They should make the firmware open source and hope the community fixes their mess. It is probably the only way this issue looks like it will be resolved if they do not fix it themselves.
If it were up to me, I would just write drivers for Linux and replace their embedded stuff with embedded Linux. The debugging tools for Linux are at least decent.
Um, it *is* Linux, most of the code is on Sourceforge, and I've been digging through it quite a bit. One of these days I'm going to try to recompile the kernel with a few tweaks, I just haven't had time.
Their drivers are likely closed source. I have never heard of DOCSIS hardware with open source drivers. I was also rather certain that the SB6190 was closed source from the lack of third party firmware based on an open source release for it,
In the ever popular show, "Is It Intel Or Broadcom" we have 3 new contenders who even Jeff Baumgartner decided to not call out, when you KNOW he knows its important info. »www.multichannel.com/new ··· s/416661
It question will be,,,, can i add any of these to my bad modem list EVEN BEFORE PRODUCTION BEGINS !!
In yet another installment of IS IT INTEL...
Tricky... Arris appears to have become a bit reluctant to share internal pics ? They seem to be a bit chip shy...
There is a stunning key bit of info missing from your story.
Intel or Broadcom based ?
We hear that Arris is laywered up, MSOs are having strained relationships and there is yet to be a fix deployed / tested for the long list of Puma issues, like the DoS that effects Puma 7. So who is inside the boxes you mention is really interesting info and its missing from your story.
Also its interesting for the first time Arris has not published internal pics of the device with the FCC and, obviously, set them as confidential to hide whats inside.
I think any new D 3.1 device should be called out for whats inside at this point.
Im hoping to figure that out shortly so I can maybe add one, or more, of these to my badmodem.com list even before production
I think everybody making modems does not want to admit to having Intel inside. It appears that they are going to lengths to keep people from finding out. This is pointless of course cuz someone will get one and crack it open and post pics. OR we could just run the DoS test against it and see what happens. If it DoSs,,, we have a Puma 7. Assuming by the time these ship we still dont have a fix..
i think I put Jeff in a bad spot. He has to keep relationships with all these guys. He knows what the deal is with this issue. Im sure he did not say because, i would bet, they wont tell him ? I betcha..
So.... I cant find info on any of these devices.. As they are cable modems and must have either a Broadcom or Intel inside, I need to know whats in these boxes for the list..
_ _ WELL FOLKS,, THIS IS QUITE THE INSTALLMENT OF OUR CONTINUING SHOW "WHATS INSIDE"
Unlike previous episodes where the FCC made this easy, the details of whats inside have been carefully hidden by the manufacturers in a vain attempt to hide this from us all. WILL THEY SUCCEED ? of course not, we will figure it out..
In the ever popular show, "Is It Intel Or Broadcom" we have 3 new contenders who even Jeff Baumgartner decided to not call out, when you KNOW he knows its important info. »www.multichannel.com/new ··· s/416661
It question will be,,,, can i add any of these to my bad modem list EVEN BEFORE PRODUCTION BEGINS !!
In yet another installment of IS IT INTEL...
Tricky... Arris appears to have become a bit reluctant to share internal pics ? They seem to be a bit chip shy...
And Jeff not mentioning Intel/Broadcom..
Hmmm............
Well, as far as the ARRIS TG3452 is concerned, I think this line on the datasheet for it strongly indicates that it uses the Intel Puma 7 D3.1 chipset:
In the ever popular show, "Is It Intel Or Broadcom" we have 3 new contenders who even Jeff Baumgartner decided to not call out, when you KNOW he knows its important info. »www.multichannel.com/new ··· s/416661
Tricky... Arris appears to have become a bit reluctant to share internal pics ? They seem to be a bit chip shy...
There is a stunning key bit of info missing from your story.
Intel or Broadcom based ? : I think any new D 3.1 device should be called out for whats inside at this point.
Im hoping to figure that out shortly so I can maybe add one, or more, of these to my badmodem.com list even before production
Well it looks like Jeff Baumgartner saw and answered your email because he added this update to the article at 8:06 PM ET tonight:
quote:Update: A reader inquired about which vendor is supplying the silicon in Arris's new model, pointing to recent complaints about the performance of DOCSIS 3.0 cable modem products with Intel Puma 6 made by suppliers that included Arris and Netgear that has since been dismissed »www.multichannel.com/new ··· d/415289 . Arris declined to identify the chipmaker for the new TG3452, adding that it "works with a variety of partners" and that the company typically does not "disclose our technology providers or our plans or agreements with them." Arris also did not disclose the chipmaker for its first D3.1 model, a retail device called the SB8200, though it's believed to be powered by Broadcom silicon.
Well it looks like Jeff Baumgartner saw and answered your email because he added this update to the article at 8:06 PM ET tonight:
quote:Update: A reader inquired about which vendor is supplying the silicon in Arris's new model, pointing to recent complaints about the performance of DOCSIS 3.0 cable modem products with Intel Puma 6 made by suppliers that included Arris ... . Arris declined to identify the chipmaker for the new TG3452, adding that it "works with a variety of partners" and that the company typically does not "disclose our technology providers or our plans or agreements with them."
Hmmmm...... I guess that's why it appears that the TG3452 does not sport one of those "Intel Inside" decals/stickers that they used to have on their earlier modems and gateways!
Well it looks like Jeff Baumgartner saw and answered your email because he added this update to the article at 8:06 PM ET tonight:
quote:Update: A reader inquired about which vendor is supplying the silicon in Arris's new model, pointing to recent complaints about the performance of DOCSIS 3.0 cable modem products with Intel Puma 6 made by suppliers that included Arris ... . Arris declined to identify the chipmaker for the new TG3452, adding that it "works with a variety of partners" and that the company typically does not "disclose our technology providers or our plans or agreements with them."
Hmmmm...... I guess that's why it appears that the TG3452 does not sport one of those "Intel Inside" decals/stickers that they used to have on their earlier modems and gateways!
They even hid the internal pics from us on the FCC documentation.
I think this might be a Intel requirement rather then a Arris decision. The overreaching legal agreement Arris had to sign to use the Intel chips might include provisions that allow Intel to suppress disclosing anything to the public including what chip a modem is made with.
I think Broadcom needs to start pushing to get logos on the box..
while it may seem im anti-Intel this is not true
YES im unhappy with Intel. Dismayed. Disillusioned. Perplexed. Pissed... BUT.. The industry needs Intel.. We NEED 2 players in the market. Having a Atom in the box brings possible powerful innovation and disruptive change in the market. Nothing wrong with all that as long as it does not involve anti-trust actions and brings higher performance devices.
Im just having my doubts right now. I still see Pumageddon and so far no way out of that pumapocalyptic future thats a mix of THX1138 and Mel Gibson.
Once we have a real fix being deployed that fixes the Puma 7, then i will be Pro-Intel. Well we need Puma 6/5 fixes too of course.
_________________________________________________________________ On a different note.. _________________________________________________________________
Ive been looking at the published Puma code.. A few things strike me. A huge chunk of it is TI and Puma 5 ( or earlier ) era. SOme of the Arris edits are interesting. Remember the code is from the 6190 D firmware..
The attached is typical. Not sure if this section tells anything too important. But its very interesting for a flavor of whats going on in how the code is written and who did what where.
They even hid the internal pics from us on the FCC documentation.
I think this might be a Intel requirement rather then a Arris decision. The overreaching legal agreement Arris had to sign to use the Intel chips might include provisions that allow Intel to suppress disclosing anything to the public including what chip a modem is made with.
I think Broadcom needs to start pushing to get logos on the box..
[att=1]
while it may seem im anti-Intel this is not true
YES im unhappy with Intel. Dismayed. Disillusioned. Perplexed. Pissed... BUT.. The industry needs Intel.. We NEED 2 players in the market. Having a Atom in the box brings possible powerful innovation and disruptive change in the market. Nothing wrong with all that as long as it does not involve anti-trust actions and brings higher performance devices.
Im just having my doubts right now. I still see Pumageddon and so far no way out of that pumapocalyptic future thats a mix of THX1138 and Mel Gibson.
Once we have a real fix being deployed that fixes the Puma 7, then i will be Pro-Intel. Well we need Puma 6/5 fixes too of course.
we don't need Intel for cable equipment chips, there are many other chip makers with better track records!.
we don't need Intel for cable equipment chips, there are many other chip makers with better track records!.
There are 2. Broadcom and Intel. Because cable modem chips are very analog and there are lots of patents there will likely never be anyone else. Intel bought the cable modem chip from TI because they could not make one on their own.
No new chip maker is gonna try and make cable modem chips.
My current phone. No joke. Its SOOOOooo much better then a cell phone.. WAY clearer, easier to hold and chat on. Awesome ringer. Should last at least another 50 years.
Not the right number in the dial. It just came with the phone so I left it.
I love POTS and will grant ya the second two points, but the first is no longer true, not with the advent of AMR-WB (aka HD Voice). It’s a night and day difference between the two for voice calls.
I love POTS and will grant ya the second two points, but the first is no longer true, not with the advent of AMR-WB (aka HD Voice). Itâs a night and day difference between the two for voice calls.
Hmmmm.... A analog only POTS connection has its merits. But those dont exist anymore.
As far as digital goes, HD Voice uses a lot of compression tricks and lots of, well, math. Also, a cell phone just cant have a good earpiece speaker. Ive noticed i get some actual bass thru the ear piece if im talking to someone else with a WE 500 or WE 2500.
I do a lot of work with super high end audio reproduction and at the super high end nothing beats Vinyl and tubes. Even digital at extreme resolutions and formats like octuple DSD with sampling rates of 22.5792 MHz »en.wikipedia.org/wiki/Di ··· rate_DSD cant match a analog recording done with tubes back in the day and mastered 100% using tubes and tape onto Vinyl.
Digital is really only good for convince and cost savings. Kinda like film. 70mm film still out performs any form of digital video with 70mm running about 12k-18k resolution.
Perceptual coding is a evil bastard, not a feature. If a phone did 20 Mhz sampling without compression and kept that on decoding and reproduction it would be closer to a analog link. However this bitrate is no where in our future with phone calls.
Hmmmm.... A analog only POTS connection has its merits. [...] I do a lot of work with super high end audio reproduction and at the super high end nothing beats Vinyl and tubes.
I'd say that's a matter of the distortions and limitations you're _used to_ hearing, not of correct reproduction of the signal.
It's more obvious in visuals: If you emulate an old arcade video game, you have to add filtering and scan lines to reproduce the output of old CRTs - not because it's accurate, but rather because it's familiar and what those games were _designed_ for. An accurate reproduction of the video looks terrible, but only because it exhibits how poor detail there was in computer graphics back in the days...
Another example: Old TV series remastered in HD. It's so ridiculous when you see all too obviously when a stunt man is filling in for the actor. The output of NTSC/PAL TVs just wasn't good enough for that detail - or you might say "it was designed for no better than that". If they hadn't redone all the space scenes in Star Trek TOS, they would've been terrible, when HD allows you to all-too-painfully see that there's just a miniature plastic space ship being moved in front of a painted golf ball...