Intel SandyBridge... I don't get it.

solarflow99 solarflow99-Re5JQEeQqe8AvxtiuMwx3w at public.gmane.org
Thu Jan 27 21:41:05 UTC 2011


isn't it true that in several months, all those analog cards won't work anymore?



On Thu, Jan 27, 2011 at 4:32 PM, D. Hugh Redelmeier <hugh-pmF8o41NoarQT0dZR+AlfA at public.gmane.org> wrote:
> On Thu, 27 Jan 2011, Lennart Sorensen wrote:
>
> | Date: Thu, 27 Jan 2011 14:06:23 -0500
> | From: Lennart Sorensen <lsorense-1wCw9BSqJbv44Nm34jS7GywD8/FfD2ys at public.gmane.org>
> | Reply-To: tlug-lxSQFCZeNF4 at public.gmane.org
> | To: tlug-lxSQFCZeNF4 at public.gmane.org
> | Subject: Re: [TLUG]: Intel SandyBridge... I don't get it.
> |
> | On Thu, Jan 27, 2011 at 12:46:25PM -0500, D. Hugh Redelmeier wrote:
>
> | >     + I want to stay with analog as long as I can because DRM prevents
> | >       recording digital cable signals (the record-component-out hack is a
> | >       little hacky, but I will try that; it requires a lot of Rogers STBs
> | >       and whacky open-loop control of the tuner)
> |
> | Open loop?
>
> http://en.wikipedia.org/wiki/Open-loop_controller
>
> The controller gets no feedback from the controllee so it doesn't know
> for sure that its commands have had the desired effect.  IR remotes
> are only open-loop because they assume that the human can monitor the
> results and correct for failures.  An IR blaster cannot observe the TV
> screen or set-top-box display to see if the commands were registered.
>
> Is firewire control of the STB yield feedback to the controller?
>
> Another question:
>
> I bought a firewire controller just for controlling a couple of STBs.
> The controller has one ordinary FW connector and two 9-pin connectors
> ones for 800MHz firewire (I think).
> http://www.dynexproducts.com/products/computers/DX-PCI2PF.html
> I guess that I need a 9-Pin to old-fashioned FW converter or cable.
>
> Is that correct?  Are such cables common?
>
> | > So: how much better are 1156 systems than 775 systems (and Socket AM2+
> | > systems) for my purpose?
> |
> | Depends what you are doing.
>
> TV stuff.
>
> [I've just ordered DDR3 RAM, so I'm getting committed to this
> project: nothing I currently have would accept DDR3 DIMMs.]
>
> | > Will 5 SD streams swamp + whatever Myth is doing swamp the memory bus?
> | > If so, 1156 would be much better.
> |
> | Well if you use hardware mpeg2 compression cards, then I doubt it would
> | be much load at all.
>
> Yes.
>
> | > Will anything need much CPU?  Since I've been living with an Athlon
> | > 1700 XP, I think the CPU issue is minor.  I do hope to add a couple of
> | > Hauppauge HD PVRs and I don't know what load that adds.
> |
> | The HD PVR compresses to MPEG 4 in hardware, so the data isn't that much,
> | although USB being extremely inefficient may take a decent chunk of CPU
> | to handle.
>
> Are USB3 ports (as on the socket 1156 I'm looking at, (when used as
> USB 2 (for the Hauppauge HD PVR) likely to be less CPU intensive than
> ordinary USB2 ports?
>
> | Of course there are crazy solutions out there:
> | http://www.magma.com/4slot.asp
> |
> | Box with 4 PCI slots, attaches by cable to the host and uses one slot
> | there (PCIe or PCI, your choice).  Too bad it is stupidly expensive
> | ($1500).  It would be cheaper to build multiple backend machines with
> | a couple of PCI cards each.
> |
> | Another method is a PCI bridge riser like this one:
> | http://www.orbitmicro.com/global/rc2-019-p-686.html
> | Turns one PCI slot into 3 slots.  Add a PCI extender cable and you could
> | actually use multiple.  How you mount the cards in the case is another
> | interesting issue then.
>
> A big honking motherboard looks to be cheaper.
> --
> The Toronto Linux Users Group.      Meetings: http://gtalug.org/
> TLUG requests: Linux topics, No HTML, wrap text below 80 columns
> How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists
>
--
The Toronto Linux Users Group.      Meetings: http://gtalug.org/
TLUG requests: Linux topics, No HTML, wrap text below 80 columns
How to UNSUBSCRIBE: http://gtalug.org/wiki/Mailing_lists





More information about the Legacy mailing list