On April 12, 2005 11:36 pm, Kevin P. Fleming wrote:
Also, keep in mind that a DS3 is _only_ 45 megabits per second. Any PCI
bus (even lowly 33MHz 32-bit PCI) can easily handle 90 megabits per
Yes, but then what are you doing with it? You're shuttling the new data
to/from a network card in a
Matt Klein wrote:
Kevin,
Mmm. Yep.
-m
On Tue, 12 Apr 2005, Kevin P. Fleming wrote:
Matthew Boehm wrote:
So, no hardware encoding on this beast?
The announcement on the website makes no mention of transcoding, echo
cancellation or toast-and-jam making, so at this time, no, there is
no
Andrew Kohlsmith wrote:
Yes, but then what are you doing with it? You're shuttling the new data
to/from a network card in a lot of cases. Combined with other traffic over
the PCI bus for normal system operation I could see you coming close to the
limitations of regular ole PCI.
Absolutely.
Steve Underwood wrote:
Since encoding typically requires 5 times as much compute as decoding,
for CELP based codecs, an encode onyl board would not be as dumb as it
seems at first sight :-)
Hah! I knew someone would say that!
___
Asterisk-Users mailing
On April 13, 2005 10:57 am, Kevin P. Fleming wrote:
Very true; realistically, modern PC hardware has more than enough
bandwidth to do what is required. The real issue is timing, based on
contention for resources, and how that impacts latency. The existing
boxes out there (not PCs) that handle
Kevin P. Fleming wrote:
Andrew Kohlsmith wrote:
Yes, but then what are you doing with it? You're shuttling the new
data to/from a network card in a lot of cases. Combined with other
traffic over the PCI bus for normal system operation I could see you
coming close to the limitations of regular
It means something to me.
Kevin
I am far from needing one of these yet but I am interested in them. I
think that clearing up who does the transcoding would help some
people. I wonder if someone will use this for large scale network
needs. Image Stream is just down the Highway from me and I know
[EMAIL PROTECTED] wrote on 04/12/2005 11:36:47 PM:
[EMAIL PROTECTED] wrote:
In other words, a PCI-based co-processor would double the PCI bus
bandwidth necessary. And with a latency-sensitive product like voice,
bus
contention is not something you want to add to! :)
It only
Andres wrote:
Can you confirm if there will be some sort of DSP daughther card add on
of some sort for the DS3000 so that we can run G729 transcoding? I
don't see how the DS3 interface would be usefull unless we could offload
transcoding stuff to onboard DSPs. Or is Digium only going to
Andrew Kohlsmith wrote:
secondary card for DSP functions is very inefficient of the PCI bus. I'd be
curious to know if the Digium cards can even do PCI-PCI DMA.
The Digium TDM cards can DMA into any RAM accessible over the PCI bus,
regardless of whether it is located on the motherboard or on a
[EMAIL PROTECTED] wrote:
Andres wrote:
Can you confirm if there will be some sort of DSP daughther card add
on of some sort for the DS3000 so that we can run G729 transcoding?
I don't see how the DS3 interface would be usefull unless we could
offload transcoding stuff to onboard DSPs. Or
Bicom Systems wrote:
What is target release date for DS3000P?
That has not been announced; sometime after today would be a safe
assumption :-)
___
Asterisk-Users mailing list
Asterisk-Users@lists.digium.com
[EMAIL PROTECTED] wrote on 04/12/2005 10:51:49 AM:
Andrew Kohlsmith wrote:
secondary card for DSP functions is very inefficient of the PCI
bus. I'd be
curious to know if the Digium cards can even do PCI-PCI DMA.
The Digium TDM cards can DMA into any RAM accessible over the PCI bus,
PM
Subject: Re: [Asterisk-Users] DS3000P - 20 E1 capacity on single card
[EMAIL PROTECTED] wrote on 04/12/2005 10:51:49 AM:
Andrew Kohlsmith wrote:
secondary card for DSP functions is very inefficient of the PCI
bus. I'd be
curious to know if the Digium cards can even do PCI-PCI DMA
[EMAIL PROTECTED] wrote:
Bicom Systems wrote:
What is target release date for DS3000P?
That has not been announced; sometime after today would be a safe
assumption :-)
Sure thing...
In any case as you wrote sometime after today we all should know :-)
Ta
Senad
Kevin P. Fleming wrote:
Andres wrote:
Can you confirm if there will be some sort of DSP daughther card add
on of some sort for the DS3000 so that we can run G729 transcoding?
I don't see how the DS3 interface would be usefull unless we could
offload transcoding stuff to onboard DSPs. Or
[EMAIL PROTECTED] wrote:
In other words, a PCI-based co-processor would double the PCI bus
bandwidth necessary. And with a latency-sensitive product like voice, bus
contention is not something you want to add to! :)
It only 'doubles the bandwidth required' when compared to a single-board
Matthew Boehm wrote:
So, no hardware encoding on this beast?
The announcement on the website makes no mention of transcoding, echo
cancellation or toast-and-jam making, so at this time, no, there is no
hardware transcoding apparently included. (Besides, would you really
want a board that could
Kevin,
Keep in mind that according to Wiki there are no DSP's on the
board.
-m
On Tue, 12 Apr 2005, Kevin P. Fleming wrote:
[EMAIL PROTECTED] wrote:
In other words, a PCI-based co-processor would double the PCI bus bandwidth
necessary. And with a latency-sensitive product like voice,
Kevin,
Mmm. Yep.
-m
On Tue, 12 Apr 2005, Kevin P. Fleming wrote:
Matthew Boehm wrote:
So, no hardware encoding on this beast?
The announcement on the website makes no mention of transcoding, echo
cancellation or toast-and-jam making, so at this time, no, there is no
hardware transcoding
Andrew Kohlsmith wrote:
Now they MAY have incorporated the TJ320 chip logic in the Xilinx Spartan II
FPGA but I would be **VERY** surprised if they did that. Just my opinion,
but I think that level 2 digium tech is full of shit.
Andrew is correct; there are no TigerJet parts on the quad-span
Remco Barende wrote:
like it says, the equivalent of 20 E1's or 28 T1's
and I guess you know how many channels a E1 or T1 PRI is
That is correct; the DS3000P will support full access to every channel
on the DS-3 (or E-3), however it is provisioned. In a T1-RBS signaling
mode, that means 672
Andrew is correct; there are no TigerJet parts on the quad-span cards.
There will also not be any TigerJet parts on the DS3000P card, as they
don't make anything that would be useful there :-)
Kevin,
Can you confirm if there will be some sort of DSP daughther card add on
of some sort for the
Andres wrote:
Andrew is correct; there are no TigerJet parts on the quad-span
cards. There will also not be any TigerJet parts on the DS3000P
card, as they don't make anything that would be useful there :-)
Kevin,
Can you confirm if there will be some sort of DSP daughther card add
on of
On April 11, 2005 03:17 pm, Andres wrote:
Can you confirm if there will be some sort of DSP daughther card add on
of some sort for the DS3000 so that we can run G729 transcoding? I
don't see how the DS3 interface would be usefull unless we could offload
transcoding stuff to onboard DSPs. Or
(Note, I do not work for nor speak for Digium.)
G711 only; if you want transcode do that on a cluster of boxes feeding the box
with this card in it.
-A.
Or wait another 10 generations of Pentium Processors to catch up with
DS3 G729 requirements :)
--
Andres
Network Admin
[EMAIL PROTECTED] wrote:
On April 11, 2005 03:17 pm, Andres wrote:
Can you confirm if there will be some sort of DSP daughther card add
on of some sort for the DS3000 so that we can run G729 transcoding?
I don't see how the DS3 interface would be usefull unless we could
offload transcoding
Come June/July an USB/PCI DSP cost effective solution should be
available
to address this issues. It will transcode nearly all codec's.
I am not in position to reveal the company name
at this stage unless MN wants to speak up :)
Put me on the mailing list for the PCI DSP card, I'll beta test if
On April 11, 2005 06:43 pm, Bicom Systems wrote:
Come June/July an USB/PCI DSP cost effective solution should be available
to address this issues. It will transcode nearly all codec's.
I am not in position to reveal the company name
at this stage unless MN wants to speak up :)
secondary card
On April 9, 2005 08:25 pm, Eric Wieling wrote:
Which specific Digium card does not use the TigerJet chip (as shown in
lspci)?
TE405P:
05:03.0 Communication controller: Xilinx Corporation: Unknown device 0314 (rev
01)
I imagine the TE410 and TE110 are both also similarly lspci'd.
I
On April 10, 2005 12:01 pm, Matthew Boehm wrote:
I have a TE405P and mine shows up as Xilinx but a lvl 2 tech a digium says
it still uses the TigerJet chipset. That's why it won't work in my Dell.
I'll paypal you US$100 if you can find a TJ320 chip on either the TE410P or
TE405P. It doesn't
I just checked digium's site. Looks like next big thing is coming to town
DS3 on single card. Would be nice to know how many channels it can handle.
Anybody had his hands on this card or knows some details ?
regards
m.
___
Asterisk-Users mailing list
izo wrote:
I just checked digium's site. Looks like next big thing is coming to town
DS3 on single card. Would be nice to know how many channels it can handle.
Anybody had his hands on this card or knows some details ?
Please God, if you can hear me, don't let them use a TigerJet chipet.
--
like it says, the equivalent of 20 E1's or 28 T1's
and I guess you know how many channels a E1 or T1 PRI is
On Sat, 9 Apr 2005, izo wrote:
I just checked digium's site. Looks like next big thing is coming to town
DS3 on single card. Would be nice to know how many channels it can handle.
Anybody
On April 9, 2005 02:13 pm, Eric Wieling wrote:
izo wrote:
I just checked digium's site. Looks like next big thing is coming to town
DS3 on single card. Would be nice to know how many channels it can
handle. Anybody had his hands on this card or knows some details ?
Please God, if you can
Andrew Kohlsmith wrote:
On April 9, 2005 02:13 pm, Eric Wieling wrote:
izo wrote:
I just checked digium's site. Looks like next big thing is coming to town
DS3 on single card. Would be nice to know how many channels it can
handle. Anybody had his hands on this card or knows some details ?
Please
On April 9, 2005 08:25 pm, Eric Wieling wrote:
Which specific Digium card does not use the TigerJet chip (as shown in
lspci)?
TE405P:
05:03.0 Communication controller: Xilinx Corporation: Unknown device 0314 (rev
01)
I imagine the TE410 and TE110 are both also similarly lspci'd.
-A.
Andrew Kohlsmith wrote:
On April 9, 2005 08:25 pm, Eric Wieling wrote:
Which specific Digium card does not use the TigerJet chip (as shown in
lspci)?
TE405P:
05:03.0 Communication controller: Xilinx Corporation: Unknown device 0314 (rev
01)
I imagine the TE410 and TE110 are both also similarly
38 matches
Mail list logo
