OnLive demoed: lag, graphics are a problem. GASP!!!

[racofer]

BROS I FORESEE A TECHNOLOLOGIC DISCOVERY AFTER THIS ONLIVE CRAP WEAR PEEPLO ARE GIVEN A HOME DEVICE WHER THEY RENDER GRAPHICS IN REELTIME AND PROCESS STUFF LOCALLY FOR BETTER USER EXPERIENCE AN NO LAG ANYMORE AND U CAN BUY SOFTWARE NEXT TO UR HOUSE INSTEAD OF OVER THE INTERWEB ITS THE FUTURE BROS I TELL YE1!

 

[Xi]

And they somehow mysteriously ported UT3 and other games to sikritly developed OnLive OS nobody knows about which is also capable of running multiple instances of virtual machines or whatever on systems with custom built OnLive motherboards, CPUs, GPUs produced on supersikrit electronics factories nobody knows about that transfer data through hyperspace gates to user TV screen and all systems have supersikrit specs about which nobody knows too! Oh yes! Haha

Actually, I believe Onlive has a team that does porting for the big studios, however, independent development studios can get access to tools specifically for porting a game to the Onlive platform. The cluster does not run a bunch of simulated XP desktops. Was it designed with portability in mind? Probably, but again we can only speculate. There is no virtual machine, so to speak.

Individual users do not have dedicated hardware per user necessarily. This is also part of the need to port a game to Onlive's system. CPU usage, RAM, possibly even GPU are all virtualized across the cluster. It's not that no one knows about these things, it's just very high level technology that only so many people in the country even have the skill to understand. Some joe blow, like yourself, will not be able to understand how it works because you have a meager brain that only understands desktop technology. However, I'm sure you could read some books on the hardware of cloud computing. They probably are available: Here is a link to one with a few recommendations also toward the middle of the page.

these have GPU's, lots of them, that handle the rendering of the graphics.

Seriously only a complete moron that hears about PC for the first time will buy into this bullshit

Like I said in the Onlive thread that I started a few months/weeks ago, no one ever dreamed that we could fly, or understood the concept of a sub-atomic particle, or thought we would leave the earth's orbit, or land a rover onto mars that could send back video footage. It's all a matter of science and the technology that utilizes what we know about science. Just because something seems impossible does not make it so.

However, maybe you should be telling me what you see in a few Rorschach Inkblots instead?

 

[Melcar]

Es un misterio.

 

[MapMan]

Porting UE3 is pretty much impossible atm and the only company that can do that is Epic. They wanted to release UE3 linux build but due to 3rd party software licenses they couldnt do it. Unless they drop the 3rd party software(They havnt told us what exactly is it) and rewrite it themselves I dont see a port happening anytime soon.

 

[Overweight Manatee]

It's not that no one knows about these things, it's just very high level technology that only so many people in the country even have the skill to understand. Some joe blow, like yourself, will not be able to understand how it works because you have a meager brain that only understands desktop technology.

Yeah, I'm sure that I am one of only the handful of people who understand technology we have had in almost the exact same form for at least 10 years. Almost certainly longer then that but I'm just being conservative.

Virtualizing something doesn't mean that you magically cut the hardware needed in half or something. In fact it means that your system requirements go up (only slightly if the implementation is good, but it does go up). The fact is they still need X CPU time and Y GPU time to render a game. Virtualizing doesn't change that, it just lets you be a bit more flexible to adjust to the demands of different games.

 

[bozia2012]

However, maybe you should be telling me what you see in a few Rorschach Inkblots instead?

I see tachyons.

 

[ChristofferC]

I don't see how you could virtualize games over several computers in a general way since you must use a dedicated video card for the graphics calculations. You can't just split up those calculations over several video cards in a general way for the same reasons that you can't just check a box to make your application multithreaded. There is also not much benefit in doing so (in fact you would lose some performance). All I can see them doing is running a few instances of less resource intensive games on the same computer as single current title. Sounds a bit less cool than what they're trying to make it sound, huh?

 

[Melcar]

It's because of the midichlorians guys.

 

[Overweight Manatee]

All I can see them doing is running a few instances of less resource intensive games on the same computer as single current title. Sounds a bit less cool than what they're trying to make it sound, huh?

Say its not so. I heard Onlive got their technology straight the UFO held in Area 51.

 

[MetalCraze]

No man, you are just like those stupid guys who said that man won't fly!
Well man doesn't fly still but... eh

 

[rpgcodexusername]

Strangely OpenGL was created for this from the ground-up
http://en.wikipedia.org/wiki/GLX

 

[Xi]

http://www.techcrunch.com/2008/07/09/ot ... echnology/

Onlive isn't the only company pursuing this type of computing. Also, both Nvidia and ATI both have designed video cards specifically for this type of computing. Even if the initial services have a rough start, technology will improve to make such things a reality.

http://www.techcrunch.com/2009/06/16/vi ... -see-this/

Gaikai
OTOY
Onlive

http://www.joystiq.com/2009/07/01/dave- ... of-gaikai/

In the end, the only really good argument centers around how well an FPS or twitch style action game might work. All other types of games are no problem as input lag has less of an effect. A system that promotes more than just action? Sold.

It's not just limited to games either, any program could effectively run on these type of systems. Theres a video that shows Photo Shop via Gaikai, including WoW, EVE, Spore, etc, in action.

 

[Overweight Manatee]

In the end, the only really good argument centers around how well an FPS or twitch style action game might work. All other types of games are no problem as input lag has less of an effect. A system that promotes more than just action? Sold.

Yes, the sad thing is that onlive could conceivably revive interest in turn based strategy games and adventure games. But pretty much EVERY other genre is fairly twitchy. Maybe not enough that you would notice lag if you were lazily laying on a couch 20 feet away from the TV with a control in one hand while scarfing down chips in the other, but for PC gaming I just don't see it being relevant.

 

[MetalCraze]

Onlive isn't the only company pursuing this type of computing. Also, both Nvidia and ATI both have designed video cards specifically for this type of computing.

You do know that Tesla is not for gaming purposes right?

So tell me Xi, what is this mysterious technology OnLive utilizes? Tell me. Tell me specs. If it's real where are specs? What OS do they use (you can only play UT3 on Windows)? What videocard (Tesla isn't supported by any game to render everything fully)? What processor (otherwise it's still 1 cpu per videocard) and who manufactures all of it for them? Who designed it for them? Who hacked the closed-source games for them to make it work on something else than 1 CPU-1/SLI/CF GPU with Windows so they work on OnHoax?

There is currently no technology from Intel, AMD, ATI, NVidia or even fucking VIA-S3 that is capable of what OnLive promises (and isn't planned I wonder why) , so who manufactured their mysterious servers for them that are capable of rendering multiple instances of UT3 at once and hyperspace data transfer?

 

[Xi]

http://www.nvidia.com/object/io_1256065754359.html

It does exist. Here is an article from the source. It utilizes the Tesla GPUs even, which I posted earlier.

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

Here's info on how it all works, or at least all Nvidia will say.

The Tesla RS system is configurable from a small 8 Tesla GPU system designed for small workgroup collaboration environments to large 100 Tesla GPU systems that are ideal for serving many consumers over the internet. Enabling anyone to interact remotely with complex, photorealistic 3D models and environments, Tesla RS-based server clusters deliver more than 10 times faster performance than CPU-based clusters while consuming less power and needing less space. Featuring the revolutionary NVIDIA® CUDA™ parallel computing architecture and powered by 240 parallel processing cores in each Tesla processor, the NVIDIA RealityServer with iray rendering technology delivers streaming realism to web-connected devices. Users will be able to interact with photorealistic 3D scenes and customize lighting conditions, using web-connected PCs, netbooks, and smart phones.

Consumer Service Platform
• 100+ Tesla GPUS
• 1000+ concurrent users

The RealityServer platform runs in GPU-based cloud computing environment, accessible using web-connected PCs, netbooks, and smart phones, enabling 3D web applications to dynamically scale based on utilization requirements.

If you don't want to believe Nvidia, then I can't help you. AMD has a similar offering that the Gaikai doods were using, though I think Nvidia is slightly ahead on the Cloud Computing front.

It's really coming... What else can I say?


AMD Offering:

Cloud Computing Technology from AMD



Cloud service providers interested in leveraging this powerful IT delivery methodology must select underlying technologies that offer a high level of performance, reliability, power efficiency, and cost-effectiveness.

An essential component of any cloud computing environment is servers equipped with processors that deliver a balance of performance, I/O capabilities, low energy consumption, and hardware-assisted virtualization. AMD Multi-Core Opteron™ processors offer these qualities, along with the value needed to drive a cost effective cloud environment.

Balanced Architecture: The Multi-Core AMD Opteron™ processor is designed to deliver performance that balances power consumption, memory management, I/O, and throughput. This enables the performance and power efficiency needed to help to drive a cost effective cloud computing environment.


Low Energy Consumption: The Multi-Core AMD Opteron processor contains AMD-P technology, a suite of features that can help to significantly reduce energy usage. This technology can help cloud service providers manage their vast virtual server farms while helping to minimize downtime.

All Multi-Core AMD Opteron processors feature AMD PowerNow!™ technology, which dynamically manages power utilization across processor cores, and AMD CoolCore™ technology, which enables reduced power consumption within individual cores.
They also feature Dual Dynamic Power Management™, which can enable reduced idle processor power consumption while enabling per-processor power management in multi-socket systems.
Together, those technologies and others can help enable cutting-edge energy efficiency. That can be a major competitive advantage for cloud service providers, whose data centers typically house large volumes of power-hungry servers.


Hardware-assisted Virtualization Technology: Multi-Core AMD Opteron processors also feature AMD-Virtualization (AMD-V™) technology, hardware extensions designed to improve efficiency and reduce overhead of virtualization software. AMD-V™ enables:

Virtualization software to operate with optimal efficiency and enables a quick transition between virtual machines.
Features Rapid Virtualization Indexing (RVI) technology, which offers fast and efficient performance by executing functions typically handled in software within the CPU.
Fast switching between virtual machines, and its functionality enables running virtual machines to be migrated across AMD Opteron processor generations.
More resources:

"Managing the Cloud"
Archive of Cloud Computing articles from IDG
ACP - The Truth About Power Consumption Starts Here

http://sites.amd.com/us/cloud/Pages/about.aspx

 

[Xi]

Again, it's all pretty new. I don't think there's a lot of info out about how it all functions without purchasing the hardware and working directly with Nvidia/AMD engineers.

Just gonna have to wait and see, but it's definitely a reality. Hell, even if Onlive fails, someone will figure out how to make it work. It's a LOT of serious $$ if you looked at what I posted earlier.

 

[Xi]

Need a job? Go work for Onlive!

http://san-francisco-jobs.jobfox.com/te ... plyhired75

Job Description:
Game Development Engineer, OnLive Platform
Location: Palo Alto, CA, United States

Job Type: Full-Time
We are looking for an exceptionally agile and flexible engineer that has experience porting large games to new platforms. We are interested in candidates that have worked on AAA console and PC games, are familiar with leading game engines, and are great communicators. You will work with leading game developers worldwide to integrate the OnLive SDK into existing games and teach outside development studios about the OnLive SDK.

RESPONSIBILITIES:
Drive integration of SDK for OnLive game features. Work closely with leading game developers to develop innovative next-generation video games
Understand game features and game development process
Triage incoming support issues via forums, email, and phone calls
Write whitepapers and code samples that illustrate common problems encountered by customers
Develop and present SDK training sessions as required
Pro-actively learn and implement new technologies quickly and independently

REQUIRED COMPETENCIES:
B.S degree with 3 or more years experience in software engineering or equivalent experience
Excellent software architecture skills
Demonstrated consultative customer management skills, including expectation setting, collaborative problem resolution, teaching of services/capabilities
Must have game console programming and game middleware experience
Independent, can identify issues and opportunities and act on them while keeping the team informed
Positive attitude and a strong work ethic, passion for cutting edge real-time tech, passion for video games
Ability to maintain balance under pressure and hit tight deadlines
Excellent written & verbal communication, documentation skill
Previous experience writing game network code using TCP and UDP
Experience using defect tracking and revision control tools. Perforce experience a plus
Strong Windows SDK experience
Preferred Languages - C++, C, C#, STL, Python, SQL, PHP, Perl, LUA
Preferred Technologies – Xbox360 SDK, PS3 SDK, Unreal game engine, MSVC

NICE TO HAVE:
Experience in field engineering organization providing onsite customer support
Experience with web technologies - Soap/Web Services, Tomcat, IIS, and Apache, HTML, XML, HTTP, HTTPS

 

[Xi]

What is GPU Computing?

GPU computing is the use of a GPU (graphics processing unit) to do general purpose scientific and engineering computing.
The model for GPU computing is to use a CPU and GPU together in a heterogeneous computing model. The sequential part of the application runs on the CPU and the computationally-intensive part runs on the GPU. From the user’s perspective, the application just runs faster because it is using the high-performance of the GPU to boost performance.


The application developer has to modify their application to take the compute-intensive kernels and map them to the GPU. The rest of the application remains on the CPU. Mapping a function to the GPU involves rewriting the function to expose the parallelism in the function and adding “C” keywords to move data to and from the GPU.

GPU computing is enabled by the massively parallel architecture of NVIDIA’s GPUs called the CUDA architecture. The CUDA architecture consists of 100s of processor cores that operate together to crunch through the data set in the application.

The Tesla 10-series GPU is the second generation CUDA architecture with features optimized for scientific applications such as IEEE standard double precision floating point hardware support, local data caches in the form of shared memory dispersed throughout the GPU, coalesced memory accesses and so on.

"GPUs have evolved to the point where many real-world applications are easily implemented on them and run significantly faster than on multi-core systems. Future computing architectures will be hybrid systems with parallel-core GPUs working in tandem with multi-core CPUs."

Prof. Jack Dongarra
Director of the Innovative Computing Laboratory
The University of Tennessee

History of GPU Computing

Graphics chips started as fixed function graphics pipelines. Over the years, these graphics chips became increasingly programmable, which led NVIDIA to introduce the first GPU or Graphics Processing Unit. In the 1999-2000 timeframe, computer scientists in particular, along with researchers in fields such as medical imaging and electromagnetics started using GPUs for running general purpose computational applications. They found the excellent floating point performance in GPUs led to a huge performance boost for a range of scientific applications. This was the advent of the movement called GPGPU or General Purpose computing on GPUs.

The problem was that GPGPU required using graphics programming languages like OpenGL and Cg to program the GPU. Developers had to make their scientific applications look like graphics applications and map them into problems that drew triangles and polygons. This limited the accessibility of tremendous performance of GPUs for science.

NVIDIA realized the potential to bring this performance to the larger scientific community and decided to invest in modifying the GPU to make it fully programmable for scientific applications and added support for high-level languages like C and C++. This led to the CUDA architecture for the GPU.

CUDA Parallel Architecture and Programming Model

The CUDA parallel hardware architecture is accompanied by the CUDA parallel programming model that provides a set of abstractions that enable expressing fine-grained and coarse-grain data and task parallelism. The programmer can choose to express the parallelism in high-level languages such as C, C++, Fortran or driver APIs such as OpenCL™ and DirectX™-11 Compute.



The first language support NVIDIA provided is for the C language. A set of C for CUDA software development tools enable the GPU to be programmed using C with a minimal set of keywords or extensions. Support for Fortran, OpenCL, et cetera will follow soon.

The CUDA parallel programming model guides programmers to partition the problem into coarse sub-problems that can be solved independently in parallel. Fine grain parallelism in the sub-problems is then expressed such that each sub-problem can be solved cooperatively in parallel.

The CUDA GPU architecture and the corresponding CUDA parallel computing model are now widely deployed with 100s of applications and nearly a 1000 published research papers. CUDA Zone lists many of these applications and papers.

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

 

[ChristofferC]

Why do you keep posting links to stuff you don't understand? CUDA is not for games so it has no relevance in this thread.

 

[Xi]

Why do you keep posting links to stuff you don't understand. CUDA is not for games so it has no relevance in this thread.

They are designed for computing in a massive cluster, not for general consumer rendering. They can virtualize rendering across a cluster, in the same way you virtual CPU usage. A single chip alone is nothing to brag about, but when you have a thousand of them in a server farm you can host 100s of thousands of games and load the cpu/gpu/ram requirements across the cluster.

They are specifically for Cloud computing. Onlive will probably use this system as they were at an Nvidia conference show casing Tesla last year.

CUDA is a part of the RealityServer tech, not the entire thing. It's one component.

RealityServer® is the 3D web services software platform that is highly scalable to thousands of servers. It is ideal for creating and deploying 3D web services for interactive and collaborative 3D applications.

http://www.mentalimages.com/index.php

RealityServer is Nvidia's attempt to bring 3D to any computing platform that supports a browser or standard Web services calls, including netbooks and smartphones.

SLIDESHOW (7)

Slideshow|All Shots
By moving the CPU-crushing rendering requirements of creating high-resolution images and animations off of the client and onto a back-end computer, Nvidia hopes to bring complex graphics applications like fluid dynamics, architectural design, real-time product styling and design, 3D video games, to computing platforms that don't have the processing power to run them locally. RealityServer could mean the transformation of the Web and its applications into a 3D world complete with photorealistic ray-traced images and high-resolution animations that can be scrolled, rotated, and painted in real time.
To find out more about Nvidia's plans to 3D-enable the Web, I interviewed key people from Nvidia and Mental Images, the company Nvidia recently acquired to provide the rendering software behind RealityServer. But first, we'll take a look at the hardware and software technology underlying RealityServer and the tools Nvidia will provide to application developers. RealityServer will be available starting November 30, 2009.

http://www.extremetech.com/article2/0,2 ... 677,00.asp

"As we discussed last month, RealityServer 3.0 is Nvidia's attempt to bring photo-realistic 3D images to any Internet-connected device, including the likes of Android and iPhone. RealityServer 3.0 pushes the CPU-killing 3D rendering process to a high-power, GPU based, back-end server farm based on Nvidia's Tesla or Quadro architectures. The resulting images are then streamed back to the client device in seconds; such images would normally take hours to compute even on a high-end unassisted workstation. Extreme Tech has up an article containing an interview with product managers from Nvidia and Mental Images, whose iray application is employed in a two-minute video demonstration of near-real-time ray-traced rendering."

http://tech.slashdot.org/story/09/11/15 ... +(Slashdot)&utm_content=Google+Feedfetcher

 

[Overweight Manatee]

http://kotaku.com/5490382/onlive-starts-streaming-games-on-demand-this-june-for-15-a-month


What a great deal! Onlive is selling at $15 a month for the purpose of locking you into DRM, making your games useless if OnLive fails or if you can't connect to the internet, providing a laggy, low settings game with compression artifacts, and it doesn't even include games, you still have to 'buy' them yourselves.

 

[Grunker]

Keep in mind that $14.95 USD monthly fee does not include the rental and ownership fees associated with actually playing those games. The subscription fee pays for things like "instant-play free game demos; multiplayer across PC, Mac and TV platforms; massive spectating; viewing of Brag Clips video capture and posting; and cloud-saving of games you've purchased." Yeah, don't toss your PC in the nearest Dumpster just yet.

Lol. If this is succesful, I'm boycotting everything but work-related computer-use for some time, just to get away from the stupid.

 

[MetalCraze]

http://kotaku.com/5490382/onlive-starts-streaming-games-on-demand-this-june-for-15-a-month


What a great deal! Onlive is selling at $15 a month for the purpose of locking you into DRM, making your games useless if OnLive fails or if you can't connect to the internet, providing a laggy, low settings game with compression artifacts, and it doesn't even include games, you still have to 'buy' them yourselves.

FINALLY THIS IS A TECHNOLOGICAL BREAKTHROUGH $15 A MOTNH IS BETTAR THAN UPGRADING PC FOR $2000 EVERY MONTH!!!!

 

[Xi]

$15 a month is dirt cheap. Sign me up. I'll activate the service for 1 month, play the latest game I'm interested in, then cancel the service after a month. I'll do this rather than investing 500+ dollars every time I want to play a certain game.

 

[Overweight Manatee]

$15 a month is dirt cheap. Sign me up. I'll activate the service for 1 month, play the latest game I'm interested in, then cancel the service after a month. I'll do this rather than investing 500+ dollars every time I want to play a certain game.

Absolutely, because everyone knows you only play 1 game for 1 month once every 5 years.