Welcome, Guest. Please login or register.
Amiga Kit Amiga Store Iridium Banner AMIStore App Store A600 Memory

AuthorTopic: AmigaOne X1000  (Read 2808 times)

0 Members and 1 Guest are viewing this topic.

Offline Pyromania

AmigaOne X1000
« on: June 25, 2010, 01:34:25 AM »
If the X1000 sells well and the XMOS technology is innovative and allows 3D rendering to be spead up we will add support for XMOS in Aladdin 4D 6.0 or 7.0. The original creator of Aladdin 4D Greg Gorby wanted to support the Transputer project which XMOS descends from. We will respect Greg’s wishes and take advantage of this cool feature of the X1000.
 

Offline Piru

  • \' union select name,pwd--
  • Hero Member
  • *****
  • Join Date: Aug 2002
  • Posts: 6946
  • Total likes: 0
    • http://www.iki.fi/sintonen/
Re: AmigaOne X1000
« Reply #1 on: June 25, 2010, 02:30:10 PM »
Quote from: Pyromania;567093
If the X1000 sells well and the XMOS technology is innovative and allows 3D rendering to be spead up

XMOS has no FPU, is slow and has so poor connectivity to the rest of the system it hardly makes any sense to spend resources trying to add 3D rendering support.

I'm certain there will be others that claim otherwise.

Anyhow, GPU is the future of rendering.
 

Offline Zac67

Re: AmigaOne X1000
« Reply #2 on: June 25, 2010, 05:09:57 PM »
Quote from: Piru;567179
Anyhow, GPU is the future of rendering.


Yeah - hard to beat 1+ TFLOPS...
 

Offline Bif

Re: AmigaOne X1000
« Reply #3 on: June 25, 2010, 08:08:13 PM »
Quote from: Pyromania;567093
If the X1000 sells well and the XMOS technology is innovative and allows 3D rendering to be spead up we will add support for XMOS in Aladdin 4D 6.0 or 7.0. The original creator of Aladdin 4D Greg Gorby wanted to support the Transputer project which XMOS descends from. We will respect Greg’s wishes and take advantage of this cool feature of the X1000.


This would seem like a pretty poor use of the chip. You could possibly get some kind of speedup, but it would be a lot of effort, maybe for 1% improvement. Wrong tool for the job I think. But hey don't let my thoughts stop you, it might be fun to try anyway.

I can't see much use for the XMOS in this machine given it is fixed point and weak, but if I was going to use it, I would probably reserve it for OS functionality where it can increase responsiveness of the machine by not stealing significant time slices from CPU via interrupt routines or threads. Off the top of my head I can't even think of what those would be though where that would also be advantageous as I don't know what OS services take enough CPU to make it worth offloading. Perhaps the sound engine. And I'm not even sure the chip can run independently enough to give a speedup if the CPU needs to constantly page code on to it and stuff.

Perhaps it makes sense to let the chip run classic emulation routines. It might be interesting that the chip can support multiple threads with low latency, might make for interesting emulation of multiple chips at one time so that they almost appear to run concurrently.

Realistically I think any of what I just said is a waste of time though, even if X1000 had a large user base, writing code for the XMOS is probably a dead end because it isn't present in all Amiga machines, nor can we necessarily expect these chips to continue to get more powerful as CPUs do. A 5% speedup in the clock rate of the CPU in the X1000 should be more than enough MIPS to cover anything the X1000 would have done and render it pointless. Plus any code you write for the X1000 is going to be limited in performance to what the X1000 can do - if the sound engine did run on it, it would suck if it just crapped out at 20 channels of audio instead of being able to say handle up to 200 on the CPU.
 

Offline DAX

Re: AmigaOne X1000
« Reply #4 on: June 26, 2010, 10:43:07 AM »
Quote
no fpu...can't beat 1 GFlops

As we are entering the era of heterogeneous computing any additional data crunching facility the system has acces to is good.

Talking about 3D rendering (my field) and GP-GPU, there have been a lot of misconceptions flying around, with fancy RT demonstrations that quite frankly demonstred very little. The problem there, is that they use a lot pre-baked data while GP-GPU finishes the rest (AtiCinema2.0 is a good example), but when the rendering is 100% handled without baked data (as shown in some horrid raytracing examples), the quality of the rendering drops dramatically and can't hold a candle to non RT algorithms, while when the CPU intervenes in a hybrid mode, the slow downs are as heavy as it's the improvement in picture quality you set (one example would be the recent "mostly gpu based" Quicksilver engine for 3DS max, VS the same scene done in plain mental-ray, but it gets deeper than that).

To solve this dilemma ex Nvidia/Ati employees have formed Caustics Graphics, a company that produces a board which only intervene into a more complex process in order to aid where CPUs and GPUs are not very good at, preparing data for their consumption (heterogeneous computing at its finest we might say, everything is used) more infos about this here:

http://www.vizworld.com/2009/09/caustics_3dsmax_demonstration/ (whatch the video and listen to what they ave to say on the subject)

Considering only some data is sent to the board for prepping (the rest is done by the CPU and GPU) the bandwidth required is modest and I wonder if a minicluster of Xmos chips could handle the same kind of data (that card is expensive, you can build a 256 Xmos cluster for less) and work together with both CPU and GPU in a similar manner (or at least have the trio achieve a 1 GI frame per second result, which is still very good).

On Xmos forum you can read about the SoftFloat they implemented, it is extremely fast for being "soft", and on a cluster of 500Mhz chip you might still get what you need for "trio" operartions as the one above (or other, a different thing would be if it had to handle everything but this is not the case).

@Pyromania
In any case go for it, it's gonna be a fun exercise and you will grant the original creator's wish :)
« Last Edit: June 26, 2010, 11:07:16 AM by DAX »
 

Offline F1Lupo

Re: AmigaOne X1000
« Reply #5 on: February 21, 2011, 04:42:57 AM »
Quote from: Pyromania;567093
If the X1000 sells well and the XMOS technology is innovative and allows 3D rendering to be spead up we will add support for XMOS in Aladdin 4D 6.0 or 7.0. The original creator of Aladdin 4D Greg Gorby wanted to support the Transputer project which XMOS descends from. We will respect Greg’s wishes and take advantage of this cool feature of the X1000.


Bring it on man:afro:
____________________________________________________________________
c64-dual sids, A1000, A1200-060@50, A4000-CSMKIII
Indivision AGA & Catweasel MK4+= Amazing
! My Master Miggies-Amiga 1000 & AmigaOne X1000 !
--- www.mancave-ramblings.blogspot.ca ---
  -AspireOS.com & Amikit- Amiga for your netbook-
***X1000- I BELIEVE *** :angel:
 

Offline Karlos

Re: AmigaOne X1000
« Reply #6 on: February 21, 2011, 07:39:36 AM »
Quote from: Piru;567179
XMOS has no FPU, is slow and has so poor connectivity to the rest of the system it hardly makes any sense to spend resources trying to add 3D rendering support.

I'm certain there will be others that claim otherwise.


I have to agree. The XMOS looks handy for implementing very low latency interrupt handlers and things like that, but as a render accelerator? Hard to see how it's going to be as fast as the PA6T for that job.

Quote
Anyhow, GPU is the future of rendering.


Indeed. Actually, some of the best demonstrations I've seen use hybrid techniques where conventional rasterization (which is what the existing hardware excels at anyway) is used for primary rays and then compute kernels running on the GPU are invoked for the secondary rays where it is able to take advantage not only of the massive parallelism of the GPU but also the already-rendered depth and stencil buffers to cull certain rays.
int p; // A