Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

S.T.A.R.S. Officer — Sat, 14 Nov 2009 05:36:19 GMT

It makes sense to me.

I can't agree more that if Nintendo DID in fact release the numbers, majority of developers would have drop the Wii head first into a concrete casket.

But I can't help but think, does this mean as the Wii goes on, developers will be able to produce better graphics on the Wii in terms of textures and sprites? Honestly, SMG vs. SMG2, SMG2 looks way more detailed.

WaW vs. MWR, MWR is way more detailed in textures and the sprites are better.

I know graphics are supposed to improve over time, but what we've seen is already pretty much the cap on the Wii's capabilities. So we can expect to see is a little more...?

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

tapionvslink — Mon, 09 Nov 2009 23:34:10 GMT

Nintendo´s strategy is clear, they knew that GPGPU programming was to knew and expensive when they launched the Wii, and that there were few tools to ease the development using GPGPU software, that´s why they decided to concentrate their efforts in the wiimode at the beginning instead of trying to compete against 360 and ps3 in terms of graphics, since if they had taken that route they would have experienced something similar that happened to the ATARI Jaguar(64bits) that tried to eliminate the super nes an the sega Saturn from competition using hardware technology that was too new by that time.

http://en.wikipedia.org/wiki/Atari_Jaguar

If you have seen Angry Nerd´s review on ATARI Jaguar, I am sure you can tell that none game showed improved graphics over the SNES or the Sega Saturn games. Thats because the system was to complex to use.

If Nintendo had revelaed the secrets of Wii Hollywood at the start, I am sure that many people would have been dissapointed knowing the theory capabilities of Wii, so Nintendo probably thought that the best path to follow was to demostrate the Wii power by games and not by numbers, but unfortunately, Nintendo focused to much their system on the casual market, that the core market was almost forgoten.

It´s been 3 years since Wii was launched, and Nintendo earned a lot of profit since then, so, I do not think that hiding wii specs is necesary anymore, but the only way for nintendo to realize that soon enough is by listening to the people who own a Wii and realize that they already know the Wii true hardware arquitecture.

I have worked to hard to find what the Wii true specs, so please help me making the others of the wii comunity realize this.

Un purpose, if someone can post this in neogaf forums, I would be very happy.

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Imbalance — Mon, 09 Nov 2009 23:07:24 GMT

See, this is where I'm torn.

Ever since that old rumor that you could play Genesis games on NES with a few simple modifications, I've had a craving for standardized format in the industry. Y'know, like how anybody's DVD will play on any manufacturer's player? I had almost given up hope on the matter until Nintendo finally went to disc, but here we are, umpteen generations later, and we're not really any closer. In fact, where Nintendo is concerned, the industry is leaving them behind because of their coding stinginess. How many top-tier games gloss over the Wii because developers either fail at or don't want to gamble the expense of "dumbing down" graphics requirements for their games? Yet, the above goes a long way to confirm what many CONSUMERS already know: that the truth is that Wii exclusive titles would actually have to be dumbed down to be compatible with 360 or PS3, and that the developers have only proved how lax they have been in regards to rising to the challenge of producing games to fully utilize the system's capabilities.

With that in mind, I'll cross my fingers that maybe sometime before I get too old to care the industry will embrace the need to employ a universal media format (not DL-only, either) and that the innovative technologies should be much more broadly accessible. That goes for the next generation - though I doubt they're listening - but Nintendo needs to get their isolated heads out of their butts, and the other two need to wake up and realize that they've already stepped over the perception threshold (graphics) and can refocus on making good games as opposed to games that are just pretty.

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

S.T.A.R.S. Officer — Mon, 09 Nov 2009 22:30:16 GMT

Sooo.... we're gonna get better looking Wii games as time progresses?

I think I said that a million times on old GI.

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

SvenHudson — Sun, 08 Nov 2009 07:13:32 GMT

I read about the first screen's-worth of text and then I noticed how short my scroll bar was. Kinda scanned the rest.

So the point is basically that the Wii can out-perform its hardware's numbers, but developers are only just now finding out?

I guess that's pretty cool.

Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

srkelley — Sat, 07 Nov 2009 11:50:28 GMT

Basically with a gpgpu part of the gpu can be used like a processor. When used intelligently it can easily out perform comparable processors. It can't replace a processor as there are some things that it cannot process or do efficiently enough.

Just reading that makes me think that Zelda Wii will be a step above Super Mario Galaxy and Dead Space Extraction.

Good post, some parts were a bit too close but I got a lot of info out of it,

Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

S.T.A.R.S. Officer — Sat, 07 Nov 2009 00:27:49 GMT

Sooo... the reason why 3rd party games on the Wii aren't as good as they could be is because GPGPU is more expensive to develop on than a GPU? I tried my best to fully understand this, while it is interesting, it would be nice if TS could dumb it down a tinge for me...

Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

tapionvslink — Fri, 06 Nov 2009 21:12:13 GMT

I just want to tell you all the following article information dates from 1 year ago, but the information provided is so important that many of you will get interested in no time, not to mention that if you put close attention to the statements made by Dan Emerson, you will realize that the secret of the Wii Hollywood is that it has been a GPGPU all this time. The Wii Hollywood benefits from stream processing through the Floodgate stream processing engine of the Gamebryo 2.3 and later versions, and it is even possible that Nvidia CUDA is running inside the Hollywood.

Q&A: Emergent's Dan Amerson Opens The Floodgates
January 16, 2008
http://www.emergent.net/en/News/News-Ar ... n-Amerson/

Among the most important things mentioned about the Nintendo Wii and the Gamebryo 2.3 you will find:
"
What is Floodgate used for?

At its core, Floodgate is a stream-processing engine. Developers provide execution kernels that are just code that runs per element of a stream and specify the inputs and outputs to the kernel. Inputs can be fixed across all elements or can be a stream of elements. These items allow developers to build up tasks that Floodgate will execute asynchronously.

On top of that, developers can link tasks together by specifying outputs from one task as inputs to another task. Floodgate uses that information to schedule tasks allowing complex sets of behavior to emerge. Most importantly, all of this work occurs through a single API allowing developers to target a new platform just by recompiling their kernels and application code.
"

What is stream processing? .- http://en.wikipedia.org/wiki/Stream_processing
"
Stream processing is a computer programming paradigm, related to SIMD, that allows some applications to more easily exploit a limited form of parallel processing. Such applications can use multiple computational units, such as the floating point units on a GPU, without explicitly managing allocation, synchronization, or communication among those units.

The stream processing paradigm simplifies parallel software and hardware by restricting the parallel computation that can be performed. Given a set of data (a stream), a series of operations (kernel functions) are applied to each element in the stream. Uniform streaming, where one kernel function is applied to all elements in the stream, is typical. Kernel functions are usually pipelined, and local on-chip memory is reused to minimize external memory bandwidth. Since the kernel and stream abstractions expose data dependencies, compiler tools can fully automate and optimize on-chip management tasks. Stream processing hardware can use scoreboarding, for example, to launch DMAs at runtime, when dependencies become known. The elimination of manual DMA management reduces software complexity, and the elimination of hardware caches reduces the amount of die area not dedicated to computational units such as ALUs.
"

Continuing with the interview...

"
Now that Gamebryo is coming to the Wii, what sacrifices has the technology had to make?

I wouldn't say that we've made any sacrifices with our technology in taking it to Wii. We've always strove for maximum compatibility across platforms with Gamebryo, and our commitment to Wii is no exception.

With Gamebryo 2.3 on Wii we have almost all of the core features from Gamebryo 2.3 that the hardware will support. Obviously, there are some things that just aren't portable to Wii from the other platforms. Then again, I can't run a D3DX Effect on PS3, so I don't consider limitations of the platform a sacrifice.

Is Floodgate's technology applicable to the Wii without a multicore processor?

That's a tricky question to address, because it really depends on how you define "applicable." We have Floodgate running on Wii, and you are right that there are not as many opportunities for performance gains with the technology on Wii, since it's running on a single core.

However, we do our best to keep the overhead of using Floodgate minimal, so that companies can leverage code from their titles across all platforms without experiencing an overly high performance impact on platforms that don't have as many cores. Remember that one of the goals with Floodgate was to reduce the workload on programmers. With that fact in mind, I'd say that it's absolutely applicable.

Can you give a percentage of performance and features that the Wii can handle as compared to the 360 and PS3? Given your technical knowledge of the platform, are any games currently on the market pushing the tech as far as it can go? Can it be pushed further?

Coming up with a percentage is a perilous task, and it's one that I won't engage in these days. Any number that I give will be disputed the moment I say it. It's not just an apples-to-oranges comparison; it's like comparing apples to steaks. They're both food, but the comparison really ends there.

It's well understood that Nintendo chose a different strategy than Microsoft and Sony, and that the Wii is not as powerful from a raw numbers or features standpoint as PS3 and Xbox 360, so arguing over whether the GPU in Wii is 30% or 70% as powerful as the RSX or Xenos is a moot point.

As far as your second question goes, I'm pretty comfortable saying that no one is pushing the platform as far as it can go. Necessity is the mother of all invention, and developers will find ways to get more power out of all of these consoles. I think we saw that pretty clearly with the last round of titles in the previous generation. There really was no comparison between them and the launch titles.
"

I am pretty sure that most of you are wondering how is it that the Floodgate stream processing engine can make a difference in performance in a console that has only one core in it´s CPU and that the overall hardware power isnt that much compared to PS3 or Xbox 360. The answer is simple, the Wii´s CPU may not be very powerful and it´s single core, but the Hollywood has a lot stream processors that can be used for that job .

Graphics processing unit

http://en.wikipedia.org/wiki/Graphics_processing_unit

“

A graphics processing unit or GPU (also occasionally called visual processing unit or VPU) is a specialized processor that offloads 3D graphics rendering from the microprocessor. It is used in embedded systems, mobile phones, personal computers, workstations, and game consoles. Modern GPUs are very efficient at manipulating computer graphics, and their highly parallel structure makes them more effective than general-purpose CPUs for a range of complex algorithms

Stream Processing and General Purpose GPUs (GPGPU)

Main articles: GPGPU and Stream processing

A new concept is to use a modified form of a stream processor to allow a general purpose graphics processing unit. This concept turns the massive floating-point computational power of a modern graphics accelerator's shader pipeline into general-purpose computing power, as opposed to being hard wired solely to do graphical operations. In certain applications requiring massive vector operations, this can yield several orders of magnitude higher performance than a conventional CPU. The two largest discrete (see "Dedicated graphics cards" above) GPU designers, ATI and NVIDIA, are beginning to pursue this new market with an array of applications. Both nVidia and ATI have teamed with Stanford University to create a GPU-based client for the Folding@Home distributed computing project (for protein folding calculations). In certain circumstances the GPU calculates forty times faster than the conventional CPUs traditionally used in such applications.^[10]^[11]

Recently NVidia began releasing cards supporting an API extension to the C programming language CUDA ("Compute Unified Device Architecture"), which allows specified functions from a normal C program to run on the GPU's stream processors. This makes C programs capable of taking advantage of a GPU's ability to operate on large matrices in parallel, while still making use of the CPU where appropriate. CUDA is also the first API to allow CPU-based applications to access directly the resources of a GPU for more general purpose computing without the limitations of using a graphics API.

Since 2005 there has been interest in using the performance offered by GPUs for evolutionary computation in general and for accelerating the fitness evaluation in genetic programming in particular. There is a short introduction on pages 90–92 of A Field Guide To Genetic Programming. Most approaches compile linear or tree programs on the host PC and transfer the executable to the GPU to run. Typically the performance advantage is only obtained by running the single active program simultaneously on many example problems in parallel using the GPU's SIMD architecture^[12]^[13]. However, substantial acceleration can also be obtained by not compiling the programs but instead transferring them to the GPU and interpreting them there^[14]^[15]. Acceleration can then be obtained by either interpreting multiple programs simultaneously, simultaneously running multiple example problems, or combinations of both. A modern GPU (e.g. 8800 GTX or later) can readily simultaneously interpret hundreds of thousands of very small programs.

”

This may not come to a surprise to some of you but, do you remember that there were many rumors pointing that the Nintendo Revolution would be based on the ATI Radeon r520?

Did you know that all the ATI video cards that come from the x1000 family like the ATI Radeon r520 are all GPGPUs by hardware?

ATI stakes claims on physics, GPGPU ground
by Scott Wasson — 11:09 AM on October 11, 2005

http://techreport.com/discussions.x/8887

“One of the more surprising aspects of ATI's Radeon X1000 series launch is something we didn't get a chance to talk about in our initial review of the graphics cards: ATI's eagerness to talk about using its GPUs for non-graphics applications” …..etc

And if you pay good attention, we have recently seen titles like Monster Hunter 3 that support HDR+Antialaising, one of the feautures of the ATI x1000 Radeon video cards. We have also seen not just good graphics, but great physics in games like Dead Space Extraction, so, if the Wii Hollywood is just an overcloked Flipper, how is it that can achieve things like that?

ATI Radeon X1000 Series(HDR + anti-aliasing)

http://www.hardwaresecrets.com/article/227

Monster Hunter 3 demo impressions

By Nick Roussounelos 14th Jun 2009
Nintendo, 3,354 views

http://news.vgchartz.com/news.php?id=3996

“The demo consists of two quests, hunt either a Kurubekko (little wyvern much like Kut-Ku) or an alpha carnivore, Dosjagi. You can choose between 5 weapons: Great Sword, Sword & Shield, Hammer, Light Crossbow and Heavy Crossbow. Make your selections and off you go! The first thing you'll notice is the spectacular graphics; no, that's no hyperbole. The lighting, the texture detail, anti-aliasing (yes, you read that right, no jaggies), HDR (High dynamic range rendering for us geeks) and other technologies make this one of the prettiest Wii games on the market, period”

[Game Review] Dead Space Extraction (Wii)

By Zac Pritcher

Published: October 13, 2009Posted in: Featured, Games

http://www.everyview.com/2009/10/13/game-review-dead-space-extraction-wii/

“The developers have done a fantastic job of creating an enveloping environment with a vice-like grip that holds your attention the entire way through the game. The lighting effects are great, particles are nice, character and enemy models are all fantastic, and level design is creepy as hell. On top of that, the physics are excellent. Blasting off enemy limbs with pinpoint precision is a very satisfying feeling and works fantastically.”

Well, let´s continue with more about the Gamebryo and the Nvidia CUDA

CUDA and Gamebryo

http://www.nvidia.com/content/nvision20 ... mebryo.pdf

http://developer.nvidia.com/object/nvision08-gamebryo.html

http://www.gamedev.net/columns/events/g ... sp?id=1759

In the last links it is mentioned that CUDA is a bit more flexible than Gamebryo´s Floodgate and that this version doesn´t support scatter-gather option(we will see more about this in the definition of DMA)

What is Nvidia CUDA? http://www.nvidia.co.uk/object/cuda_what_is_uk.html

http://developer.download.nvidia.com/compute/cuda/1_0/NVIDIA_CUDA_Programming_Guide_1.0.pdf

" NVIDIA CUDA™ technology is the world’s only C language environment that enables programmers and developers to write software to solve complex computational problems in a fraction of the time by tapping into the many-core parallel processing power of GPUs. With millions of CUDA-capable GPUs already deployed, thousands of software programmers are already using the free CUDA software tools to accelerate applications—from video and audio encoding to oil and gas exploration, product design, medical imaging, and scientific research.
"

Nvidia's CUDA: The End of the CPU?

9:40 AM - 06/18/2008 by Fedy Abi-Chahla

http://www.tomshardware.com/reviews/nvidia-cuda-gpu,1954.html

“Conversely, the operation of a GPU is sublimely simple. The job consists of taking a group of polygons, on the one hand, and generating a group of pixels on the other. The polygons and pixels are independent of each other, and so can be processed by parallel units. That means that a GPU can afford to devote a large part of its die to calculating units which, unlike those of a CPU, will actually be used.

GPUs differ from CPUs in another way. Memory access in a GPU is extremely coherent – when a texel is read, a few cycles later the neighboring texel will be read, and when a pixel is written, a few cycles later a neighboring pixel will be written. By organizing memory intelligently, performance comes close to the theoretical bandwidth. That means that a GPU, unlike a CPU, doesn’t need an enormous cache, since its role is principally to accelerate texturing operations. A few kilobytes are all that’s needed to contain the few texels used in bilinear and trilinear filters.

”

Emergent's Gamebryo Game Engine to Incorporate NVIDIA PhysX Technology
http://www.iminers.com/render.php?eid=1 ... =pressroom

Y Nvidia tambien a confirmado que Wii usa el Nvidia Physx sdk
http://www.tomshardware.com/news/Wii-Nv ... ,7325.html

If you have understand a Little of what we have seen right now, then you might be wondering what is a GPGPU

GPGPU
http://en.wikipedia.org/wiki/GPGPU

"General-purpose computing on graphics processing units (GPGPU, also referred to as GPGP and to a lesser extent GP²) is the technique of using a GPU, which typically handles computation only for computer graphics, to perform computation in applications traditionally handled by the CPU. It is made possible by the addition of programmable stages and higher precision arithmetic to the rendering pipelines, which allows software developers to use stream processing on non-graphics data.

GPU functionality has, traditionally, been very limited. In fact, for many years the GPU was only used to accelerate certain parts of the graphics pipeline. Some improvements were needed before GPGPU became feasible
"

As you might have already guessed, the reason of why many Wii games haven´t demostrated to be very different from Gamecube games is due that almost nobody takes adventage of the stream processing features of the Hollywood.

Just to mention that GPGPUs adecuately programmed can do floating point calculations between 100 y 250 times faster than high end CPUs, but the software most be there in order to achieve that.

Another good reason of why most Wii games do not make proper use of the Hollywood´s GPGPU could be because of the high price that GPGPU programming comes according to declarations of Tim Sweeny of Epic Games :

Tim Sweeney: GPGPU Too Costly to Develop - Tom's Hardware

http://www.tomshardware.com/news/Sweeney-Epic-GPU-GPGPU,8461.html

Ok, now let´s take a closer look at the ability of Gamebryo 2.3 into using locked cache in Wii.

Beautiful pixels

Saturday, August 23, 2008

Multi-Platform Multi-Core Architecture Comparison (PC, Wii, Xbox 360, PS3, CUDA, Larrabee)

http://www.backfox.com///index.php/010110A/687474702s62656175746966756p706978656p732r626p6s6773706s742r636s6q2s323030382s30382s6q756p74692q706p6174666s726q2q6q756p74692q636s72652q6172636869746563747572652r68746q6p

“Getting into consoles, we start with the Wii. There are two types of memory, both accessible by CPU and GPU. However, what's really interesting is the ability to lock a portion of the L1 cache and explicitly manage it with DMA transfers. In one test case, we saw 2.5 times performance improvement by explicitly managing Floodgate transfers with the locked cache!”

DMA(Direct memory access)

http://en.wikipedia.org/wiki/Direct_memory_access

“Direct memory access (DMA) is a feature of modern computers and microprocessors that allows certain hardware subsystems within the computer to access system memory for reading and/or writing independently of the central processing unit. Many hardware systems use DMA including disk drive controllers, graphics cards, network cards and sound cards. DMA is also used for intra-chip data transfer in multi-core processors, especially in multiprocessor system-on-chips, where its processing element is equipped with a local memory (often called scratchpad memory) and DMA is used for transferring data between the local memory and the main memory. Computers that have DMA channels can transfer data to and from devices with much less CPU overhead than computers without a DMA channel. Similarly a processing element inside a multi-core processor can transfer data to and from its local memory without occupying its processor time and allowing computation and data transfer concurrency.

Without DMA, using programmed input/output (PIO) mode for communication with peripheral devices, or load/store instructions in the case of multicore chips, the CPU is typically fully occupied for the entire duration of the read or write operation, and is thus unavailable to perform other work. With DMA, the CPU would initiate the transfer, do other operations while the transfer is in progress, and receive an interrupt from the DMA controller once the operation has been done. This is especially useful in real-time computing applications where not stalling behind concurrent operations is critical. Another and related application area is various forms of stream processing where it is essential to have data processing and transfer in parallel, in order to achieve sufficient throughput.

DMA is an essential feature of all modern computers, as it allows devices to transfer data without subjecting the CPU to a heavy overhead. Otherwise, the CPU would have to copy each piece of data from the source to the destination, making itself unavailable for other tasks. This situation is aggravated because access to I/O devices over a peripheral bus is generally slower than normal system RAM. With DMA, the CPU gets freed from this overhead and can do useful tasks during data transfer (though the CPU bus would be partly blocked by DMA). In the same way, a DMA engine in an embedded processor allows its processing element to issue a data transfer and carries on its own task while the data transfer is being performed.

A DMA transfer copies a block of memory from one device to another. While the CPU initiates the transfer by issuing a DMA command, it does not execute it. For so-called "third party" DMA, as is normally used with the ISA bus, the transfer is performed by a DMA controller which is typically part of the motherboard chipset. More advanced bus designs such as PCI typically use bus mastering DMA, where the device takes control of the bus and performs the transfer itself. In an embedded processor or multiprocessor system-on-chip, it is a DMA engine connected to the on-chip bus that actually administers the transfer of the data, in coordination with the flow control mechanisms of the on-chip bus.

A typical usage of DMA is copying a block of memory from system RAM to or from a buffer on the device. Such an operation does not stall the processor, which as a result can be scheduled to perform other tasks. DMA is essential to high performance embedded systems. It is also essential in providing so-called zero-copy implementations of peripheral device drivers as well as functionalities such as network packet routing, audio playback and streaming video. Multicore embedded processors (in the form of multiprocessor system-on-chip) often use one or more DMA engines in combination with scratchpad memories for both increased efficiency and lower power consumption. In computer clusters for high-performance computing, DMA among multiple computing nodes is often used under the name of remote DMA.

IO Accelerator in Xeon

As an example of DMA engine incorporated in a general-purpose CPU, newer Intel Xeon chipsets include a DMA engine technology called I/O Acceleration Technology (I/OAT), meant to improve network performance on high-throughput network interfaces, in particular gigabit Ethernet and faster.^[3] However, various benchmarks with this approach by Intel's Linux kernel developer Andrew Grover indicate no more than 10% improvement in CPU utilization with receiving workloads, and no improvement when transmitting data.^[4]

AHB (possibly Wii Hollywood)

In systems-on-a-chip and embedded systems, typical system bus infrastructure is a complex on-chip bus such as AMBA High-performance Bus. AMBA defines two kinds of AHB components: master and slave. A slave interface is similar to programmed I/O through which the software (running on embedded CPU, e.g. ARM) can write/read I/O registers or (less commonly) local memory blocks inside the device. A master interface can be used by the device to perform DMA transactions to/from system memory without heavily loading the CPU.

Therefore high bandwidth devices such as network controllers that need to transfer huge amounts of data to/from system memory will have two interface adapters to the AHB bus: a master and a slave interface. This is because on-chip buses like AHB do not support tri-stating the bus or alternating the direction of any line on the bus. Like PCI, no central DMA controller is required since the DMA is bus-mastering, but an arbiter is required in case of multiple masters present on the system.

Internally, a multichannel DMA engine is usually present in the device to perform multiple concurrent scatter-gather operations as programmed by the software.

Cell

As an example usage of DMA in a multiprocessor-system-on-chip, IBM/Sony/Toshiba's Cell processor incorporates a DMA engine for each of its 9 processing elements including one power-processor element (PPE) and eight synergistic processor elements (SPEs). Since the SPE's load/store instructions can read/write only its own local memory, an SPE entirely depends on DMAs to transfer data to and from the main memory and local memories of other SPEs. Thus the DMA acts as a primary means of data transfer among cores inside this CPU (in contrast to cache-coherent CMP architectures such as Intel's coming general-purpose GPU, Larrabee).

DMA in Cell is fully cache coherent (note however local stores of SPEs operated upon by DMA do not act as globally coherent cache in the standard sense). In both read ("get") and write ("put"), a DMA command can transfer either a single block area of size up to 16KB, or a list of 2 to 2048 such blocks. The DMA command is issued by specifying a pair of a local address and a remote address: for example when a SPE program issues a put DMA command, it specifies an address of its own local memory as the source and a virtual memory address (pointing to either the main memory or the local memory of another SPE) as the target, together with a block size. According to a recent experiment, an effective peak performance of DMA in Cell (3 GHz, under uniform traffic) reaches 200GB per second

”

From what you can read in last link, it appears that Wii might be using the AHB option.

Types of DMA.- http://wiki.answers.com/Q/What_are_types_of_DMA

“Different Types of DMA

DMA is a mechanism through which a hardware device can transfer data to or from memory without using the processor. The processor is required to set up the transfer, and the device will signal the processor when it has completed the transfer. The advantage of this system is that the processor can perform other tasks while the DMA transfer is being performed.

There are several types of DMA used in Windows 2000 and later:

Common-buffer DMA

Common-buffer DMA is performed when the system can allocate a single buffer that is accessible by both the hardware and the software. The driver is responsible for synchronizing accesses to the buffer. The memory is not cached, making this synchronization easier for the driver. After setting up a common buffer, both the driver and the hardware can write directly to the addresses in the buffer without any intervention from the HAL.

Packet DMA

Packet DMA is performed when there is a single existing buffer that must be mapped for use by the hardware. An example of using packet DMA is the transfer of a file from memory to a disk. Using common-buffer DMA in this situation would be wasteful, because the file would have to be transferred to the common buffer before the hardware could transfer it to the disk. Instead, the HAL is consulted; it gives the driver the information it needs to help the hardware find the actual buffer in memory. This operation is complicated by the need for the routines involved to work across different architectures.

Scatter/gather DMA

Scatter/gather DMA is a shortcut method that sets up several packet DMA transfers at once. If you are transferring a packet over the network, for example, each part of the network stack adds its own header (TCP, IP, Ethernet, and so forth). These headers are all allocated from different places in memory. In this case, the scatter/gather DMA saves time by issuing a batch request to the HAL to map each header plus the data segment for access by the hardware. Instead of having to call the packet DMA routines on each part of the packet, this method calls each routine once, and lets the HAL be responsible for mapping each one individually.

Note Scatter/gather capability does not mean that the device can use the scatter/gather routines. Scatter/gather capability refers to a flag in the device description that indicates that the device is able to read or write from any area in memory, instead of just a certain range.

System DMA

System DMA is performed by programming the system DMA controller on the motherboard to do the transfer directly. Only ISA cards can use system DMA.”

“nVidia's CUDA is certainly an interesting architecture. It differs significantly from other systems, being a large collection of fairly small microprocessors. Each microprocessor block has a shared register file, and a large number of threads that are very efficiently switched by a hardware implemented scheduler. Each block also has a shared memory cache that must be explicitly managed by code. The left side of the diagram is the CPU of the system, I left it as a dual-core just for an example.”

Gamebryo Ligthspeed version 2.6

This engine is an update of the Gamebryo 2.5, and also includes the Nvidia Physx SDK

http://www.emergent.net/en/News/Press-R ... VELOPMENT/

Read with attention these parts that are mentioned in the previous link:

Gamebryo Leverages the Wii's Unique Architecture
"
Gamebryo for the Wii includes optimizations to the exporter pipeline, input mechanisms and rendering pipeline to ensure that you get the most out of this popular console. Gamebryo also supports the Wii Remote™ controller, and other standard Gamebryo features are mapped to the unique capabilities of this platform.
Cross-Platform Game Development With the Wii

Emergent’s Floodgate™ stream processing engine has been made Wii-aware for cross-platform game development projects: execution is now automatically optimized using the locked cache functionality, which reduces risk when porting multiprocessor code. Projects created with Gamebryo for other platforms can now be easily re-targeted to the Wii.

"

ATI and PhysX Co-exist on the Nintendo Wii
http://www.tomshardware.com/news/Wii-Nv ... ,7325.html

Read with careful attention the following which is mentioned in the article available at the previous link

"
PS3
Yesterday Tom's reported that Nvidia signed a deal with Sony Computer Entertainment Inc that gives PlayStation 3 developers access to the PhysX software development kit (SDK). According to the company, the kit is now available as a free download on the SCEI Developer Network and consists of a full-featured API and "robust" physics engine. But because the console's RSX GPU--based on Nvidia's G70 architecture (think GeForce 7800)--doesn't support PhysX in a hardware (or CUDA) sense, the middleware thus relies on the Cell's Synergistic Processing Units (SPUs) to process the physics rather than dumping the entire load on the Cell's Power Processor Unit (PPU).

Wii
what makes the whole announcement rather curious is just how the Nintendo Wii hardware can even handle physics processing. Of the three major consoles on the market today, the Nintendo Wii is the least powerful in a visual sense, relying more on the interaction provided by the Wii Remote.

Let's look at it this way: the Nintendo Wii relies on the PowerPC-based "Broadway" processor clocking in at 729 MHz and developed using 90 nm SOI CMOS processing. On the graphic side, the visuals are rendered by ATI's Hollywood" GPU, clocking in at 243 MHz and developed using a 90 nm CMOS process; there's a 3 MB embedded GPU texture memory and framebuffer thrown in there as well. As for the console's memory, there's 88 MB total: 64 MB "external" GDDR3 SDRAM and 24 MB "internal" 1T-SRAM integrated into the graphics package.

So how will the Nintendo Wii carry the burden? That question has yet to be answered, however after closer inspection of the Gamebryo LightSpeed announcement released last week (link), reporting that Emergent Game Technologies integrated PhysX into its Gamebryo 2.6 development platform for the Wii, today's announcement should not have come as a surprise. According to a Nvidia rep, PhysX has been a part of game development for some time; the company merely made it official with today's announcement. With the new SDK implimentation, Nvidia can now make changes directly to the middleware without the need for developer involvement.

"
We have to take into account that tom´s hardware is basing his question in specs that come from the first development kits that were realesed for Wii, which were denied as the final hardware of Wii in many occasions, like when Matt said that the specs came from 3^rd party development houses that had early development kits of the Wii, which were distributed by Nintendo even before the Hollywood chip was finished, or when NEC announced in June of 2006, Matt´s specs of Wii based in early development kits were announced in March 2006, that they would supply eDRAM for Wii(embedded in Hollywood) and that Mosys would be the designer of the macros based on the latest 1T-SRAM macros by that time . Even Nintendo denied the hardware present at E32006(similar to the first development kits as the final hardware.

Revolution's Horsepower

Studios give us the inside scoop on the clock rates for Broadway and Hollywood. How do the CPU and GPU stack up on paper?

by Matt Casamassina

March 29, 2006

http://wii.ign.com/articles/699/699118p1.html

Nintendo Confirms GameCube Housing for Wii at E3

May 19, 2006 | By: James Brightman

http://www.gamedaily.com/articles/features/nintendo-confirms-gamecube-housing-for-wii-at-e3/68959/?biz=1

NEC and MoSys Announce Wii Hardware Partnerships

Companies will work together to provide graphics LSIs and 1T-SRAM for the Wii.

by Gerry Block

June 19, 2006

http://gear.ign.com/articles/713/713254p1.html

Mosys speaks

http://www.thefreelibrary.com/MoSys'+1T-SRAM(R)+Embedded+Memory+Technology+Enables+Nintendo's+Next+...-a0147193666

“SUNNYVALE, Calif. -- MoSys, Inc. (Nasdaq:MOSY), the industry's leading provider of high-density system-on-chip (SoC) embedded memory solutions, today announced that Nintendo's upcoming Wii(TM) video game console uses MoSys' patented 1T-SRAM technology to implement the very high performance memory within the Wii console's graphics system.

This is the latest result from the more than six-years-long technical collaboration between the two companies. Previous generations of MoSys' 1T-SRAM technology were incorporated in Nintendo's earlier game console, the Nintendo GameCube(TM), in 1999, and shipped in multi-million quantities. The newest 1T-SRAM implementations embedded within the Wii console are fabricated using NEC Electronics' advanced 90nm CMOS-compatible embedded DRAM process technology. These high speed and ultra low latency memories are used as the main embedded memory on the graphics chip and in an additional external memory chip.

"We are pleased that Nintendo has again chosen our 1T-SRAM to power their latest generation game console," said Chet Silvestri, chief executive officer at MoSys, Inc.

"Designing the Wii console required an incredible list of breakthroughs in technology and innovation. The performance delivered by MoSys' 1T-SRAM technology is an important element of our solution," said Genyo Takeda, Senior Managing Director, General Manager, Integrated Research & Development Division at Nintendo. "The graphic performance of Wii benefits from MoSys' ability to develop highly innovative and dependable embedded memory products."

”

By that time, the newest macros available by Mosys were the 1T-SRAM-MIM(1T-SRAM-Q fabricated using MIM by TSMC instead of FAC), but since the manufacturer would be NEC and not TSMC, the macros were fabricated employing NEC´s MIM-2 technology.

As you can see, the Wii Hollywood was not yet finished even in June 2006, making it very suspicious because the memories UX6D (1T-SRAM-Q made with mim-2 and non-FAC) were already available for a long time, moreover the Wii Hollywood was finalized in dates approaching the dates when the ATI Radeon x1900 was released(November 2006), with this I do not say that Hollywood is an ATI Radeon x1900, just that as this was the unique from all other ATI Radeon X1000 family cards(the first cards of this family as the X1700 hit the market in late 2005) to include ATI FireStream processor technology, this must be a good reason of why Nintendo wanted to delay the release date of the Wii, so that way they could include the same technology used in the ATI Radeon x1900, or may even actually implemented their redesign known as AMD Stream Processor, as this was commercially available on 16 November 2006, and the Wii came out on 20 November 2006

http://en.wikipedia.org/wiki/AMD_FireStream

Jajaja, maybe this is why AMD priceses itself when it gives congratulations to Nintendos Wii Hollywood .

Apparently, Nintendo's Wii is silent and beautiful. Like a tiger. Meanwhile, AMD is prancing around like a school girl, taking all the credit by handing out backhanded compliments.

http://apcmag.com/amd_thanks_self_for_awesome_wii.htm

“"Nintendo's vision is to do something radically new and different with Wii," AMD says el presidente of Nintendo America said, continuing "... the technology hidden in the system allows us to deliver an intuitive and fun gaming experience in a silent and beautiful form factor," wrestling back some credit for the Wii's gameplay.”

And why is it that the declarations of AMD coincide so well to what ATI said earlier about the Wii Hollywood´s technology?

ATI: Wii Graphics at E3 "Tip of the Iceberg"

June 02, 2006 | By: James Brightman

http://www.gamedaily.com/articles/features/ati-wii-graphics-at-e3-tip-of-the-iceberg/69014/?biz=1

“ATI is also responsible for providing the custom GPU for Microsoft's Xbox 360, so we tried to find out how the "Hollywood" chip compares to what's in the 360. Once again, however, Swinimer sidestepped the question. "They're different chips for different platforms and different uses.”

Different chips for different uses?

But independently of the power, a GPU is a GPU, which means in other words that Swimmer said that Wii Hollywood is was a GPGPU.

Well, it was about time for an engine to take adventage of Wii´s architecture design.

Want to know more about Nividia Physx SDK?
http://en.wikipedia.org/wiki/PhysX

What is PhysX
" PhysX was originally the product of Ageia, a company based in California, which was founded in 2002. PhysX was originally designed to be a hardware physics solution. Specifically, it was sold on a PCI or PCI-E 1x style card, called a Physics Processing Unit, or PPU. It was designed to offload the physics calculations from the main Central Processing Unit, allowing the CPU to focus on other things, resulting in a higher framerate in video games. Also, since the PPU was dedicated to handling physics calculations, physics based effects became more accurate and often had more physics based objects flying about. The Ageia PPU was a large parallel processor, similar to a Graphics Processing Unit, or GPU."

En mi opinion, creo que este engine ha sido usado en el Dead Space Extraction dado a que este juego tiene muy buenas físicas segun varios reviwes disponibles en la red.

http://www.everyview.com/2009/10/13/game-review-dead-space-extraction-wii/

http://wii.nintendolife.com/news/2009/04/dead_space_extraction_screens_emerge_from_the_shadows

EA purchased the Nvidia Physx somewhere around 2008· Posted Dec 9, 2008 12:23 pm PT

http://www.gamespot.com/news/6202121.html

“Havok--the company Intel acquired--was the first to introduce physics into games and bring out a physics library. Havok's physics has been run on the CPU in a time-scheduled way, Peddie said. "Because of that, there weren't many CPU resources to really do a great job on the physics," he said. "Nothing would really happen. What happened, at most, is that you would hit this thing (a window or a wall, for example), and it would apply a decal to indicate that there was some change in it. It's not very realistic."”

Nintendo also purchased Nvidia physx for Wii

ATI and PhysX Co-exist on the Nintendo Wii

Next news

3:31 PM - March 19, 2009 by Kevin Parrish

http://www.tomshardware.com/news/Wii-Nvidia-PhysX,7325.html

As we all know, the Wii´s main CPU is so underpowered that handling physics through Nvidia Physc SDK would be a lot of burden, even high end CPUs like quad cores have troubles when you enable Nvidia physx through the CPU instead of the GPU; if a game like Batman or the Dead Space Extraction make use of a lot of physics, the CPU should be really powerful for calculating all complex algorithms required for the physics, not to mention the bandwidth consumption that would be required. And since Dead Space Extraction for Wii not only shows extraordinary physics but also high end graphics, which is why achieving the physics through the Broadway would not be an option..

Review de Dead Sapce Extraction

http://www.neoteo.com/dead-space-extraction.neo

dsHorror exclusive interview with Dead Space director Glen Schofield

http://www.deadspacehorror.com/dshorror-exclusive-interview-dead-space-director-glen-schofield

There was a question involving if the Wii had enpugh power to achieve the physics presented in the Dead Sapce game.

“dsH: Recently, Dead Space Extraction for Wii has been announced. Do you plan to integrate any new features? Will map sizes and number of enemies have to be reduced compared to Dead Space? Will it still be physics heavy?

Glen: You should see it. It looks fantastic on the Wii. In my opinion (and others) its one of the best looking games so far on the Wii. Obviously we will make a couple compromises but in other ways we can add more, especially taking advantage of the Wii controllers. There are new enemies, characters, story, puzzles, co-op and weapons. And yes, physics and even zero-g play a huge part in the game.”

Nvidia confirms that Wii has been using their previous versions of Nvidia Physx for some time

http://hothardware.com/News/DICE-Uses-NVIDIA-PhysX-in-Mirrors-Edge/

Some games that have made use of Nvidia Physx (including Wii titles)

http://developer.nvidia.com/object/physx_good_company.html

Wii floodgate locked cache DMA transfer

http://www.backfox.com/index.php/010110A/687474702s62656175746966756p706978656p732r626p6s6773706s742r636s6q2s323030382s30382s6q756p74692q706p6174666s726q2q6q756p74692q636s72652q6172636869746563747572652r68746q6p

NVIDIA PhysX y and other applications based on CUDA register more than one million downloads in two weeks

18 DE SEPTIEMBRE DE 2008

http://www.nvidia.co.in/object/io_1221741114287.html

Nvidia® PhysX™ offers offers more than 20 times realism in videogames
07 Dic, 2008 - 19:29:10

http://www.nvidia.com/object/io_1219046456259.html

NVIDIA wants ATI to implement PhysX

August 12, 2008

http://blogricardo.wordpress.com/2008/08/25/nvidia-quiere-que-ati-implemente-physx/

http://pc.ign.com/articles/898/898190p1.html

Nvidia will offer CUDA for Wii in the GDC 2009(what´s up ATI?)

http://www.ntra-net.com/2009/03/19/physx-on-the-wii-console/

http://www.theinquirer.es/2009/03/19/physx-en-la-consola-wii-de-nintendo.html

Emergent optimises Wii development

http://www.itexaminer.com/emergent-optimises-wii-development.aspx

El Nvidia Physx SDK runs in the cell of the ps3 and not in it´s gpu since the gpu version is g7(not gpgpu by hardware), and only starting from g8 physx can be running through CUDA

http://www.tomshardware.com/news/Wii-Nvidia-PhysX,7325.html

“Yesterday Tom's reported that Nvidia signed a deal with Sony Computer Entertainment Inc that gives PlayStation 3 developers access to the PhysX software development kit (SDK). According to the company, the kit is now available as a free download on the SCEI Developer Network and consists of a full-featured API and "robust" physics engine. But because the console's RSX GPU--based on Nvidia's G70 architecture (think GeForce 7800)--doesn't support PhysX in a hardware (or CUDA) sense, the middleware thus relies on the Cell's Synergistic Processing Units (SPUs) to process the physics rather than dumping the entire load on the Cell's Power Processor Unit (PPU).”

Nvidia Physx performance in the cpu and gpu comparasion(second page)

http://techgage.com/article/nvidias_physx_performance_and_status_report/

http://techgage.com/article/nvidias_physx_performance_and_status_report/2

http://www.elitebastards.com/index.php?option=com_content&task=view&id=599&Itemid=29&limitstart=3

All ATI Radeons that support gpgpu by hardware like those from the x1000 family, can run Nvidia Physx with modified drivers

http://www.tomshardware.com/news/nvidia-physx-ati,5764.html

Ok, why don´t we take a look into the Nintendo´s displacement mapping ptents and see how are they related with all previous information?

Read specially the parts that talk about reduction of the strain on the CPU and the inclusión of floating point units, as well as dot-product units in the rendering pipelines of the Hollywood and compre it with the definition of what a GPGPU is.

Nintendo Displacement mapping patents:

http://www.backfox.com///index.php/010110A/687474702s6r696r74656r646s2q7265766s6p7574696s6r2r626p6s6773706s742r636s6q2s323030355s31325s30315s617263686976652r68746q6p

“

2. Why could displacement mapping be Nintendo related?

Firstly, there is a Nintendo patent that has caused this topic to crop up in this community before. It is entitled Method and apparatus for efficient generation of texture coordinate displacements for implementing emboss-style bump mapping in a graphics rendering system.Its abstract is a bit of a mouthful, unfortunately. Read my highlights, though:

A graphics system including a custom graphics and audio processor produces exciting 2D and 3D graphics and surround sound. The system includes a graphics and audio processor including a 3D graphics pipeline and an audio digital signal processor. Emboss style effects are created using fully pipelined hardware including two distinct dot-product computation units that perform a scaled model view matrix multiply without requiring the Normal input vector and which also compute dot-products between the Binormal and Tangent vectors and a light direction vector in parallel.

The resulting texture coordinate displacements are provided to texture mapping hardware that performs a texture mapping operation providing texture combining in one pass. The disclosed pipelined arrangement efficiently provides interesting embossed style image effects such as raised and lowered patterns on surfaces.

This proves that Nintendo has not only been interested in this technique but is a patent holder. The section entitled ´cross-reference to related applications´ references 25 separate provisional patent applications that are thereby incorporated into the patent. Almost all of them date back to 2000. This would suggest that it is an important patent that has kept Nintendo busy but doesn´t date back too far to be cutting edge.

Secondly, relating back to making the process of displacement mapping more efficient and less of a strain on the CPU, one way of adaptive tessellation might actually be the last Nintendo patent I talked about in great detail, called Three-dimensional image generating apparatus, storage medium storing a three-dimensional image generating program, and three-dimensional image generating method.

A number of readers pointed out that the patent had nothing to do with actually visualising graphics in 3D, but rather optimizing a 3D world to be viewed on a 2D display. Then, that patent made little sense to me. But in the context of trying to reduce the computational strain on the CPU involved in displacement mapping, this may make perfect sense.

Lastly, whether the Revolution´s graphics chip will turn out to be based on the R520 or R530, it will be Radeon technology. And its manufacturer ATI has the following advice for developers on their Designing for Radeon development support page:

Use multi-texturing effects for realistic low polygon primitives. For example, you can use emboss style bump mapping to achieve the illusion of a bumpy surface that would take a lot more polygons to approximate otherwise. Similarly, other intelligent use of texture maps can reduce the polygon count of your mesh designs.

This may not be unusual, since nVidia will undoubtedly have similar advice on their development support pages, but at least it shows that ATI is also very concerned with this technique. In fact, ATI supported this technology earlier than nVidia, it seems. While the Radeon 9500/9700 was capable of displacement mapping, the GeForce FX was only partly so. The Radeon 9700 Pro already supported adaptive tessellation. In fact, ATI has an exclusive technology called ´Truform 2.0´, which is a kind of tessellation.

Now, there have been numerous rumours about Nintendo having discovered some kind of secret development technique. This may not be secret per se, but it would make sense if Nintendo had discovered a way of implementing displacement mapping efficiently. They have a patent relating to this technology and they have a strong ally who has some expertise in this field.

It may explain why Nintendo have not yet talked about the graphics chip or shown any real game footage yet. It would also explain why the basic hardware features that have been suggested seem to be underpowered at face value, yet Nintendo maintains that their graphics will be on par. This may yet turn out to be the Revolution´s last secret. "

Added.- The displacement mapping technique can achieve 1.4 millon polygons with just 1360 of them but……

"However, it comes at a price. While the poly count is significantly lower, there is some strain on the CPU. Johannes Hirche writes:

Rendering displacement mapped surfaces is a process that involves a significant number of geometric and arithmetic operations. When applied to a triangle mesh, it involves prior retessellation of the base domain surface and transformation of the vertices and normals. Even on fast CPUs, it is a time consuming operation, wasting bandwidth and processing power.

This is why displacement mapping has not been widely used in real-time graphics. However, new and refined techniques allow for displacement mapping to be implemented in real-time. Again, Johannes Hirche writes:

The main focus was to explore new techniques suitable for hardware implementation in order to reduce the bandwidth strain on the system bus by moving the tessellation process onto the graphics subsystem. (...) A possibility to overcome these problems is to tessellate the individual triangles sequentially and to adaptively add triangles where necessary, until a desired level of accuracy is

reached. (...) With only minor user interaction or conservatively predefined input parameters the sampling schemes produce adaptive tessellations with very low error measures.

ATI says that Gamecube was capable of Displacement mapping but not for realtime graphics

http://ati.amd.com/developer/gdc/GDC2003-DisplacementMappingNotes.pdf

Neverthless, since the Hollywood architecture is based in GPGPU, it can achieve the displacement mapping in a more efficient way, just like the ATI Radeons of the x1000 family.

http://developer.amd.com/media/gpu_assets/R2VB_programming.pdf

One of the reasons of why Wii has not been used well to produce amazing graphics is because the GPGPU programming was still to knew when Wii was launched, not to mention that programming for GPGPU requires more effort and can make the development too costly if there are no good tools to ease the development.

Tim Sweeney: GPGPU Too Costly to Develop

http://www.tomshardware.com/news/Sweeney-Epic-GPU-GPGPU,8461.html

Another reason is because most developers just saw and still see Wii as a platform for making money without much effort; the studios just take adventage of the of the popularity of the Wiimode gameply to lure people into buying games of low quality.

Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Re: Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Wii Hollywood is a GPGPU according to Dan Emerson of Gamebryo and Nvidia

Graphics processing unit

Stream Processing and General Purpose GPUs (GPGPU)

Monster Hunter 3 demo impressions

By Nick Roussounelos 14th Jun 2009Nintendo, 3,354 views

[Game Review] Dead Space Extraction (Wii)

Nvidia's CUDA: The End of the CPU?

Saturday, August 23, 2008

Multi-Platform Multi-Core Architecture Comparison (PC, Wii, Xbox 360, PS3, CUDA, Larrabee)

IO Accelerator in Xeon

AHB (possibly Wii Hollywood)

Cell

“Different Types of DMA

Revolution's Horsepower

Studios give us the inside scoop on the clock rates for Broadway and Hollywood. How do the CPU and GPU stack up on paper?

NEC and MoSys Announce Wii Hardware Partnerships

Companies will work together to provide graphics LSIs and 1T-SRAM for the Wii.

June 02, 2006 | By: James Brightman

ATI and PhysX Co-exist on the Nintendo Wii

Nvidia® PhysX™ offers offers more than 20 times realism in videogames07 Dic, 2008 - 19:29:10

http://www.nvidia.com/object/io_1219046456259.html

NVIDIA wants ATI to implement PhysX

By Nick Roussounelos 14th Jun 2009
Nintendo, 3,354 views

Nvidia® PhysX™ offers offers more than 20 times realism in videogames
07 Dic, 2008 - 19:29:10