Welcome to our community

Be a part of something great, join today!

Log in
  • Hey all, just changed over the backend after 15 years I figured time to give it a bit of an update, its probably gonna be a bit weird for most of you and i am sure there is a few bugs to work out but it should kinda work the same as before... hopefully :)

NAB - Apple announces Special Event

OEM aja kona's in Discreet Lustre I think.

I thought that Discreet Lustre uses a Nvidia Quadro card with HD-SDI out (and of course DVI out).

It also supposedly has a DVS Centaurus board.
 
The 4500sdi is the model # of the nvidia quadro card?

It might be that only Combustion can use Aja hardware.
 
ElysianChris said:
I'm really hoping Apple has integrated Final Touch into FCP or at least updated the standalone app. so that it works correctly with FCP. I can get by with Colorista for the time being, but it would be nice to have a really good DI that integrates with FCP, especially if it is cost effective.
Click to expand...

I would expect some Final Touch tools to end up in FCP, but I really don't want my NLE to be weighed down with too many extras.

Apple needs to work on the Final Cut Studio integration.
 
I think that if apple does an extreme version it will be priced not a lot higher than fpc but will require a high end station configuration that they will be happy to sell you. It will all be a bargain compared to other solutions.
I don't think this will happen till maybe the end of the year; it would make a Hell of a Christmas gift ! (no pun intended)
Aloha
-A
 
IMO, I don't see Apple doing an extreme version.

A- If they use hardware acceleration, it will probably be off a commodity GPU. A lot of high-end systems are using GPU accleration... Discreet/Autodesk Flame, Mistika are two such systems. These are six-figure systems that use commodity hardware. So I doubt they will be using any expensive custom hardware, since commodity GPUs will suffice.

B- A big part of their business model is to sell in volume instead of making a high margin on each sale. According to their marketing department, Final Cut has sold 500,000 units. They could probably make multiple versions of their software to target the high-end market (i.e. sell cinema tools at a higher price); however, they are probably sacrificing that small amount of money for marketing reasons. If you buy Final Cut Studio, it is the same Final Cut Studio that is being used to cut Hollywood films. They are using their (small) high-end market to prop up the much larger mid and low-end markets.

2- As far as turnkey solutions go, that market is already being serviced by the VAR market (value-added reseller). AFAIK, those VARs are able to offer better support AND price than apple (i.e. they can configure systems with cheaper, better solutions than XSERV RAID).
 
Glenn,
I think that my previous post covers your A,B and 2. Apple can hit the high end by selling those 500K users a slightly higher priced software with a slightly more turbo hardware configuration that will do what the other guys six figure systems do. They would easyly sell the volume necessary and have the bragging rights as well, perfect marketing. If you beliaeve in the red model, this should make sense.
Aloha
-A
however, it won't happen at this nab.
 
Poi Boy said:
Glenn,
I think that my previous post covers your A,B and 2. Apple can hit the high end by selling those 500K users a slightly higher priced software with a slightly more turbo hardware configuration that will do what the other guys six figure systems do. They would easyly sell the volume necessary and have the bragging rights as well, perfect marketing. If you beliaeve in the red model, this should make sense.
Aloha
-A
however, it won't happen at this nab.
Click to expand...

What do you get when you spend $500k?

It's more than just software and hardware, it's 24/7 on site support, rock solid performance, industry standard workflow and format support, not just Quicktime.

The people who buy the $500k systems have clients who are willing to pay premium prices for speed and reliability, to get the job done while they wait.

I don't know if Apple wants to deal with a client base this demanding.

There is much more profit in the middle ground, and they can add/update as they please.

It's true that Shake is a high end app, but it already had a loyal user base before Apple bought it.
 
I know that for SGO Mistika, it's just as buggy as Final Cut. They also don't offer on-site support, and don't offer 24/7 support to customers not in their time zone (Spain). It also doesn't support most Quicktime or AVI codecs (or firewire).

However, it's really really fast. It runs circles around Final Cut... with a $800 gaming video card.
 
AppliedVisual said:
IBM and Sony have a joint copyright on the "Cell" name as it pertains to CPUs.

Nothing wrong with Cell, it's a great multi-core CPU (available in 3, 7 and 9 core configs right now). It scales very well using multiple Cell processors in a NUMA-style configuration (memory banks are isolated to their respective Cell processors). The Cell on it's own is rather weak compared to a Core 2 series x86 processor or most desktop style CPUs. But as I said, they scale well and have their purpose. They're ideal for embedded applications, like the PS3 or a 3-core Cell would do really well inside a new Samsung refrigerator with built-in DVD player and coffee maker.

I could see Cell being used on video processor boards as maybe a host processor, but it's not going to do the bulk of the actual video processing. If this whole hardware-accelerated FCP Extreme were true, I would look for a custom set of ASICs from TI, Broadcom or nVidia, possibly ATI/AMD. Apple isn't going to re-invent the wheel here and they're probably not going to start from scratch either... there will be an underlying hardware partner like nVidia, AJA or BlueFish helping to at least design the hardware and possibly even supply a good bit of the firmware and software for it.
Click to expand...

I agree ... but ...

always tricky to compare processors, even when they are similar -

Cell was designed specifically to speed up image render and SONY is looking to sell these processors to the CGI community -

see page 29 - 3D WORLD MAGAZINE - April 2007 issue
 
I have little idea what I'm talking about here (hehe), but the raw specs for the cell processor would make it a very good candidate for speeding up image processing. It can do roughly ~200 GFLOPs/sec, which is several times more than a typical Pentium/AMD processor. It's ~100 GLOPs/sec (single precision) if you don't count multiply-adds as two operations.

It can also have the benefit of a high bandwidth memory subsystem, but that depends on what kind of system you stick it in (dual Cell blades have custom hardware).

It's in the same ballpark as GPUs and FPGAs in terms of raw performance. The advantage may be that the Cell is easier to program (or not; depends on what happens on the GPU front with ATI+Nvidia).

Sony has been making many press releases about Cell's possibility in media content creation systems (i.e. possibly like a Discreet Flame?)... although it doesn't seem like they are shipping anything.

2- On a slightly different note, Sony has this really neat image searching system that runs on a Cell workstation. It organizes images based on visual similarity. Very fast and it works pretty well (for a very difficult problem). If you're at NAB, check out the Petasite booth.
 
I think the problem with Cell isn't Cell. It's Sony's own marketing department getting ahead of themselves and some of the top brass at Sony buys into their own hype. So they try to make products that just aren't quite do-able yet with the current hardware.

Look at what IBM is doing with Cell... That is far more realistic at this point in time. Embedded AIX and Linux/UNIX applications -- yes, even Cell being used as additional media processors in workstations. One of the most logical and impressive implementations I've seen for Cell thus far have been as media processors for streaming media servers. Think of a media server that can ingest video clips in nearly any format availalbe and stream to any user on the web in any format they desire all on-demand.

Don't get caught up in Sony's GFLOPS numbers, it only tells part of the story. Cell in its current form still has a lot of factors working against it that can cripple it severely. Starting with external bus structure and bandwidth. IBM is working on it and Sony is hyping it up generating investor revenue.
 
You must log in or register to reply here.
Back
Top