Threaded Mode | Linear Mode

feinicks · 17/02/2009 06:29 AM

ZiNgA BuRgA Wrote:GTX295 = unnecessary. Even disregarding the cost, it's two GPUs in one = BIG power sucker (and probably noisy too).
There's no game these days, or in probably in a year's time, that'll max out the card, and even later, just play the thing on lower settings (can you really tell the difference between 4x Anisotropic filtering and 16x? even if you can, I doubt you'll be paying much attention when you're actually focused on the gameplay).

CPUs are a bit different in that they're always somewhat max-out-able, but the i7 probably won't make your games run much faster than a Core 2.

they do infact, though not by much margin. i7 965 (?) at any rate. not i7 920, as much..

Vegetano1 · 17/02/2009 06:44 AM

ZiNgA BuRgA Wrote:GTX295 = unnecessary. Even disregarding the cost, it's two GPUs in one = BIG power sucker (and probably noisy too).
There's no game these days, or in probably in a year's time, that'll max out the card, and even later, just play the thing on lower settings (can you really tell the difference between 4x Anisotropic filtering and 16x? even if you can, I doubt you'll be paying much attention when you're actually focused on the gameplay).

Disagree!!

the GTX295 got CUDA technology meaning >>converting video with GPU "transcoding HD videos", CUDA suported program's eg Photoshop and other program's."editing digital images using Adobe CS4"

Video converting with gpu is way faser then with cpu,..

the 295gtx also got PhysX, "Experience jaw-dropping PhysX gaming effects"

this card also suports multiple monitor's way better,.

http://www.nvidia.com/object/product_gef...95_us.html

ZiNgA BuRgA · 17/02/2009 06:52 AM

Vegetano1 Wrote:the GTX295 got CUDA technology meaning >>converting video with GPU "transcoding HD videos", CUDA suported program's eg Photoshop and other program's."editing digital images using Adobe CS4"

Video converting with gpu is way faser then with cpu,..

Hardly any *good* video encoders run on CUDA. There are a number of technical difficulties with this. Primarily, encoding is mostly a serial issue. CUDA is MUCH slower than the CPU at serial processing - it only really starts to shine when you run like 256+ simultaneous threads, which optimising an encoder for is extremely difficult. I recall the x264 developers trying to make a CUDA version, but they dropped it as a typical quad core CPU ran much faster.

Vegetano1 Wrote:the 295gtx also got PhysX, "Experience jaw-dropping PhysX gaming effects"

So do all the lower GeForce cards, starting from the 9xxx range. Plus PhysX doesn't have a heap of support anyway, and with ATi not supporting it, unlikely to happen in the near future.

Have no idea about "better multiple monitor support" (how can it get "better"?), but don't use nVidia's site for reference - it's obviously full of marketing material to try and get you to buy the thing. This is the case with all "extreme" products - typically heavy advertising is done on these. (it's quite interesting, cause there's many similarities to the fashion industry)

Vegetano1 · 17/02/2009 07:35 AM

ZiNgA BuRgA Wrote:
Vegetano1 Wrote:the GTX295 got CUDA technology meaning >>converting video with GPU "transcoding HD videos",  CUDA suported program's eg Photoshop and other program's."editing digital images using Adobe CS4"

Video converting with gpu is way faser then with cpu,..
Hardly any *good* video encoders run on CUDA.  There are a number of technical difficulties with this.  Primarily, encoding is mostly a serial issue.  CUDA is MUCH slower than the CPU at serial processing - it only really starts to shine when you run like 256+ simultaneous threads, which optimising an encoder for is extremely difficult.  I recall the x264 developers trying to make a CUDA version, but they dropped it as a typical quad core CPU ran much faster.

Vegetano1 Wrote:the 295gtx also got PhysX, "Experience jaw-dropping PhysX gaming effects"
So do all the lower GeForce cards, starting from the 9xxx range.  Plus PhysX doesn't have a heap of support anyway, and with ATi not supporting it, unlikely to happen in the near future.

Have no idea about "better multiple monitor support" (how can it get "better"?), but don't use nVidia's site for reference - it's obviously full of marketing material to try and get you to buy the thing.  This is the case with all "extreme" products - typically heavy advertising is done on these. (it's quite interesting, cause there's many similarities to the fashion industry)

;p./ i get whot you mean,. but if you see this pic,. you really tend to believe Gpu is faster then cpu encoding,.>>
[Image: heterogenousComputing.png]

So whot you'r saying is that there are't many encoders yet wich use the gpu>!?

I do think working with Photoshop or any other imageing program works best with Cuda,. my 8800gts already suoprts some feature's in PS like the scale grid but when i have more then 15 picture's open,. PS stops suporting this,. there is also some additional feature's wich don't work with the 8800gts,.

here is the history of gpu computing,. (how nvidia,. likes/sees it)
http://www.nvidia.com/object/GPU_Computing.html

PhysX gaming prob works best with a faster graphics card,.. if you want to be able to use the highest settings,.

Multiple monitors will work better,. using different gpu core's for different monitors,. ;p

Assassinator · 17/02/2009 08:05 AM

Ok, put it this way. Compare the GTX 295 with lets say a normal 9800.

You pay 400% the price, for at most 200% the real performance. And most of the time (for all but a few games), that improvement in performance would be worthless anyway, since you can't see the difference between something like 60 and 120fps (ie. overkill).

So IMO the GTX 295 is pretty pointless. Spending that much more money just to get a few odd games to run better is not an investment I would make.

Vegetano1 Wrote:
ZiNgA BuRgA Wrote:
Vegetano1 Wrote:the GTX295 got CUDA technology meaning >>converting video with GPU "transcoding HD videos", CUDA suported program's eg Photoshop and other program's."editing digital images using Adobe CS4"

Video converting with gpu is way faser then with cpu,..
Hardly any *good* video encoders run on CUDA. There are a number of technical difficulties with this. Primarily, encoding is mostly a serial issue. CUDA is MUCH slower than the CPU at serial processing - it only really starts to shine when you run like 256+ simultaneous threads, which optimising an encoder for is extremely difficult. I recall the x264 developers trying to make a CUDA version, but they dropped it as a typical quad core CPU ran much faster.
;p./ i get whot you mean,. but if you see this pic,. you really tend to believe Gpu is faster then cpu encoding,.>>

So whot you'r saying is that there are't many encoders yet wich use the gpu>!?

It could maybe be faster if the encoder can properly encode using 240 threads simultaneously.

Using x264 as an example (since that's what Zinga was talking about), using 240 threads is un-viable. All the splicing will kill your quality. Even 24 theads is probably pushing it.

Like Zinga said, encoding videos, at least as of this point of time, is the CPU's field. GPUs are good for decoding video though, but decoding video isn't particularly hard anyway, so petty much any GPU should be able to do it at a good enough speed (not lag), and you won't need a monster GPU for that.

feinicks · 17/02/2009 08:09 AM

There are hardly any good GPU Based encoders due to the fact that it is a new field to explore. Akin to the scenario: Have water, will fish. But the fish will only come after sometime.

GPU are very powerful, indeed. There is that ongoing debate about whether in the future, the GPU will replace the CPU as the complete processing unit or not. Wee have a thread here about the same.

Assassinator · 17/02/2009 08:19 AM

feinicks Wrote:There are hardly any good GPU Based encoders due to the fact that it is a new field to explore. Akin to the scenario: Have water, will fish. But the fish will only come after sometime.

The current encoders don't work well with too many threads, and are also quite heavily dependent on current CPU features such as the SSEx and other instruction sets, which the GPU doesn't have, to speed up the process.

feinicks Wrote:GPU are very powerful, indeed. There is that ongoing debate about whether in the future, the GPU will replace the CPU as the complete processing unit or not. Wee have a thread here about the same.

The GPUs strength comes in parallel processing. Like that previous picture V1 posted, processing 240 threads simultaneously. (Likewise, the CPU is good at serial processing)

But some stuff simply cannot be processed parallel, and so I don't think the GPU will ever completely replace the CPU. At least not any time soon.

Vegetano1 · 17/02/2009 08:29 AM

@Assassinator i am not sure but i read somewhere that the 8800 is still better then the 9800,. atleast faster,. although the 8800 does't have physx,.

you can't denie that the GTX 295 is way faster then the 9800 and you are 200% better off playing a game like Crysis with the 295gtx then with a 9800,. its a simple fact,. and other games too,. although i don't really want to question you or Zinga,. but some games have very high settings available,..
prices of the 9800 were also high when the card was first presented,.

anyway the most importend thing besides game suport now is also the suport for the CS4 series and other high end graphics program's and with maybe some good video GPU encoders on the way a expensive graphics card could be a good choice,.

Blur

check this and click "video processing",.>> they sure make it look worth awhile to encode with the GPU,. nvidia claims 20x faster then with cpu.
http://www.nvidia.com/content/graphicspl...index.html

o and skip intro,. ;p

feinicks · 17/02/2009 10:43 AM

Assassinator Wrote:
feinicks Wrote:There are hardly any good GPU Based encoders due to the fact that it is a new field to explore. Akin to the scenario: Have water, will fish. But the fish will only come after sometime.

The current encoders don't work well with too many threads, and are also quite heavily dependent on current CPU features such as the SSEx and other instruction sets, which the GPU doesn't have, to speed up the process.

Not denying that current gen encoders, or for that matter any parallel process application, are too CPU dependent. However, my argument is based on the fact that, in the future, application design may change from CPU to GPU to harness the more free and efficient power of the latter. CUDA is a an example of GPU-based instruction set. And already, the importance of CUDA has been recognised. Give it time, and you may have GPUs comparing with CPUs.

Assassinator Wrote:
feinicks Wrote:GPU are very powerful, indeed. There is that ongoing debate about whether in the future, the GPU will replace the CPU as the complete processing unit or not. Wee have a thread here about the same.

The GPUs strength comes in parallel processing. Like that previous picture V1 posted, processing 240 threads simultaneously. (Likewise, the CPU is good at serial processing)

But some stuff simply cannot be processed parallel, and so I don't think the GPU will ever completely replace the CPU. At least not any time soon.

As I quoted above, I am more interested in the future of the GPUs. CPU are reaching a set limit on how smaller and powerful they can get, without compromising on hardware estate. There is a finite value as how small and how many cores you can fit onto a board, without increasing its size (and hence cost) and, more importantly, heat generation. Ever since the advent of discreet graphics cards, GPUs have been spared this dilemma. This is quite evident in the ever increasing sizes of the cards, to allow more efficient channeling of information and dissipation of heat.

In the future, wee might even have developed a hybrid CPU-GPU system. I'm sure that this will be the goal due to the simple fact that though some basic level instruction have to be serial processed, parallel processes have more efficiency and in general greater flexibility and power.

GPUs are far too young as of now and too uni-directional to seriously threaten, but forget not that if properly constructed, a super computer comprising of only GPU core is as potent as its CPU counterpart.

ZiNgA BuRgA · 17/02/2009 06:37 PM

I can't say much about Photoshop, since I rarely use it, and ATi cards don't have CUDA anyway.

Vegetano1 Wrote:PhysX gaming prob works best with a faster graphics card,.. if you want to be able to use the highest settings,.

Not really. The aim of PhysX is to provide a hardware acceleration interface for physics calculations. Having a faster graphics clock doesn't improve it.

Vegetano1 Wrote:Multiple monitors will work better,. using different gpu core's for different monitors,. ;p

Playing 2 games at once? Otherwise, no. Drawing stuff to a screen is extremely fast.

Vegetano1 Wrote:@Assassinator i am not sure but i read somewhere that the 8800 is still better then the 9800,. atleast faster,. although the 8800 does't have physx,.

The 9800 is a renamed 8800 (not exactly sure which versions however). In other words, speeds shouldn't be different at all, just like the 8600GT and 9600GSO being identical.

Vegetano1 Wrote:you can't denie that the GTX 295 is way faster then the 9800 and you are 200% better off playing a game like Crysis with the 295gtx then with a 9800,. its a simple fact,. and other games too,. although i don't really want to question you or Zinga,. but some games have very high settings available,..

Don't use the high settings then >_> Do you really need it? Willing to pay all that for an image that has two smoother lines?

Vegetano1 Wrote:prices of the 9800 were also high when the card was first presented,.

That's why you generally don't buy top-of-the-range cards >_>

Vegetano1 Wrote:anyway the most importend thing besides game suport now is also the suport for the CS4 series and other high end graphics program's and with maybe some good video GPU encoders on the way a expensive graphics card could be a good choice,.

GPGPU computing is still in its infancy. Unless you actually have a specific need for it, wait for it to mature instead.

Vegetano1 Wrote:check this and click "video processing",.>> they sure make it look worth awhile to encode with the GPU,. nvidia claims 20x faster then with cpu.
http://www.nvidia.com/content/graphicspl...index.html

I claim the CPU is 5x faster than that. Do you believe me?
Either case they give very little information on it.

feinicks Wrote:Not denying that current gen encoders, or for that matter any parallel process application, are too CPU dependent. However, my argument is based on the fact that, in the future, application design may change from CPU to GPU to harness the more free and efficient power of the latter. CUDA is a an example of GPU-based instruction set. And already, the importance of CUDA has been recognised. Give it time, and you may have GPUs comparing with CPUs.

This is difficult.  A good metaphor for parallel processing is getting multiple people to do a task.  Some tasks, 240 (less trained) people can do a lot better than 4 (well trained).  However, some tasks simply won't make sense to have 240 people do it.  For example, playing a game on a computer - you can't really have 240 people doing the thing at once.  Whereas, 4 people, which are much better at doing it, will easily beat 240 people who aren't so good.

Encoding is something that is largely serial, but can be somewhat made parallel.  It's serial because a lot of aspects depend on the events occurring beforehand, ie you must do action A before action B because action B requires information obtained from action A.  You can't "parallelize" this because it requires this sort of ordering.
You might take the example of adding two 100-digit numbers.  For one person, this would take a while.  But if wee got 10, you could split up and give 10 digits to each person, then do a "merge" at the end.  It's not perfect, as a carry can potentially make things a lot slower, but in most cases, it's somewhat like "parallizing" a serial task.

See also: http://en.wikipedia.org/wiki/GPGPU#Misconceptions

Threaded Mode \| Linear Mode whot do you guess??
Author	Message