Slashdot Mirror

nVidia has dropped support for cards older than the GForce4. I have a GForce2 with 64MB and TV tuner that would benefit from this driver.

No, NVIDIA has not dropped support for older cards - they support a driver for legacy cards which continues to be updated for new X and kernel releases. Look for the "Legacy GPU" release:
http://www.nvidia.com/object/unix.html

Cards older than GeForce4 were fixed function pipeline parts - programmable pipeline cards are a different beast, so a split in the drivers is reasonable.

Applications that use OpenGL or Direct3D are subjected fairly heavily to Amdahl's Law. The problem is that there is a (quite severe) penalty to submitting data to the APIs from different threads, which generally means that all submission to the API is done from a single thread.

If the limiting framerate issue for your title is submission to the API (and for a lot of 3-D graphics applications, it is [warning: PDF]), then you're not going to get any speedup on multi-core systems, and there won't even by a way to improve the latency when typing messages because displaying those messages is what is taking so long.

Use GPU-accelerated decoding, ala PureVideo. Not that I wouldn't also want the CPU power to do it too.

Who, exactly, do they think is going to use this besides amateurs and little tools companies (like the ones linked to in the article) who cater to amateurs...?

Sony (PS3 SDK). Epic (Unreal Engine 3). Nvidia (FX Composer). AGEIA (physics).

XML provides more than just a way of serialising a tree into text - I've not looked into the details very far, but what I've seen is that COLLADA uses XML Schema for validation, URIs for references between different locations (e.g. defining a piece of geometry, then adding several instances of it into a scene definition - and then changing it into an external URI if you don't want everything in the same file, or if you want to point to binary data instead of more XML, and having the standard XML tools deal with that correctly), and you can stick custom bits of XML into certain extension points (which standard tools can't parse but can pass along unchanged for later tools). You could do all that without XML, but the designers decided it would be more successful if they did use it.

If your artists can export a model from 3ds Max, load it into FX Composer and tweak the shaders to make it look good, load it into a physics simulator to make sure it reacts sensibly, then have it converted into the optimised native format for whatever engine you're using - and if you're no longer constrained in choice of tools (maybe you want to change from Max to Maya, or support a modding community with Blender, or load assets from your last game into your new engine) because they all support the same standard format, and you don't have to write all the code yourself - then it seems like it can have a practical benefit. I'm sure it doesn't work perfectly in practice, and it's not going to give groundbreaking improvements to the game development process, but it appears to go a long way in the right direction and it looks like it's gaining some real support.

IANAL, but I've read about this a lot and I think I understand it well.

The software in question is divided into three parts: the Linux kernel itself, the open-source stub written by nvidia, and the binary nvidia driver. The first is licensed under the GPL, and the other two are licensed under the nvidia software license. However, there is one part of that license that only applies to the stub, not the binary:

2.1.2 Linux/FreeBSD Exception. Notwithstanding the foregoing terms of Section 2.1.1, SOFTWARE designed exclusively for use on the Linux or FreeBSD operating systems, or other operating systems derived from the source code to these operating systems, may be copied and redistributed, provided that the binary files thereof are not modified in any way (except for unzipping of compressed files).

So, here's the theory (from nvidia's perspective): the binary part is not a derivative work of the Linux kernel; it's a derivative work of the Windows nvidia driver. The open-source stub is a derivative work of the kernel, but that's okay because it's redistributable (again, according to nvidia). In other words, only distributing the stub as open source satisfies the GPL.

Now, here's the catch: once you compile the kernel and the stub together the end result is a derivative work of the kernel, but not entirely GPL, so you're not allowed to redistribute it (does it seem like I contradicted myself? It should...). And the binary part still isn't a derivative work of the kernel because it's linked in at runtime in a similar way that a userspace program would link in to make system calls.

Now, here's the other perspective: some people say that having redistribution rights to that stub isn't good enough, that it must be GPL (or better*). This makes perfect sense to me, because the clause I quoted obviously fails to preserve the "four freedoms" that the GPL requires. So, let's imagine that nvidia distributed the stub as GPL, to make it legal. Then what? Well, since the stub is also part of the nvidia driver, it would then have to be GPL too because otherwise you'd have GPL code being a derivative work of non-GPL code, which the GPL doesn't allow. In other words -- unless I'm mistaken -- you can't make a closed-source extension to GPL code (which is obvious) but you also can't make a GPL'd extension to closed code unless you write an exception to the license. And nvidia can't write an exception because the Linux kernel doesn't have one. Therefore, distribution of the nvidia driver by anyone whatsoever (including nvidia) violates the GPL, even if the kernel itself isn't distributed along with it (because of the definition of "derivative work").

So, here's the bottom line: it may or may not be illegal to distribute the nvidia code by itself, and it's certainly illegal to distribute it linked with the kernel. Either way, though, it's the GPL (and/or copyright law) that causes it to be so, regardless of nvidia's license (unless it were to become genuinely GPL-compatible).

Hmm... you know what? After writing that, I'm less certain that I understand it. Oh well... : \

(*i.e., less restrictive -- like the BSD or MIT license)

My GeForce 6200 can do more GFLOPS than the fastest Core 2 Duo, but you're not going to run Linux on it.

O rly?

(Yes, I'm aware that the 6200 isn't CUDA-compatible. And yes, I'm aware that CUDA doesn't compile all the code to run on the GPU. It is getting closer, however...)

The shim is not under the GPL, it's under this license. Which permits nothing but redistribution in distros, with optional modification to the wrapper (or shim) but not the binary.

Linux supports more devices "out of the box" than any other operating system ever has. Yes, even FreeBSD.

The other key highlight of this talk was:

Closed source Linux kernel modules are illegal.
Closed source Linux kernel modules are unworkable.
Closed source Linux kernel modules are unethical.

So who the hell is this guy? He's Greg Kroah-Hartman. Who the hell is that? He's a kernel developer. His name appears 149 times in my kernel sources (Ubuntu patched, 2.6.15). And, perhaps more tellingly it appears at the top of the files:

    drivers/pci/pci-sysfs.c and
    drivers/pci/search.c

both of which contain many functions which are called from functions in this file:

    NVIDIA-Linux-x86-1.0-8776-pkg1/usr/src/nv/os-inter face.c

What's that? It's wrapper for the closed source NVIDIA kernel module. What license is that under? The NVIDIA Software License. It's basically a proprietary EULA with a redistribution (without modification) exception for distros. It sure aint the GPL, or "as free" as the GPL (which is techically what the GPL requires for derived works).

So Greg.. why don't you sue them? You've made your position clear, fight them. If you havn't got the money, contact the FSF, assign your copyright to them, get them to fight. Given the choice between opening their source code or not being able to distribute their software at all, NVIDIA will choose to open their source code. How can I be so sure? Cause people buy their chipsets to integrate into things like set top boxes and other devices that run Linux. They need that embedded market, that's why they released the drivers in the first place. The problem is that no-one is making them choose.

anandtech's review of the Core 2 Duos indicated that power consumption was really about the same across the board. The AMD EE chips were the least-power hungry, with just about every other chip all in a small span. Methinks someone's test procedures aren't quite accurate

A Tech Report article on AM2 chipsets shows that the nForce 590 chipset consumes about 20 watts more than the ATI RD580 (CrossFire XPress 3200) chipset at load. However, the power-hungry nForce 590 chipset has more integrated features: 2 gigabit ethernet chips (RD580 has none), 6 SATA ports (RD580 has 4), and more RAID options. That's probably why most high-end CPU reviews I've seen have used the nForce 590 chipset.

An Anandtech article shows that the Intel 975X chipset consumes about 3-5 watts more than the more current P965 chipset for certain apps, but unfortunately they don't have "load" comparisons. The P965 chipset lacks SLI support, so that's probably why reviewers are using 975X.

So a comparison of platforms that "most people buy" should probably use the Intel P965 chipset and the nForce 570 Ultra chipset. It's too bad review sites tend to use the SLI uber-chipsets.

This has been a big complaint of mine for a long time. We all know that under Windows, NVidia has had this PowerMizer thing that allows you to lower GPU power consumption. And we all know about Coolbits, which allows you to over/underclock your GPU in both Windows and desktop linux.

Now here's the thing. Both Coolbits and PowerMizer are disabled for mobile GPUs under Linux. So when you're not needing full 3D performance, that NVidia card is sitting there sucking up your laptop battery power. Might as well load up Beryl and go nuts.

Oh, and nvclock does not work on all GPUs. My mobile QuadroFX GPU, for example, seems particularly immune to it. When I run it nvclock seems to think it is underclocking my GPU, but I see no effects whatsoever---no reduction in the GPU temperature, for example, which is constantly sitting at 85C. Any other suggestions?

I keep on hearing about how bad the binary linux ATI drivers for linux are, but hey, at least they've got their PowerPlay (GPU downthrottler) thing working in Linux. Maybe for my next laptop I'll consider an ATI card.

New systems are staring to come with mxm slots
The 24-inch I-mac uses one.
http://en.wikipedia.org/wiki/MXM
http://www.mxm-upgrade.com/
http://www.nvidia.com/page/mxm.html

you can take their TNT2 card and still get it to work with their -current- drivers (even on Linux!)

As demonstrated below, (one "single-consistent idea" can still be broken into logical pieces:

Integrating CPU with the GPU? That would be cool, however in the long term I don't really see it happening, or most anything else mentioned here. Rather this 'fusion' of technology will most likely just end up as a new processor, based around a concept similiar to Cell, but probably without an emphasis on the differant 'processing units'. We'll probably end up with a new cpu design, hopefully x86 compatible, (although if they could create a really really good cpu, I mean the next big thing, they really shouldnt feel like they have to work with the x86 ISA), with all these smaller units (not cores) doing everything.

Think of the latest 8800 GTX, it has 128 unifed 'Streaming Processors',that were previously known as pixel shader units (or a variation thereof). Graphics cards also have/had vector shader units, and the ability to use geometry shaders, however these (at least for g80) are all unified now, in the sense that these 128 'Streaming Processors' (notice the name change AND it's significance) can on-the-fly change their function to either of the three categories. Cool idea for the GPU, but it also hints at where things are going. The Voodoo2 is more of a 'graphics processor' than g80, IMHO, in the sense that gpus are just essenestially becoming 128 'tiny'-cored general purpose coprocessors, that area really good at floating point math, (essential for graphics).

As far as today's graphics cards go we can run C code on these things. Sounds more like an auxillary, differant, type of CPU, that's all today's graphics cards are. check it out... http://developer.nvidia.com/object/cuda.html/ [nvidia.com] So the wheel of reincarnation has started. It began with GLQUAKE, with an addon card (that was an addon to your 2D card...) and will now end with either a CPU w/ GPU (fusion) or even better, a CPU that incorporates processing concepts from the GPU into a new design that is really really good, especially in the areas needed for Quake VI.

PPSh-41: Well still have pci-e express, and better still we'll probably have PCIe x16 slots too. Whether or not the consolidated market will still continue to sell graphics coprocessors is beyond my skill of intuition, however they wont stop selling them for at least 6 years. Another possible scenario is an Intel/NVIDIA buyout (I like nvidia, dont buy them out, in fact just establish a partnership). In that case, I'd venture to say that with each cpu/gpu maker in their own business couplings the GPU as we know it (as of right now) is dead.

Remember what tim sweeney said, "something something blah blah blah, what we'll find is that [graphics] is a problem that turns out was just general computing". That definatly seems like the case, when you consider the ps3's graphics (which are probably good, havent checked it out, not that I'd buy a console). If a new (more than just an extention to the x86 spec) cpu comes out of all this technology, I really really doubt well be using a GPU, as it'd be kinda like just having another CPU, but in a pci slot, with a two slot cooler, loud noise, more cords, bigger cases, more heat etc etc. Provided the new cpu specification can be used buy both companies to make compatible cpus, well all see the extinction of the GPU, as well as numerous benefits mentioned a second ago. So bottom line, and in my best hopes, a Slashdot reading Dual Core, Dual GPU fat person, will eventually have a new box in 2012, right before the world ends. It will be running GNU/Linux), and will no more than the size of a mid tower. However the CPU in the system will be a really really fast, and will have hundreds of the 'streaming processors' (that are the basis of out latest gpu's processing ability) that can be used for games, graphics, or anything else, or even porn.

Seriously though, you did just hit HTML formatted by accident. Come on you're amongst men now, you can admit it bravely...

Integrating CPU with the GPU? That would be cool, however in the long term I don't really see it happening, or most anything else mentioned here. Rather this 'fusion' of technology will most likely just end up as a new processor, based around a concept similiar to Cell, but probably without an emphasis on the differant 'processing units'. We'll probably end up with a new cpu design, hopefully x86 compatible, (although if they could create a really really good cpu, I mean the next big thing, they really shouldnt feel like they have to work with the x86 ISA), with all these smaller units (not cores) doing everything. Think of the latest 8800 GTX, it has 128 unifed 'Streaming Processors',that were previously known as pixel shader units (or a variation thereof). Graphics cards also have/had vector shader units, and the ability to use geometry shaders, however these (at least for g80) are all unified now, in the sense that these 128 'Streaming Processors' (notice the name change AND it's significance) can on-the-fly change their function to either of the three categories. Cool idea for the GPU, but it also hints at where things are going. The Voodoo2 is more of a 'graphics processor' than g80, IMHO, in the sense that gpus are just essenestially becoming 128 'tiny'-cored general purpose coprocessors, that area really good at floating point math, (essential for graphics). As far as today's graphics cards go we can run C code on these things. Sounds more like an auxillary, differant, type of CPU, that's all today's graphics cards are. check it out... http://developer.nvidia.com/object/cuda.html/ So the wheel of reincarnation has started. It began with GLQUAKE, with an addon card (that was an addon to your 2D card...) and will now end with either a CPU w/ GPU (fusion) or even better, a CPU that incorporates processing concepts from the GPU into a new design that is really really good, especially in the areas needed for Quake VI. PPSh-41: Well still have pci-e express, and better still we'll probably have PCIe x16 slots too. Whether or not the consolidated market will still continue to sell graphics coprocessors is beyond my skill of intuition, however they wont stop selling them for at least 6 years. Another possible scenario is an Intel/NVIDIA buyout (I like nvidia, dont buy them out, in fact just establish a partnership). In that case, I'd venture to say that with each cpu/gpu maker in their own business couplings the GPU as we know it (as of right now) is dead. Remember what tim sweeney said, "something something blah blah blah, what we'll find is that [graphics] is a problem that turns out was just general computing". That definatly seems like the case, when you consider the ps3's graphics (which are probably good, havent checked it out, not that I'd buy a console). If a new (more than just an extention to the x86 spec) cpu comes out of all this technology, I really really doubt well be using a GPU, as it'd be kinda like just having another CPU, but in a pci slot, with a two slot cooler, loud noise, more cords, bigger cases, more heat etc etc. Provided the new cpu specification can be used buy both companies to make compatible cpus, well all see the extinction of the GPU, as well as numerous benefits mentioned a second ago. So bottom line, and in my best hopes, a Slashdot reading Dual Core, Dual GPU fat person, will eventually have a new box in 2012, right before the world ends. It will be running GNU/Linux), and will no more than the size of a mid tower. However the CPU in the system will be a really really fast, and will have hundreds of the 'streaming processors' (that are the basis of out latest gpu's processing ability) that can be used for games, graphics, or anything else, or even porn.

It's straightforward to do this for 3D games, because the system has real depth information. Just use shutter glasses and render alternate frames with the viewpoint shifted by one eye separation distance. That's easy, and looks good if the system can render upwards of 70 fps.

Nvidia already offers stereo 3d drivers which work great with my eMagin 3d visor. As you say the games already have Z depth information to render a realistic environment, which the software just renders from a slightly offset perspective. I believe it is 30Hz from each perspective for a combined 60Hz. Though, the HUD and crosshairs could use a little work in some games, since the game creators always give them depth. But if it is distracting then they can usually be disabled if you can't get them to line up properly with the 3d background.

The best stereo 3d effects are in games which have a lot of close at hand objects. Looking down the barrel of the gun in Call of Duty for instance is pretty darn cool. Or climbing up a ladder where the ladder looks like it has real depth. I'd say virtual objects out to about a virtual 10 feet or so, just like in the real world, are really compelling in 3d.

What nVidia's been doing for years then? http://www.nvidia.com/object/3d_stereo.html for more info, but I remember having a pair of these back in the GeForce 3 days. It would work with any 3D game, though how good it looked depended on the game, my guess is it had to do with how it handled the 3D data. The most impressive were (of course) the nVidia demos which could actually appear to jump out of the screen a bit.

Looks like these days they aren't doing their own glasses, but rather have drivers that will work with 3rd party stuff.

So sounds like this is not only nothing but hype, but it's not even original in any way.

actually with the nvidia purevideo codec and the right drivers playback for blu-ray and hd-dvd is currently available http://www.nvidia.com/page/purevideo_hd.html

1) Add universal repositories in Synaptic package manager.
2) Type this in in the terminal:
sudo apt-get install nvidia-glx nvidia-kernel-common
sudo nvidia-glx-config enable
3) Type Ctrl-Alt-Backspace to restart your display, or reboot if you prefer.

Of course, you might want to wait until the root exploit in NVIDIA's driver is fixed in a non-beta release.

Already done.

...the best that an attacker can accomplish is a DoS for as long as you are visiting that site...

Then perhaps you can explain why this isn't a working javascript exploit proof of concept:
(Taken from a post further down this very page)

http://nvidia.com/content/license/location_0605.as p?url=';a='a';i=18;while(i--)a%2B=a;location=a;//

I mean... if the overflow is that easy, wouldn't someone adept at hitting the right targets in memory be able to do a lot worse with nothing more than javascript?

and it seems nvidia driver download chooser is broken. go to
http://www.nvidia.com/content/drivers/drivers.asp, choose "graphics driver" -> "geforce and tnt2" -> linux ia32 (or amd64). wtf ? why a redirect to some quadro page ?

going to linux driver page at http://www.nvidia.com/object/unix.html also shows 1.0-8774 as the latest for all supported oses/archs, so no, it is not fixed.

hi, nvidia. could you please provide open specs ? thanks.

and it seems nvidia driver download chooser is broken. go to
http://www.nvidia.com/content/drivers/drivers.asp, choose "graphics driver" -> "geforce and tnt2" -> linux ia32 (or amd64). wtf ? why a redirect to some quadro page ?

going to linux driver page at http://www.nvidia.com/object/unix.html also shows 1.0-8774 as the latest for all supported oses/archs, so no, it is not fixed.

hi, nvidia. could you please provide open specs ? thanks.

That version seems to have come out of beta 3 days ago:
http://www.nvnews.net/vbulletin/showthread.php?t=7 8188
http://www.nvidia.com/object/linux_display_ia32_1. 0-9626.html

as well as the 1.0-9626 QuadroPlex driver:
http://www.nvidia.com/object/linux_display_ia32_1. 0-9626.html
http://www.nvidia.com/object/linux_display_amd64_1 .0-9626.html

Thanks

as well as the 1.0-9626 QuadroPlex driver:
http://www.nvidia.com/object/linux_display_ia32_1. 0-9626.html
http://www.nvidia.com/object/linux_display_amd64_1 .0-9626.html

Thanks

Actually Nvidia updated the driver last week and it is now at 9626.
http://www.nvidia.com/object/linux_display_ia32_1. 0-9626.html

And the first comment in the linked KernelTrap reports that this problem was fixed. I'm not sure if that is true though as I haven't verified it myself.

nvidia does have a cool looking box for video cards we should be looking to things like that. I think a Monitor stand with video cards in it is a cool idea.
http://www.nvidia.com/page/quadroplex.html

Intel chips sets are still behind Nvidia and Ati. The best that Intel has is x8 x8 or x1 x16 while Nvidia has x16 x16 one 590 board even has x16 x8 x16 and most 590 boards have with duel gig-e with ip/tpc offload. These are for amd 590 as the Intel ones are not out yet.
This also shows up in the workstation / severs chips sets as well. Aka the power Mac g5 has more pci-e lanes then the Mac pro and it has less bandwidth in the chip set to chip set link.
Also looking at Motherboards form super micro

Dual AMD® Opteron(TM) 2000 Series (Socket F) Support, 1000 MHz HyperTransport Link
2 (x16), 1 (x8) & 1 (x4) PCI-e, 1 64-bit 133MHz PCI-X, 1 64-bit 100MHz PCI-X
      Up to 16GB DDR2 667 SDRAM (or)
        Up to 16GB DDR2 533 SDRAM (or)
        Up to 32GB DDR2 400 SDRAM
  Dual-port Gigabit LAN/Ethernet Controller
  2-Channel Ultra320 SCSI with Zero-Channel RAID support

  Dual Intel® 64-bit Xeon® Support (667 / 1066 / 1333MHz FSB)
  Up to 32GB DDR2 667 & 533MHz FB-DIMM
  1(x16) & 1(x4 in x16) PCI-E, 2 64-bit 133MHz & 1 64-bit 100MHz
      PCI-X, 1 32-bit PCI
  Dual-Channel Ultra320 SCSI & Zero-Channel RAID Support
  Dual-port Gbit LAN

The amd board has a lot pci-e slots

Tyan

  (2) AMD Opteron(TM) (Rev.F) 2000 series processor support (1207-pin)
  (8) DDR2 DIMMs sockets; up to 32GB reg. DDR2 400/533/667 mem.
  Supports ECC memory moduels; dual channel memory bus
(4) PCI-E x16 slots
      - (1) x16 signal from IO55
      - (1) x16 from MCP55
      - (1) x16 from MCP55 with x8 signal
      - (1) x16 from IO55 with x8 signal
    (2) PCI-X 100MHz slots from NEC nPD720404
          or (1) PCI-X 133MHz slot if 133MHz card is used
    (1) PCI v2.3 32-bit 33MHz slot
    (7) Expansion slots total
  (6) SATA2 ports (3.0Gb/s), (8) SAS ports (opt.), and (2) GbE LAN ports
  (1) 1394a FireWire port and integrated audio
  SSI/Extended ATX footprint (13" x 12"; 330.2mm x 304.8mm)

Tyan does not any xeon workstation board with a full x16 slot.

To running something like this http://www.nvidia.com/page/quadroplex.html at the max you need 2 full x16 links.
http://www.nvidia.com/page/nforce_pro_workstation. html

intel is missing out on the high end workstation graphics market.

Intel chips sets are still behind Nvidia and Ati. The best that Intel has is x8 x8 or x1 x16 while Nvidia has x16 x16 one 590 board even has x16 x8 x16 and most 590 boards have with duel gig-e with ip/tpc offload. These are for amd 590 as the Intel ones are not out yet.
This also shows up in the workstation / severs chips sets as well. Aka the power Mac g5 has more pci-e lanes then the Mac pro and it has less bandwidth in the chip set to chip set link.
Also looking at Motherboards form super micro

Dual AMD® Opteron(TM) 2000 Series (Socket F) Support, 1000 MHz HyperTransport Link
2 (x16), 1 (x8) & 1 (x4) PCI-e, 1 64-bit 133MHz PCI-X, 1 64-bit 100MHz PCI-X
      Up to 16GB DDR2 667 SDRAM (or)
        Up to 16GB DDR2 533 SDRAM (or)
        Up to 32GB DDR2 400 SDRAM
  Dual-port Gigabit LAN/Ethernet Controller
  2-Channel Ultra320 SCSI with Zero-Channel RAID support

  Dual Intel® 64-bit Xeon® Support (667 / 1066 / 1333MHz FSB)
  Up to 32GB DDR2 667 & 533MHz FB-DIMM
  1(x16) & 1(x4 in x16) PCI-E, 2 64-bit 133MHz & 1 64-bit 100MHz
      PCI-X, 1 32-bit PCI
  Dual-Channel Ultra320 SCSI & Zero-Channel RAID Support
  Dual-port Gbit LAN

The amd board has a lot pci-e slots

Tyan

  (2) AMD Opteron(TM) (Rev.F) 2000 series processor support (1207-pin)
  (8) DDR2 DIMMs sockets; up to 32GB reg. DDR2 400/533/667 mem.
  Supports ECC memory moduels; dual channel memory bus
(4) PCI-E x16 slots
      - (1) x16 signal from IO55
      - (1) x16 from MCP55
      - (1) x16 from MCP55 with x8 signal
      - (1) x16 from IO55 with x8 signal
    (2) PCI-X 100MHz slots from NEC nPD720404
          or (1) PCI-X 133MHz slot if 133MHz card is used
    (1) PCI v2.3 32-bit 33MHz slot
    (7) Expansion slots total
  (6) SATA2 ports (3.0Gb/s), (8) SAS ports (opt.), and (2) GbE LAN ports
  (1) 1394a FireWire port and integrated audio
  SSI/Extended ATX footprint (13" x 12"; 330.2mm x 304.8mm)

Tyan does not any xeon workstation board with a full x16 slot.

To running something like this http://www.nvidia.com/page/quadroplex.html at the max you need 2 full x16 links.
http://www.nvidia.com/page/nforce_pro_workstation. html

intel is missing out on the high end workstation graphics market.

nVidia's latest driver's for Vista include an "OpenGL driver for compatibility testing.". OpenGL won't be a problem, it'll be provided by third party drivers like it always has been, there just won't be any software fallback provided by MS (OpenGL in software is useless anyway).

As for DRM, well. Nothing in Vista itself is going to prevent you from copying DVD's, software or music or any other such thing. Windows Media files will still be protected of course, and HDCP will HAVE to be built into all HD-DVD/Blu-ray drives and decoders (read: the hardware) for you to watch this material.

The DRM issue with respect to Vista is all mythic. The only true rights taken away from you in Vista compared to XP are in the 64bit (x64) edition, under which, you cannot install unsigned drivers (unless you add an option to the Vista bootloader which isn't that difficult).

(Full disclosure: I work for a major manufacturer of 3-D accelerators.)

There's lots of good sites that talk about GPGPU. Wikipedia has an okay article on the subject as well, and NVIDIA has a primer (PDF) on the subject. But the summary of this article is a bit overly broad.

GPGPU isn't about moving arbitrary processing to the GPU, rather it's about moving specific, computationally expensive computing to the massively parallel GPU.

Effectively, the core idea of GPGPU solutions is that you compute 256x256 (or another granularity) of solutions entirely in one pass.

NVIDIA has several examples on their website, specifically the GPGPU Disease and GPGPU Fluid samples. The Mandelbrot computation they have there could also be considered an example. (More samples here).

GPGPU has already been utilized to perform very fast (comparable to the CPU) FFTs. In an article in GPU Gems 2 (a very good book if you're interested in doing GPGPU work), they indicate that a 1.8x speedup can be had over performing FFTs on the CPU. I've heard that there are now significantly faster implementations as well.

(Full disclosure: I work for a major manufacturer of 3-D accelerators.)

There's lots of good sites that talk about GPGPU. Wikipedia has an okay article on the subject as well, and NVIDIA has a primer (PDF) on the subject. But the summary of this article is a bit overly broad.

GPGPU isn't about moving arbitrary processing to the GPU, rather it's about moving specific, computationally expensive computing to the massively parallel GPU.

Effectively, the core idea of GPGPU solutions is that you compute 256x256 (or another granularity) of solutions entirely in one pass.

NVIDIA has several examples on their website, specifically the GPGPU Disease and GPGPU Fluid samples. The Mandelbrot computation they have there could also be considered an example. (More samples here).

GPGPU has already been utilized to perform very fast (comparable to the CPU) FFTs. In an article in GPU Gems 2 (a very good book if you're interested in doing GPGPU work), they indicate that a 1.8x speedup can be had over performing FFTs on the CPU. I've heard that there are now significantly faster implementations as well.

(Full disclosure: I work for a major manufacturer of 3-D accelerators.)

There's lots of good sites that talk about GPGPU. Wikipedia has an okay article on the subject as well, and NVIDIA has a primer (PDF) on the subject. But the summary of this article is a bit overly broad.

GPGPU isn't about moving arbitrary processing to the GPU, rather it's about moving specific, computationally expensive computing to the massively parallel GPU.

Effectively, the core idea of GPGPU solutions is that you compute 256x256 (or another granularity) of solutions entirely in one pass.

NVIDIA has several examples on their website, specifically the GPGPU Disease and GPGPU Fluid samples. The Mandelbrot computation they have there could also be considered an example. (More samples here).

GPGPU has already been utilized to perform very fast (comparable to the CPU) FFTs. In an article in GPU Gems 2 (a very good book if you're interested in doing GPGPU work), they indicate that a 1.8x speedup can be had over performing FFTs on the CPU. I've heard that there are now significantly faster implementations as well.

nVidia now has a "Legacy" supported release for older cards. Ubuntu 6.06 has a "nvidia-glx-legacy" package that supports TNT through Geforce FX 5900 or so.. Should be no problem to run that on a TNT2.

For machines with http://www.xubuntu.org/
http://www.nvidia.com/object/linux_display_ia32_1. 0-7184.html

Just grab the Desktop live CD, and give it a try. (live CD probably uses the open source nv driver, just to warn you)

The video card you chose on the Dell is not the same as video card you chose on the Mac. In spite of the names, the Mac version includes specialized hardware to run 3-D stereoscopic goggles with your video output. The Dell version does not include this. It is a specialty, niche add-on that seems to make the Dell similar in price to the Apple.

Not to mention that you started out with a budget box to even get this close. Start with a real Dell pro machine and see how close you can get. It is not even a competition.

But, does it run linux*? *have good drivers? Yes. Better than ATI at least. Not open source, but the drivers are good. My 7800GS works great. Q4 runs faster under Linux than under Windows here.

Exploding triangles... last weekend it got so bad it was happening on all games. I checked and sure enough, even the PCB was getting warm and I'd done about all I could think of to keep the card cool. So the card was literally frying itself to death. Months of research on the web seeing lots of similar complaints but no solutions. Recently I found a post that pointed out the specific two chips that are the cause (see http://forums.nvidia.com/index.php?act=Attach&type =post&id=2222 for an example of triangles). They were overclocked in manufacture, *but* not the rest of the card. Which explains to me why others complained about their replacements. Well, I've finally chucked it.

Goodbye eVGA! You killed your brand equity to me.

some thing like this with pci-e and HyperTransport links and this also moves the heat form the video cards out of the case.
http://www.nvidia.com/page/quadroplex.html

Nice that they're prepackaging all this and supplying the drivers and all. But this seems to me to be a 6150 Mobo and case w/ remote. Good Mobo (use a 6100 for 2 Linux WoW machines), though sound was a pain to configure (ended up getting a external card).

Not sure how much the case is worth but the mobo is only $60-$80. Still guess it may be a good deal for those that don't wanna mess too much getting their own drivers and tweaking them properly.

Look up "inverse telecine". Film material (24fps) transmitted in 1080i is nearly always hard-telicined 1080p source. IVTC allows seamless reconstruction of original 1080p frames. Video material (30fps) transmitted in 1080i cannot be inverse telecined, but can be "decombed", which involves weaving even and odd scanlines, a process which does degrade picture quality by 25-30%. There is but a handful of consumer television sets on the market with 1920X1080 native resolution, and none can perform ivtc in real time as yet. You can, however watch broadcast transport stream caps with a dscaler or mplayer ivtc filter if you have a sufficiently fast CPU (or GPU -- see for instance nvidia's purevideo product comparison specs here). The difference in picture quality will be significant.

Quadro FX 4500 X2

Quad NVIDIA SLI Technology
Windows XP/2000

ForceWare Release 90
Version: 91.45
Release Date: August 9, 2006
Please make sure to read the Driver Installation Hints Document before you install this driver.

U.S. English
File Size: 32.8 MB
http://www.nvidia.com/object/winxp_2k_91.45.html

For all the system builders, I found the chart where I got information about hardware acceleration for nVidia cards: http://www.nvidia.com/page/purevideo_support.html

Everything from the 6200 TC 256MB on up has MPEG-2 and H.264/MPEG-4 decoding hardware acceleration. The 7xxx Series adds deinterlacing, scaling, and WMV9 acceleration.

Start off with the latest and greatest ATI All-in-Wonder. That can cost at least $300, usually more towards $500. Sorry, Nvidia can't compete with ATI in the multimedia realm. Not yet, anyways. People are going to want to play games, and impress their friends. And you need that video input/output functionality. Sure, you could use seperate cards, but this solution is more elegant.

Perhaps this will help you all...

CLICKY

Just because this was one of the first borderline on-topic posts in this discussion so far...

Can anyone explain to me how you'd use this thing? The Register article says "The machines are controlled from a PC or workstation - 32- or 64-bit, Windows or Linux - connected across a network. Nvidia reckons the boxes will interest not only content creators but folk doing scientific modelling and simulation work." That makes it sound almost like a standalone system of some sort, not just a graphics card. I mean, what kind of 'network' do you use to connect the host PC to an array of graphics cards? Usually you attach a graphics card to the PCI or AGP bus, which obviously has far more bandwidth than any networking connection I've ever heard (except for maybe some high-end clustering/HPPC interconnects). They must be using some more loose definition of network than I'm thinking, unless you deliver the data that's to be rendered to the unit in some highly-compressed, extremely high-level form, almost like a remote desktop protocol, and then it renders and displays them. That could make sense for vector graphics, where you can describe what's being displayed with only a few equations and it's really the rasterization and display that's hard, but that wouldn't fly for video and raster graphics.

I'm just going to go on the assumption that the product isn't quite that revolutionary, and the Register article is using the term 'network' a little more loosely than I'd think is particularly proper.

NVidia's page on the Quadroplex mentions on the System Requirements sub-page that a PCI Express x16 slot is required in the host system, and that each system can have a maximum of two Quadroplexes attached. No mention of any 'network;' just PCI express. I can only guess that it comes with some type of dummy header-card that you insert into the machine, which breaks out the PCI Express connection externally, so that you can connect it up to the big external box with the graphic cards in it? Is it that simple? Nothing but a box with a power supply and an external cable and a breakout card?

I've basically described two extremes of possibility. On one hand, the Register article might lead one to speculate wildly that this thing is some sort of purpose-built rasterization server, while NVidia's own site make it seem like merely two graphics cards in a pretty external case, hardly justifying the price they're asking even considering that they're probably high-end cards. I can only guess that the truth lies somewhere in the middle.

Anyone want to clue me in on exactly what this thing does and how it does it?

negative.
only 2-4 cores.
http://www.nvidia.com/page/quadroplex.html

RTFA (or, better yet, go to NVidia's site to verify). It's not 64x, it's 2x or 4x depending on the configuration. Stupidest sensationalism ever! Here is the link for the lazy.

Ahem...

ATI Linux Page
nVidia Linux Page
IBM Linux Drivers (for a random chipset)
VIA Linux Drivers (for a random chipset)

...and your point was?