup interrupt from this stupid BCM4330's undocumented "BT-AMP" function?"

Your "every time" must not have much data behind it. I've deployed about a thousand Linux systems. Most work just fine as far as drivers go. The main troublesome ones are obviously not supported just by browsing the hardware Linux claims to handle. Yes, 3D support and suspend/hibernate support are still on the weak side. Those things are pretty buggy on Windows too though; I've had problems with crap NVidia/ATI drivers crashing and not supporting features on every operating system.

Up until about two years ago, wireless drivers on Linux were still a sticking point for a lot of laptops. That's been sorted out since then for all of the popular chipsets. Former troublemakers like Broadcom even list drivers on their site. I've done about a dozen Linux laptop installs since I had one that didn't just work. This week for example I put Debian and Ubuntu on a Thinkpad T500, and there wasn't a single driver issue.

"Not sure where the hate is coming from here."
Where's the TRM? Where's the datasheet? That's where the hate is coming from. That and years of bad memories involving Broadcom WLAN chipsets on Linux due to lack of datasheet access (and, in general, lack of access to any documentation at all)

Compare the following webpages:
http://www.broadcom.com/products/BCM2835 - Marketing blurb, no block diagrams, no datasheets, no nothing

http://www.ti.com/product/am3358 - Block diagram, family parametric comparison, and an EXTREMELY complete datasheet

For someone like the OP who is planning on doing embedded computing (by necessity, more "low-level" than just using the device as an STB), having a processor datashete is an absolute and complete necessity

Since the BCM2835's datasheet is not available (in typical Broadcom style) - why even bother wasting board space on a GPIO/SPI/I2C header that no one who purchases the Pi is going to be able to be able to use?

I'll believe it when I see it.

http://elinux.org/RaspberryPiBoard#Hardware_Details

http://www.broadcom.com/products/Mobile-Multimedia/Mobile-Multimedia-Processors

The Broadcom BCM2835 looks like the ARM SOC used by the Raspberry Pi. There isn't a link for its data sheet that I could find. So it looks like just another closed hardware ARM design.

The BCM2835 is not on the website but the BCM2820 is, take a look at http://www.broadcom.com/products/BCM2820.

Full HD 1080p camcorder capabilities in a cell phone with significantly improved quality over current generation handsets (which generally have VGA or lower resolution camcorders)
Up to 20 megapixel digital camera with advanced features such as multiple shots per second, image stabilization, face and smile detection and panorama mode
The ability to render mobile games natively at up to 1080p resolution, which in combination with an on-board HDMI output, allows a console-quality gaming experience on large screen HDTVs
20% to 50% power reduction in comparison to the prior generation VideoCore® III multimedia processor
4 to 6 hours of 1080p video recording and 8 to 10 hours of mobile playback, with up to 16 hours of full HD playback over HDMI given sufficient handset storage

From the "VideoCore® III" page:

Support for 8 mega pixel camera modules enables a picture quality superior to most digital still cameras, while MPEG-4 video capability at VGA resolution offers state-of-the-art video technology for tape-less camcorders. In addition, support for the H.264 video compression standard enables next-generation cellular phones to incorporate DVB-H mobile TV capability.

WiFi and bluetooth do not use the same amount of power. That is ridiculous.

Prove it.

Like I said, they re embedded in the same chipset on most smartphones.
Further, they have about the same range, and the same bandwidth.

Plus you can get 48 10Gb ports in 1U. No one else in the industry can touch that density

Not entirely true. Arista has little secret sauce. They're using merchant silicon. Stay tuned for the plethora of switches about to be released based on the Broadcom Trident ASIC (64 10 GigE switch on chip, manifesting itself as either 64 10 GigE or 48 10 GigE + 4 40 GigE). Some (like Force10's) are already announced. The differentiator in 10 GigE ToR switching is quickly becoming software.

I'm not familiar with encoding hardware for any codec. All the encoding I've ever done was with mencoder, in software. It seems there should be one soon, though.

For decoding, Broadcom has a chip for instance.

In addition to the Rockchip that was mentioned, Broadcom has an FPGA-based chip on the market that has WebM support. The other hardware companies have not announced specific product details yet.

While you're at it, any chance of releasing the source for your video decoders?

You mean like this, or something else?

Broadcom certainly stands out

Their WL Drivers are closed, but work well. I use b43, though.

This is a repeat.

Note that there is an interesting history here concerning the H.264 baseline. According to the JVT that created this baseline there was consensus to have this be "patent free" (or, more exactly, royalty free) :

Regarding Intellectual Property Rights (IPR) for the JVT codec, JVT has agreed to the following basic
principles:

  The JVT codec should have a simple royalty free “baseline” profile (both on the encoder and
decoder) in order to promote the wide implementation and use of the JVT codec. All implementations
should have such a common baseline profile core, in order to allow minimal interoperability among all
JVT codecs. The above requirement means that all technology applied in the baseline profile shall
have no IPR, expired IPR, or valid but royalty-fee-free IPR (according to Box 2.1 or 2.2.1 of the JVT
Patent Disclosure form, as shown below).

  Special, more advanced profiles of the JVT standard may contain patents per Box 2.2 of the JVT
Patent Disclosure form (reasonable terms and conditions).

So, how (besides chutzpah) does MPEG-LA assert licensing rights for the H.264 baseline ? I don't know, but I have heard rumors that the San Diego Qualcomm case had something to do with it :

Qualcomm's Patents Rendered Unenforceable and Qualcomm Ordered to Pay Broadcom's Attorneys Fees, Expenses and Costs

IRVINE, Calif., Aug 07, 2007 -- Broadcom Corporation (Nasdaq: BRCM), a global leader in semiconductors for wired and wireless communications, today announced that a San Diego federal court ruled yesterday that Qualcomm Incorporated (Nasdaq: QCOM) engaged in aggravated litigation misconduct and standards abuse with respect to two Qualcomm patents that relate to digital video technology. The court ruled that Qualcomm has thereby waived its rights to enforce all claims of the two patents and any continuations, continuations-in-part, divisions, reissues, or any other derivatives of those patents. The court also ordered Qualcomm to pay all of Broadcom's reasonable attorneys' fees, court costs, expert witness fees, travel expenses and any other litigation costs reasonably incurred by Broadcom in defending the patent infringement case that led to the rulings.

Citing the misconduct of Qualcomm's employees, witnesses, and counsel before, during and after trial, the court found that "Broadcom proved this to be an exceptional case by clear and convincing evidence based on (1) Qualcomm's bad faith participation in the H.264 standard-setting body, the Joint Video Team (JVT); and (2) the litigation misconduct of Qualcomm through its employees, hired outside witnesses, and trial counsel during discovery, motions practice, trial and post-trial proceedings." According to the court, "Qualcomm closely monitored and participated in the development of the H.264 standard, all the while concealing the existence of at least two patents it believed were likely to be essential to the practice of the standard, until after the development was completed and the standard was published internationally. Then, without any prior letter, email, telephone call, or even a smoke signal, let alone attempt to license Broadcom, Qualcomm filed the instant lawsuit against Broadcom for infringement of the '104 and '767 patents."

This experience seems to have left a bad taste in the mouth of the video standards industry concerning royalty free patents. This is shear speculation, but I also have to wonder if this has something to do with MPEG-LA's reluctance to date to charge royalties for the H.264 baseline.

Were I to build an Atom box for the TV, I would personally go with a D510 board dual core model and a Broadcom PCIe 2d video decoder BCM70012found on Ebay for cheap. This will use very little power, be very quiet, and just as good and just as good at playing (non-3D) video as the super-hot (temperature-wise) nVidia ion systems. This is, of course, you are not using the mini PCIe slot for wireless networking. Ethernet or USB wireless ftw.

These little WRT's and such have the equivalent of 8-bit 200 Mhz CPU's.

They have what is a 32-bit 200 MHz processor. Specifically this one in the referenced Linksys model.

Maybe because this chip really does not support N?

Just because it is from that family of chips doesn't mean it has N.

Look at the designator on the end of the chipset full designation: BCM4329FKUBG That BG at the end may be telling.

[...]

the full BCM4329FKUBG designation does not appear in Broadcom's catalog. Its a custom chip.

Besides, N requires special antennas. They were not found.

The trailing characters following a family name (BCM 4329) usually specify package type and other options. Full documentation for this kind (i.e., something that contains radios dealing with new standards) of chip is only provided to users who are going to buy at least a few thousand, but I'm sure someone on /. can find a more complete PDF than this. Maybe you're right about the "BG" or maybe not. However...

What I found more interesting is that this family of chips contains an FM transmitter. Bye-bye aftermarket add-ons for transmitting to your car stereo, eh?

I am sure some old hardware does exist that still doesn't have 64bit drivers for Vista/2008, but you really really need to try to actually find such hardware.

Well, you don't need to try too hard if you have older Dell or HP equipment. See what I mean?

You do realise that your own link contains download links for the 64bit Vista and Win2008 drivers?

Oh Cmon, Give him/her a break, (s)he just suffers from this ---> http://linux.slashdot.org/story/09/07/25/1757253/Linus-Calls-MicrosoftHatred-a-Disease

Evidence? http://slashdot.org/~morgan_greywolf/journal/219467

http://slashdot.org/~morgan_greywolf/journal/226315

I am sure some old hardware does exist that still doesn't have 64bit drivers for Vista/2008, but you really really need to try to actually find such hardware.

Well, you don't need to try too hard if you have older Dell or HP equipment. See what I mean?

You do realise that your own link contains download links for the 64bit Vista and Win2008 drivers?

I am sure some old hardware does exist that still doesn't have 64bit drivers for Vista/2008, but you really really need to try to actually find such hardware.

Well, you don't need to try too hard if you have older Dell or HP equipment. See what I mean?

in-expensive cabling? you are going to compare the HDMI monstrosities with UTP, and claim they are cheap? on what planet is a 20-pin connector with shielding cheaper than UTP? Switching methods? show me an HDMI switch with 8 ports for 70$. I can buy those at any corner computer store... I can wire my entire house with them. What's the cable length limit on HDMI?

Gaming consoles already have hardware in there to do all sorts of graphics operations in hardware. MPEG encoding is right up their alley.

Cameras, etc... Already have real-time encoders, because that is how they store movies on their SD cards or whatever media. Even professionals are
doing this sort of thing:
http://www.advanceddigital.ca/products/dvb/FlashXDR_encoder_recorder.php

The cable boxes are what I was thinking might require trans-coding, purely for DRM reasons. Cableco's might want to be able to use their own special coding on their cables, so a settop box would have to de-code the other format, and reencode into MPEG.

The only case where you have a point is disk players. Frankly, the customized menus annoy the hell out of me, and I would be thrilled if the disk players could agree to some XML protocol that the TV would interpret to provide disk menus in a standardized way.

Barring that, sure, the disk players would need mpeg encoding hardware... It's no big deal:

http://www.broadcom.com/products/Cable/MPEG-2-Digital-Audio-Video-Encoders

All the devices providing this information either will receive it MPEG encoded to start with (they they can just pass it on) or need fancy hardware with cpu's and memory to do what they do anyways. This is a well trodden path, nothing new required in terms of chips. Volumes will only bring the price down over time.

HDMI is an expression of people going down a path without considering the bigger picture of what the capabilities of the devices being inter-connected are.

Things could be way simpler.

http://www.broadcom.com/support/ethernet_nic/netxtremeii.php

I expect that Broadcom will update their STA binary drivers for Linux to support this chip. They already support 4321 and 4322 802.11a/b/g/n products, so this 4329 can't be that different - probably the same wifi core on an integrated package on a smaller process...

Yes, I know, Broadcom have Linux drivers for some of their products! Shocking, eh? I know it is binary only (apart from an open-source kernel adaptor layer) but it's a move forward for Linux on the desktop.

http://www.broadcom.com/support/802.11/linux_sta.php

I meant to include this link: http://www.broadcom.com/support/802.11/linux_sta.php

From the block diagram, it looks like there are three distinct RF front-ends.

Neither the article, nor Broadcom's product page, nor the product brochure pdf mention the package size. Any guesses?

I suppose it is probably a smaller footprint than three discrete radio chips put together. One usually gets better die-level integration than board level, and you can usually eliminate redundant functions that way.

Even if it were larger footprint, the fact that you could address and power just one chip rather than three would be a winning advantage on its own.

Neither the article, nor Broadcom's product page, nor the product brochure pdf mention the package size. Any guesses?

I suppose it is probably a smaller footprint than three discrete radio chips put together. One usually gets better die-level integration than board level, and you can usually eliminate redundant functions that way.

Even if it were larger footprint, the fact that you could address and power just one chip rather than three would be a winning advantage on its own.

So broadcom documentation describes a chip with a lot of unused pins, yet we find chips broadcasting clock signals down these pins. To make things interesting, it's only on the chips we receive from Broadcom. From another lab/project, these same pins are dead. I'm pretty sure Broadcom is acting like Monsanto and enforcing their draconian NDA by watching the customers developers. If they suspect they are releasing even the slightest bit of information to the public, they turn ">around and sue that company.

Yes, Broadcom has a stranghold.. but they're cheap.

There's an orbital deathstar-like supership known only as Broadcom. Complete with storm troopers ready to quash any attempts at some process called "reverse engineering."

There is no way that an open-source GPU driver can ever achieve same quality (in terms of performance and compatibility at a given timeframe) as an actively developed proprietary driver. GPUs are very complex devices, and drivers make a huge difference on performance. In order to make a fast driver you really need to know how the target GPU works on a very low level.

Naturally, NVidia will not disclose this low level stuff about their GPUs to outsiders.

You can compare GPU drivers to compilers. There is no way that open-source compiler (GCC) will ever produce as good code for new Intel's CPUs as Intel's own compiler (ICC) as long as ICC is actively developed.

show ver

show controller <blah>

http://www.broadcom.com/products/Cellular/Mobile-M ultimedia-Processors/BCM2722

The Broadcom MP4 decoder they use has a programmable decoder built into it, one of the abilities being you can throw any kind of picture subsampling you like. Apple probably have some stock IP bilinear filter for it. At 320x240, on a 2" screen, you won't be able to tell the difference between that and anything higher quality.

it is quad band and based on this: http://www.broadcom.com/products/Cellular/HSDPA-WC DMA-EDGE-GPRS-GSM-Baseband-Processors/BCM2121 the processor is the same as the treo 650 but unlike the treo, it's got wifi. 312 mhz sounds a little underpowered for encrypted voip but it's probably doable.

Intel LaGrande aims to 'protect' every IO path inside your computer, but this is still a work in progress - first TPM on every computer, the rest will be added piece by piece until the puzzle is complete.

Gigabit ethernet controller with built-in TPM (http://www.broadcom.com/press/release.php?id=7005 09/):

"Broadcom® Controllers Integrate TPM 1.2, Enabling OEMs to Offer Hardware-Based Security as a Standard Feature on All PCs
Platforms With TPM 1.2 Hardware Will Be Ready for Enhanced Security Functionality in the Next Microsoft OS (Code Name Longhorn) Expected to Ship in 2006 Breaking the Adoption Cost Barrier, Broadcom Has Integrated TPM 1.2 Functionality in Its Latest NetXtreme® Gigabit Ethernet Controller, Which Will Be Demonstrated This Week at the Windows Hardware Engineering Conference 2005"

You might already have it and not know it (the above link is almost one year old).

Your PDA/Mobile device/... will be next (http://www.arm.com/news/8308.html/):

"NDS Announces Availability Of Mobile DRM Application Based On ARM TrustZone Technology
NDS implements the first OMAv2 DRM solution leveraging the ARM TrustZone Software API which together delivers interoperable security and reduced porting costs"

Having had to (unnecessarily as it turned out - maybe I should RTFM sometimes) look at compatability between my cheapish Fujitsu-Siemens lappy (64 bit chipset so let's only support WinXPHome - way to go guys) and Airport Express, I discovered that the Airport architecture is based on Broadcom chipsets so drivers shouldn't be impossible. If the worst comes to the worst, a cheap PCMCIA card or, perish the thought, a USB dongle, will do the job.

GSM is a psychoacoustic(sp?) audio compression model

I believe the GSM compression is just a linear-predictive coding model. It's band-limited (low pass), but that really doesn't qualify as a psycho-acoustic model like music CODECs are. Here's a link to a summary of audio CODECs for telephony: http://www.broadcom.com/products/software/mobmm_au diocodecs.php

Dogs have wider recognizable audio spectrum, focus on other frequencies than humans, and generally catch different features of the voice than humans . . . GSM is just a random noise.

Are you sure about this? If you did your thesis on canine audio perception or something like that, I would gladly defer, but that just doesn't sound right.

For instance, I've dealt with humans of varying hearing acuity. Those humans with a wider frequency range of hearing tend to find it easier to understand speech, whether in person or over the phone. Additionally, those humans with "better" hearing also tended to find it easier to deal with impairments and limitiations in the channel than those whose hearing was more limited.

I know we're talking different species here, but human experience suggests that wider range of hearing implies improved understanding of speech.

Can anyone offer a citation or some expert knowlege on this subject?