Slashdot Mirror

Sure they do. https://devtalk.nvidia.com/default/topic/522788/gtx-660-with-3-monitors/

Or how about a non-competitive clause that would have prevented them from being a direct competitor against Sony?

CUDA has been out for a couple of years now, and has gotten to be quite stable and has grown a good ecosystem of tools. As for python? Knock yourself out

> Nitpick: that's not a technique. Those rays of light are called godrays, it says nothing about the implementation technique.

Indeed. They could be using "Volumetric Light", "Occlusion Stencil", or as a post-process in Screen Space. Hard to tell which algorithm they are using.

Reference:
* http://http.developer.nvidia.com/GPUGems3/gpugems3_ch13.html

+1

Tell him to write a game. One per week would be a good start. It doesn't have to be releasable, just written. Something simple, like tic tac toe, or solitaire would be a start. Then progress to things that require collision detection and crude 2d physics and animation.

For good graphics, learn OpenGL.
There's a massive quantity of tutorials at NeHe Productions.
With OpenGL being used on tablets and web interfaces, it's the way to get an edge.

For bad graphics, take a look at Dwarf Fortress.
Even with text-only graphics that game brings in tens of thousands per year.

Read Programming Pearls and Beautiful Code for unique ideas for approaching a problem. Take a look at the Fast Inverse Square Root for a good example of combining bit-level hacking with Calculus.

Nvidia made the GPU Gems series available online. It has advanced concepts, but is currently down. Subsurface scattering is an amazing effect, as is Navier-Stokes simulations.

MIT has a course on algorithms with video lectures.

If none of that piques his interest, then he shouldn't be a programmer.

I have been using notebook drivers direct from Nvidia for quite some time; at least since I bought the laptop prior to my current one, which would have been in late 2008. Maybe it was different for the Quadro, but they have had mobile GeForce drivers available for download for years. I don't know if this will work for you, but here are the Quadro Notebook Drivers v310.90 dated yesterday.

Whether you can install the latest and greatest drivers from NVidia, or have to download from your OEM depends on the OEM. They usually tweak some settings (PCI Device ID) specifically so that the default NVidia drivers will not install. They do this for support purposes. You can modify the INF for the Nvidia drivers and force them to install, but it can be a pain to get them working right sometimes. For instance, when Vista came out, I bought a new laptop. Dell restricted the drivers so I couldn't use the Windows XP drivers for my video card. I had to find this solution to be able to run XP on the device.

I have been using notebook drivers direct from Nvidia for quite some time; at least since I bought the laptop prior to my current one, which would have been in late 2008. Maybe it was different for the Quadro, but they have had mobile GeForce drivers available for download for years.

I don't know if this will work for you, but here are the Quadro Notebook Drivers v310.90 dated yesterday.

I only bought nVidia for last decade+. But recently their drivers also suffer from minor issues.

https://devtalk.nvidia.com/default/topic/526044/linux/x-server-1-13-1-deadlocks-randomly-on-geforce-gtx680/

https://devtalk.nvidia.com/default/topic/525640/linux/evga-geforce-gtx-660-ti-problems-/

Maybe higher performance and better support than competition, but nVidia is not flawless either.

I only bought nVidia for last decade+. But recently their drivers also suffer from minor issues.

https://devtalk.nvidia.com/default/topic/526044/linux/x-server-1-13-1-deadlocks-randomly-on-geforce-gtx680/

https://devtalk.nvidia.com/default/topic/525640/linux/evga-geforce-gtx-660-ti-problems-/

Maybe higher performance and better support than competition, but nVidia is not flawless either.

Protien folding simulation is such a large and basic need globally there ought to be enough large scale interest to make development of specialized ASICs to deal with these problems cost effective and exceedingly useful for all who need to do these simulations. A quick check of google shows such chips do in fact exist with unbelivable performance figures which kick the snot out countless tens of thousands of CPU/GPUs. There is no shortage of funding for medical research so it begs the question why waste CPU/GPU resources on folding simulations?

I still do seti and milkyway at home because there are no resources allocated for seti and milkyway at home is interesting to me personally.

First of all, protein folding is not the only thing they do, the Folding@HOME infrastructure is used by many for a variety of bio-molecular studies.

Secondly, custom ASIC-based machines like Anton and MDGRAPE (which are AFAIK the only such machines around these days) consist of much more than a custom-chip, they use specialized interconnects, memory, software, etc. and cost a lot. The MDGRAPRE-4, the coming version of the Riken-developed custom molecular simulation machine costs $10M + $4M (development + manufacturing) which poses serious financial limitations to it. Moreover, these specialized machines are only able to run a handful of molecular dynamics algorithms and while fast, they are nowhere near as versatile as general-purpose codes like AMBER, GROMACS, NAMD, etc. Although it is true that these specialized machines are a few orders of magnitude faster in terms of absolute performance (i.e time to solution and not Flops), due to their limitations and the way they are used, some researchers argue that they employ a "brute force" approach to molecular simulations which is not cost-effective from the point of view of science/$ delivered. I personally wouldn't call machines like Anton and MDGRAPE a complete waste, they achieve impressive advances in hardware, software, and science results in a specific direction: pushing the limits of how fast can one run a single simulation. There are certainly other (some would say better) ways to get amazing results with general-purpose (super)-computers be it using massive clusters or cycles donated to folding Foldging@HOME.

Finally, let me explain why is there compute-resource shortage in the (bio-)molecular simulation filed which will remain for the foreseeable future no matter how much money do various governamental and non-governamental agencies pour into it. Molecular dynamics is extremely compute-intensive, a single iteration of the MD algorithm requires 10^8-10^10 Flops (not LINPACK Flops!), repeated for millions of times during a single simulation of a bio-molecular system (and such a simulation can take weeks even on a big machine). And that's still a few orders of magnitude short of what would be needed to simulate timescales at which biological processes take place. Therefore, any compute-resource available can be harnessed for molecular simulation research and Folding@HOME does a decent job at utilizing donated cycles. Admittedly, there are some in the community who think that Folding@HOME is wasteful, but that's a topic for another discussion.

Disclaimer: I am involved in the development of the GROMACS open-source molecular simulation package which is in fact on of the computational engines used by Folding@HOME. Still, I believe I have not been biased in the way I presented Folding@HOME and molecular dynamics in general.

why don't you first convince Nvidia to make their drivers install the first time without issue.

Valve is working on this. The Linux version of Steam is in beta right now , and NVIDIA has been pumping out press releases about their partnership with Valve on Linux for the last few months.

For the last decade (literally), people have been saying that the only reason why there's no games for Linux is because it's a chicken-and-egg problem - there's no games because there's no hardware support, and there's no hardware support because there's no games. Now Microsoft, with Windows 8 (and its built-in Steam competitor) has pissed off the Gaben, and I think things will change soon.

Valve throwing their weight behind it is probably going to make a lot of headway towards solving the problem within a few years; they have enough pull with both the chicken and the egg that they can encourage developers to support Linux, and hardware manufacturers to write proper drivers.

The Top500 reports actual performance as measured with LINPACK, hardware vendors report the theoretical performance of their chips, which in the case of GPUs is often quite a bit more than you'd be able to squeeze out with LINPACK.

For comparison: Tsubame 2.0 consists of 1400 nodes with approx. 4200 NVIDIA Tesla C2075, which should yield -- according to your estimate -- 2.1 PFLOPS (4200 * 0.5 TFLOPS), yet it is listed at 1.2 PFLOPS. So just add two years to your estimate and you should be fine...

Project Denver and it is ARMv8 and it uses transmeta technology and thus is a complete new core. This means NVIDIA has licensed with ARM to make cores, rather than simply purchasing IP from ARM Holdings.

The video driver for Linux is out: http://www.nvidia.com/object/linux-display-amd64-295.59-driver.html

Drivers are not the problem.

This has nothing to do with closed source. Here is the driver for the retina's video card: http://www.nvidia.com/object/linux-display-amd64-295.59-driver.html

"Windows is slowly losing relevance" - by peppepz (1311345) on Wednesday August 08, @02:22AM (#40915185)

#1 Most Used/Biggest Marketshare on PC Desktops + Servers combined, & it's "losing relevance"? Then MacOS X + Linux never had it @ all, just based on the numbers, & don't argue with me - as the saying goes, "argue with the numbers": See here, "Read 'em & weep" -> http://en.wikipedia.org/wiki/Usage_share_of_operating_systems

---

"OpenGL never caught on? I think you're talking from inside a Windows perception bubble, for everything but Windows is based on OpenGL nowadays" -

I've used AND created OpenGL screensavers for Windows since Windows 2000, XP, Server 2003 - based on the OpenGL 2.1 standard ( & that code STILL RUNS on Windows Vista/7/Server 2008, via OpenGL 'wrappers' that NVidia drivers provide)...

In fact?

---

PERTINENT QUOTE/EXCERPT:

http://developer.nvidia.com/opengl-driver

"OpenGL Driver Support - Windows driver version 305.53 and Linux drivers version 304.15.00.02 provide beta support for OpenGL 4.3 and GLSL 4.30 on capable hardware."

---

In fact, THIS is "still there" on Windows 7 (as it has been since 2000/XP/Server 2003/Vista), and it's quite "standards-current", per the above quote from NVidia:

C:\windows\system32\opengl32.dll

* Thus - it's been there, and IS STILL THERE, though DirectX is favored by Microsoft & DirectX display runs the AeroGlass desktop interface rather than OpenGL...

(DirectX controls more than display, far more - OpenGL 2.1 didn't & I coded around it, was easier to code than DirectX, I'll give it that, but it didn't control sound, display, joysticks + more, DirectX could & did)

APK

P.S.=> So, let's stop the "false propoganda" gents, ok? Thank-You...

... apk - by peppepz (1311345) on Wednesday August 08, @02:22AM (#40915185)

The drivers for gaming-related devices have got to be pretty awful on FreeBSD if nobody seems to use it at all for games, only servers

nVidia ships blob drivers for FreeBSD.

Yet they don't make the (blob) drivers for BSD either

Yes they do, at least for x86 and x86-64.

The way drivers work in Linux and BSD, aren't accommodating to "binary blobs" since they can be broken with even minor updates

FreeBSD guarantees a stable KBI across minor revisions, and we require strong justifications for breaking it between minor revisions (which means that often kernel modules will work between major revisions, we just don't guarantee it). After 10.0, we're looking at providing longer-term support for a subset of KPIs.

Perhaps not entirely coincidentally, "one month" is about the amount of time that nVidia's web forum - comically also the only route for reporting bugs, and found here - has been shut down due to a DDoS attack.

Probably not the best way to follow up their snippy response to Linus Torvald bashing their Linux support.

This is an imperfect solution which assumes that redrawing inside the window (which is the OpenGL texture) is done quickly enough. It also splits the system into two parts: the hardware-accelerated window compositing part and the unaccelerated drawing part.

More sophisticated solutions exist. Examples:
Using distance fields in OpenGL shaders for font rendering
Using textures for anti-aliased vector graphics rendering
Spline rendering in the GPU for non-affine vector graphics rendering (including text)

Such techniques assume the presence of a GPU. Forget about abstracting that away with a painter-based architecture. The fundamental problem is that the painter acts at a level that is too low. This is the reason for the higher-level QML scenegraph. As for the missing C++ bindings, thats debatable, but if some appear one day, expect a high level API. Nothing else makes sense.

It should be noted that the Enlightenment Foundation Libraries do something very similar with the combination of Evas and Edje.

Wii's AMD Hollywood GPU is roughly comparable in fillrate to a Radeon 9000, and the Xenos in the Xbox 360 is like a Radeon X1900. Which GPU on Tom's chart comes closest to the specs of a Tegra 3?

nVidia's specs page for the Tegra 3 are a bit on the light side with performance details. "3D Performance Relative to Tegra 2" is the only performance metric they give, which is, needless to say, completely useless when trying to compare them to any real 3D cards.

So, I checked with Wikipedia (which could very well be wrong).

Looking at the list of clock speeds and pixel shaders for the Tegra 3 graphics core and the other GeForce graphics cores, the fastest version of it appears to run at about the same clock speed and pixel shaders as the nVidia GeForce GT 120 (which is 3 graphics card generations old at this point).

Not only that, but both the Wii and Xbox 360 are on their way out. The Wii U is going to retail later this year. While the price is unknown, it's thought to be $299... the same price the Wii debuted at. Video specs are currently unknown, only that AMD is designing it.

The rumor mill has a new Xbox console pegged as coming out sometime in 2013. Rumor also has it that it will ship with dual-video cards (if you can call them that in consoles).

In addition, NVIDIA is working on a full release of Ice Cream Sandwich (ICS) for our Tegra 2 Dual Core and Tegra 3 Quad Core reference devices to be made available early next year. This release package will comprise of binary code, drivers developed on our reference platforms and any patches necessary to ensure stability & performance with ICS. We hope the open source community can leverage this release for their work on all Tegra devices.

If you really want to install Linux on the tablet, buy TI OMAP-based devices.

In addition, NVIDIA is working on a full release of Ice Cream Sandwich (ICS) for our Tegra 2 Dual Core and Tegra 3 Quad Core reference devices to be made available early next year. This release package will comprise of binary code, drivers developed on our reference platforms and any patches necessary to ensure stability & performance with ICS. We hope the open source community can leverage this release for their work on all Tegra devices.

If you really want to install Linux on the tablet, buy TI OMAP-based devices.

Hmm, seeing as nVidia doesn't have such a great track record with releasing source code in general (for which they deservedly received the Linus' finger) and a quick Google reveals that devs are still waiting for Tegra 2 source code, I wonder what drove them to choose the Tegra 3 for a "Nexus" branded device, which supposedly would mean a "developer friendly" device with all the features and hackability that the Nexus devices have come to be known for.

nVidia is about the only viable solution for Linux graphics.

So what will you put in your next computer, now nVidia has stopped releasing drivers?

When did they stop releasing drivers since I just looked on their site and see a release dated 15th June?

http://www.nvidia.com/object/linux-display-amd64-302.17-driver.html
http://www.nvidia.com/object/linux-display-ia32-302.17-driver.html

Or are you just a worthless troll making crap up?

nVidia is about the only viable solution for Linux graphics.

So what will you put in your next computer, now nVidia has stopped releasing drivers?

When did they stop releasing drivers since I just looked on their site and see a release dated 15th June?

http://www.nvidia.com/object/linux-display-amd64-302.17-driver.html
http://www.nvidia.com/object/linux-display-ia32-302.17-driver.html

Or are you just a worthless troll making crap up?

Apple might work with nVidia to release a driver through Apple's distribution channel but nVidia releases them separately

http://www.nvidia.com/object/unix.html

But Nvidia does make Linux drivers. He's just pissed that they're not open source.

http://www.nvidia.com/object/unix.html

"Top notch" might be overstating it a bit, there are still a few instances where they seem to be lacking. For instance, Windows users who have SLI and multiple monitors have been able to enable SLI and use both of their monitors at the same time since about 2008. But under Linux, no dice.

So if I had two monitors (which I do), and two Nvidia GPUs in SLI mode (which I do), and I wanted to run some 3D app that took advantage of SLI, I would have to: reconfigure X to disable my second monitor and enable SLI, restart X, play the game/use the app I wanted, when I was done I would have to reconfigure X again to enable my second monitor and disable SLI, restart X again, and reopen all my apps. Hardly ideal.

Here you go!

http://www.nvidia.com/docs/IO/40049/nViewDMGuide_v135-85.pdf

Minimum requirements: ftp://download.nvidia.com/XFree86/Linux-x86/295.20/README/minimumrequirements.html

[sorry for the poor formatting]
Software Element Supported versions Check With...
Linux kernel 2.4.7 and newer cat /proc/version
XFree86* 4.0.1 and newer XFree86 -version
X.Org* 1.0, 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 1.10, 1.11, 1.12 Xorg -version
Kernel modutils 2.1.121 and newer insmod --version
glibc 2.0 ls /lib/libc.so.* > 6
* It is only required that you have one of XFree86 or X.Org, not both. Sometimes very recent versions are not supported immediately following release, but we aim to support all new versions as soon as possible.

binutils 2.9.5 size --version
GNU make 3.77 make --version
gcc 2.91.66 gcc --version

The is GPU based computing but if your problem fits that paradigm then your're set: http://blogs.nvidia.com/2011/11/exascale-an-innovator%E2%80%99s-dilemma (No I don't work for nvidia)

Eventually, holes will not be necessary. Better SQUIDS + cuda = mind reading from a distance. By 2020 0r 2030 at the latest, I would conjecture.

Sounds like a brilliant design.

In many ways, it's simply a logical next step - see Nvidia's white paper for architectural details. http://www.nvidia.com/object/IO_90715.html

Thing is, we're so used to minimal innovation in the stagnant Wintel-controlled X86 world, the rapid pace of change in ARM systems is exciting. Imagine a beowulf cluster of them, for example...
http://arstechnica.com/business/news/2011/11/the-opposite-of-virtualization-calexdas-new-quad-core-arm-part-for-cloud-servers.ars

Why is it so wrong to use the proprietary drivers?

Here's an example. I bought a laptop with an Nvidia Quadro FX 3800M specifically for triple-head support (via docking station) and I can't even get dual-head to work properly due to an infinite loop somewhere in the binary drivers when mode-switching. I've done most of the investigation work, even running X through gdb and they're not interested in helping - not even some basic debug symbols.

I've been completely ignored by Nvidia via both of their official support channels: (1) the nvnews.net forum and (2) their linux-bugs@nvidia.com email address touted in their driver README. Even on Twitter.

What options do I have now with a binary driver except to run Windows (where triple-head works fine) or change hardware? I can't imagine why I would ever again support a company that can't even acknowledge my pain with their product, let alone help me to fix the problem myself.

The summary is pulled directly from the top of the article.

Here's the article from HPC Wire and some details from nvidia as well as the nvidia press release

The summary is pulled directly from the top of the article.

Here's the article from HPC Wire and some details from nvidia as well as the nvidia press release

1. Driver bugs. The worst of the bunch for me, so far, is this wonderful nVidia driver bug that rears its ugly head only when using Firefox or Internet Explorer (even with "hardware acceleration" turned off). The only workaround known at this time is to use any driver package earlier than 270.55. The latest beta driver (290.53) does not fix it. And to those considering responding to this point with "technical tips" -- read the thread, don't skim it, actually read it. Note that it's 35 pages.

2. Activation bullshit. Specifically: lack of a Volume Licensed Edition. Let me clarify what I'm talking about: Windows Vista and beyond did away with the concept of a VLK (Volume License Key), which is nothing more than a serial number that's associated with Windows XP Professional Volume Licensed Edition. (Meaning, you can't use a Retail key on a VLE install, an OEM key on a Retail install, etc. -- following so far?) VLE/VLKs mean no activation (no WPA), and they Just Work(tm). No activation. Instead, with Vista and Windows 7, each PC is required to have its own serial number, or you're forced to use a MAK or KMS. MAKs authenticate directly with Microsoft (so they have control over your license); a KMS is some piece of software running probably on a Windows 2008 Server box on your LAN which you have to purchase from Microsoft (and I have no idea what the licensing/stipulations are for getting one -- I'm sure you pay for licenses "in bulk" and the KMS probably talks to Microsoft somewhere along the lines too).

Why do I care about this? Simple: it's purely an anti-piracy effort that does nothing other than fuck tech-savvy users like myself (UNIX system administrator here, hi!) who often purchase hardware upgrades. I own legitimate copies of Windows XP Professional Retail (one for each system in my home), but I choose to use VLE simply to make my life easier and not have to deal with activation. The last time I used Retail, I happened to upgrade my RAM in my system from 2GB to 4GB, and was forced to call a telephone number + speak to someone in India and "justify" what I had done. Three months later I upgraded my video card; again, a phone call was needed. I am NOT going to do this every time I change something in my system; what if I don't like the hardware I just upgraded to, thus remove it? Oh, now I get to call Microsoft twice in one day! No thanks. Remember: Microsoft said this kind of behaviour would only happen if you changed more than 2 pieces of hardware at the same time, but that's obviously a lie. The proof is in the pudding.

Bottom line is: I paid for your OS, I therefore paid for a serial, and I should be able to run it on whatever hardware I wish. Otherwise, if you want to impose "one copy per PC" limitations to try and curb piracy, then please, PLEASE do it like how the folks at Alcohol Software do with their Alcohol 120% product (customers have a web page they can log in to and change which PC the license is associated with). If I had control over the situation (vs. making stupid phone calls and "justifying" what I'm doing with my own hardware/systems) then I'd be content with that.

3. Removal of "Windows Classic" themes such as Rainy Day. Sure, there's "Windows Basic" which does its best to "emulate" the look of 2K/XP, but they stripped out all of the themes which made it worthwhile. And yes, I have tried two Internet users' "Rainy Day" themes, both of which were horrendous.

4. Stupid UI design choices. For example, even in the most "basic" or "slim" of themes (Aero turned off, etc.), there is still an excessive amount of space wasted around application windows/borders. You can set the border size to 0 in Appearance, but you'll find 3-4 pixels of space still being used by who-the-fuck-knows-what. This combined with #3 effectively has removed a users' abi

The entire basis for this slashdot article is a Y-Combinator blog post from a poster with 3 Karma with 4 comments from 7 hours ago lamenting that there's no full package for developing for phones produced by the OS vendors. The "Boot to BASIC" thing is a retrocon issue more appropriately stated "and we wore onions on our belts" (a sin of which I am also guilty). While I applaud the /. rapid response on this critical issue the underlying assumption that these things don't exist is simply untrue. You'll find the Android SDK here. If you prefer an easier install - especially if developing for nVidia Tegra (but not exclusively) you will find that here. The equivalent iOS version is here. If you must have it, the Windows Phone SDK can be found here.

And yes, all of these things are for developing the app on a PC and deploying and testing on a phone or tablet. Much like the designers of landing gear for the 787 don't machine their prototypes out on the tarmac at Boeing field like their progenitors did, mobile app designers now prefer sitting in a comfortable workstation with a comfortable chair with decent compile times, multiple monitors, a reasonable keyboard and mouse, printing and debugging support rather than developing software on the device itself.

Booting to BASIC (or as was the case on the IBM 5100, APL) was very useful at that time because there were no apps - in some cases no media reader to store or load apps if you could get them on media. We got our apps by buying magazines or such, and manually typing the code in and debugging the typos. That day is long gone, and that's a good thing. Even Windows Phone has thousands of apps and a publicly available SDK - though for me the cyanogen and Android route would be more fun.

Now how did this manage to hit the front page of /.? I smell a rat.

Why is everybody thinking this is big news?
ftp://download.nvidia.com/CUDAOpen64/
The previous compiler, based upon Open64, has been available in source form since CUDA 1.0. They (partially) switched to LLVM in 4.1, and they also release the source code. They didn't have to, because unlike Open64 LLVM is not GPL, so it's nice of them, but it's not exactly earth-shattering news...

OpenCL is a much better path, because it can execute code on a CPU as well as a GPU.

So can CUDA, according to a graphic in one of the featured articles: "NVIDIA C or C++, PGI Fortran, or new language support, through LLVM-based CUDA compiler, to NVIDIA GPUs, x86 CPUs, and new processor support."

Well, They are making some of the best mobile/low-power solutions with the Tegra family of chip-sets.

I also believe that it's still going to take some time before the integrated solutions (Intel IGP and AMD Fusion) are good enough to replace discrete graphics for gamers - where they are strong today.

There is a new easier way to program GPUs now using Directives-based compilers.

Idea is that you add some high-level pragmas to your C or Fortran code that a parallelizing compiler
uses to map to the GPU accelerator. Of course, you have to expose parallelism in the code for
the compiler to do a decent job. Example, use more data-parallel data structures. But this is a nice
incremental way to take advantage of the GPU.

Check it out at:
http://www.nvidia.com/object/tesla-2x-4weeks-guaranteed.html?cid=dev

Sumit
NVIDIA - Tesla Group

Differences between GeForce and Tesla for compute are at:
http://www.nvidia.com/object/why-choose-tesla.html

Bottom line, GeForce is great for development, but if you want to build a cluster with GPUs,
you are better off using the commercial grade Tesla GPUs.

Sumit
NVIDIA Tesla Group

VHDL Cookbook is a good, though dated, intro.

Use ghdl to learn vhdl, without the need to have hardware, as it compiles VHDL to an executable. Icarus is similar, but for Verilog. gEDA has good tools, including the gtkwave waveform viewer. Combined, ghdl, Icarus and gtkwave are a pretty useful simulation suite. You can go a long way with simulation, since the normal design flow is to get the system 100% using simulation, then as a last step program the FPGA with maximal probability of it just working. As Bruce said, the actual partition, place and route tools are proprietary and specific to each FPGA vendor, and a google search will come up with a number of cheap FPGA boards.

Keep an eye on left field though. There is a convergence in progress between desktop CPU's, GPU's, parallel systems and FPGAs (which can be seen as an array of massively parallel simple processors). One day all I wrote may be obsolete and you will be able to program your FPGA in CUDA, or whatever results when mainstream programming figures out how to handle parallel systems properly.

VHDL Cookbook is a good, though dated, intro.

Use ghdl to learn vhdl, without the need to have hardware, as it compiles VHDL to an executable. Icarus is similar, but for Verilog. gEDA has good tools, including the gtkwave waveform viewer. Combined, ghdl, Icarus and gtkwave are a pretty useful simulation suite. You can go a long way with simulation, since the normal design flow is to get the system 100% using simulation, then as a last step program the FPGA with maximal probability of it just working. As Bruce said, the actual partition, place and route tools are proprietary and specific to each FPGA vendor, and a google search will come up with a number of cheap FPGA boards.

Keep an eye on left field though. There is a convergence in progress between desktop CPU's, GPU's, parallel systems and FPGAs (which can be seen as an array of massively parallel simple processors). One day all I wrote may be obsolete and you will be able to program your FPGA in CUDA, or whatever results when mainstream programming figures out how to handle parallel systems properly.

Nvidia project denver:

http://pressroom.nvidia.com/easyir/customrel.do?easyirid=A0D622CE9F579F09&version=live&releasejsp=release_157&xhtml=true&prid=705184

Are you sure? Might want to take a look at this.

It would help if I had read the summary. They've got a single server with eight Knight Ferry cards, each having 32 cores. That's where they get their 256 cores from. And they're calling the single server a "cloud".

What makes this most unimpressive is that nVidia has been making a GPU-accelerated real-time raytracing engine for years now (you can even download working demos), and before that they were selling a GPU-accelerated final-frame renderer (non-real-time raytracing). Intel is showing off in-house demos of stuff running on expensive hardware, while nVidia has been giving away the same thing to customers for years, and it's something that's actually out there that you can use. Heck, so far as I can tell, it's free.