Slashdot Mirror
Ouya Performance Not Particularly Exciting
hypnosec writes "Results of recent benchmark tests reveal that Ouya is not up to the mark and there are over 70 other ARM devices that perform better than the gaming console. Futuremark, which is known for its benchmarks like 3DMark and PCMark, benchmarked mobile devices and the Tegra 3 powered Ouya has been ranked 73rd."
Of course, most of the those devices cost a lot more than $100 without carrier subsidies.
As the early Nintendo days can attest.
I think it is far more interesting that it scored higher than the majority of other tegra 3 devices which cost far more. I never really expected it to be performance impressive by the time it shipped. It is running on a 1 year old chip.
Of course it is going to be outpaced by the newer devices.
So? This thing was never meant to be a PS4. The OUYA has my attention for several reasons: 1.) It's a kickstarter project and I hope it's successful for the sake of those that bet so much on it. 2.) It's cheap - consoles are never this inexpensive. The Wii was cheap, but the controllers were ungodly expensive (granted, the OUYA controllers aren't that cheap either). 3.) It's open. This is perhaps most important. I had more fun hacking a Wii and turning into an emulator box and a media streamer than I've ever had with my old, dusty Xbox 360. If I can do that with the blessing of the company who's box I just purchased, hell yes I'll buy one.
A lot of the the new, popular indie games available aren't exactly taxing on system requirements. Granted some of them could stand a bit of optimization, but having a common framework and a fixed hardware target (exacly what the Ouya provides) really will help there.
I've got a nice overclock sandy bridge i5 and a high end video card in my gaming system. While I enjoy many of the newer A-list titles with all of their eye candy, I probably put a lot more gaming hours in to titles like minecraft (mostly mod packs like tekkit or FTB), binding of issac, don't starve, super meat boy, and a lot of others that can be had for a couple of bucks on steam.
While not the fastest thing in the world, I still think the ouya could put a lot of very good games in to the hands of eager players for a very good price. The big console makers miss the mark on indie titles, requiring way too much money for development and focusing way too heavily on monitization at the expense of gameplay.
This. Sadly, I personally don't think that Ouya content is going to be able to carry it though.
Except right now even before launch it has potentially more games than xbox360, ps3; and wii combined...and cheap too, most under a dollar. Everything from throwaway games to 20hr RPG's, Lets be honest most modern game engines work on Android. In fact the only problem it has is making out the quality from the...not so quality
The raw part cost of a smartphone SoC is a tiny portion of the bill of materials (BOM), maybe 10-15%. CPU is maybe $30 at the very high end? So for a box like the OUYA where the CPU is probably the biggest cost and they don't have to worry about a display, camera, battery, cellular radios, or massive amounts of storage, they probably could have sprung for a Snapdragon 600 or Tegra 4. Only thing is it would have delayed the product by 6 months since those chips are in high demand from smartphone OEMs. Take a look at this cost breakdown analysis of the GS4: http://www.isuppli.com/Teardowns/News/Pages/Samsung-Galaxy-S4-Carries-236-Bill-of-Materials-IHS-iSuppli-Virtual-Teardown-Reveals.aspx $236 worth of parts selling for $699 just shows you how things are roughly priced (granted, MSRP - BOM != profit, but Samsung is in a pretty good position). Also you'll learn the biggest conspiracy of smartphones ever: it does NOT cost $100 to go from 16GB NAND to 32GB, or 32->64, or 64->128.
If it can power games anything like A Link to the Past and Symphony of the Night at 1080p then it'll do just fine. The only thing that worries me is the possibility of a metric ton of bad games combined with a lack of great ones like my examples. We'll all find out soon enough.
Buy your next Linux PC at eightvirtues.com
Ouya competes with non-hyped Android sticks. I just got $50 MK808B and Neo G4 ($75) and X5 ($100, but more I/O) sticks that I'm setting up to do Skype, Internet, email... and games... for friends and family. Dual-core A9, 1GB RAM, 8GB Flash, 2x or 3xUSB, BT, Wifi, Android 4.1 (4.2 on the way), SD slot, HDMI (and SPDIF for the X5)... and the full Android PlayStore,which Ouya and GameStick don't offer.
Add a $50 gamepad (a really good one, xbox or DualShock), any old keyboard and mouse, or a Logitech K400 if you want to get fancy, and you get something that can play almost as well as the Ouya/Gamestick, and do a whole lot more thanks to the PlayStore.
There aren't a whole lot of games that support gamepads or kb+ms, and quite a few games won't run at all because of lack of touch/accelerometer/gyroscope, and the portrait mode... but there are still quite a few good games, a whole bunch of emulators... and this is a lot more than what the OuyaSticks have right now. And there's a good chance that OuyaStick games will find their way to the PlayStore, too, devs would be crazy not to port them: very little extra work, a way bigger market.
I think it all comes down to the games: if Ouya or GameStick not only catch up to the PlayStore but snag good, exclusive games, it might be worth pay as much for them as for a true Android device, in spite of Ouya/GameStick being as expensive, more limited, and having bad controllers.
And quad-cores are on the way for less than $100.
The Cloud - because you don't care if your apps and data are up in the air.
The OUYA is a self-contained computer. It is only missing a display.
You also have to consider that an OUYA with a controller is $100, and that a controller by itself is $50. So this is basically a $50 self-contained computer. I expect the performance to match/or exceed that of other $50 self-contained computers.
http://blindscribblings.com - Tasty pop-culture in conceptual fashion.
The benchmark results show the OUYA (basically a $50 console bundled with a $50 controller) was faster than the HTC One S, which sells for $450 outside of a contract.
http://blindscribblings.com - Tasty pop-culture in conceptual fashion.
The raw part cost of a smartphone SoC is a tiny portion of the bill of materials (BOM), maybe 10-15%. CPU is maybe $30 at the very high end? So for a box like the OUYA where the CPU is probably the biggest cost and they don't have to worry about a display, camera, battery, cellular radios, or massive amounts of storage, they probably could have sprung for a Snapdragon 600 or Tegra 4. Only thing is it would have delayed the product by 6 months since those chips are in high demand from smartphone OEMs.
Take a look at this cost breakdown analysis of the GS4: http://www.isuppli.com/Teardowns/News/Pages/Samsung-Galaxy-S4-Carries-236-Bill-of-Materials-IHS-iSuppli-Virtual-Teardown-Reveals.aspx $236 worth of parts selling for $699 just shows you how things are roughly priced (granted, MSRP - BOM != profit, but Samsung is in a pretty good position). Also you'll learn the biggest conspiracy of smartphones ever: it does NOT cost $100 to go from 16GB NAND to 32GB, or 32->64, or 64->128.
Yes, but what you're forgetting about is the opportunity cost. If you can make a $100 device that competes equally with a $300 device simply by lowering your profit margin, that's not the whole picture. To design and start producing that device, you need funding. That capital typically comes from people who want a return on their investment, and lower risk. The smaller your profit margin, the closer you are to not being profitable at all if you miscalculated, incorrectly estimated, or failed to account for something. And even if you hit exactly your intended margin, you still end up providing a lower return on investment than if you had charged more...or, in this case, if you lower the hardware costs. Also, keep in mind that a $5 hardware cost difference matters less, profit-wise, on a $300 device than it does on a $100 device.
Hardware isn't designed and built in a vacuum; these things happen in the context of a business, as well as in the context of an entire industry.
For your security, this post has been encrypted with ROT-13, twice.
As AMD began its project, many years ago, to fuse first-class GPU circuits within the same chip as the CPU, Nvidia was forced to respond. Nvidia contemplated building an x86 processor of its own, but quickly dropped that idea to focus on building ARM SoC parts. Nvidia had but one goal- to be the number one high-end supplier of ARM solutions.
Now, many years later, we can see just how badly Nvidia has failed. Tegra 1 was a disaster. Tegra 2 and 3 were terribly late, and only gained sales when Nvidia was forced to essentially give away the chips. Tegra 4 is even later than any previous ARM part, and is such a badly conceived device, Nvidia has been forced to nigh on cancel it in order to 'rush' to release the crippled Tegra 4i that will have a better price and power consumption at the cost of CPU and GPU performance.
Tegra 3 stinks because originally the much faster Tegra 4 was supposed to be in devices by now. Tegra 4 is the last of the ULP GPU Tegra designs from Nvidia. Tegra 5 brings desktop GPU designs to ARM- at which point Nvidia loses interest in the obsolete slow ULP GPU.
Ouya is obsolete. It is only good for running a class of games associated with weak Android hardware. The cutting edge Android games will be ports from high end iPads, and will be a bad match for the Tegra 3. There are increasing numbers of Android boxes that you can also plug into your TV and controllers. What possible point does Ouya have in this light?
If Nvidia has the money and the people, it only makes sense for Nvidia to be accelerating production of Tegra 5 at this stage. Current ARM/Android is matching the early days of the decent PC (486 -> Pentium -> Pentium Pro/Pentium 2) which was also (eventually) accompanied with the birth of the PC 3D graphics accelerator. Nvidia is currently running dead last in the ARM GPU stakes, behind Mali, Adreno, and (of course) PowerVR. Given that Adreno is an old ATI design, this is incredibly humiliating for Nvidia. In fact, Google is throwing out Nvidia with its tablet refresh later this year, and going Adreno (via Qualcomm).
It gets worse. AMD's astonishing Temash APU, with 4 Jaguar x86 cores and a brilliant GPU, will be a vastly better match for a little independent console box. While a console based on Temash would likely be 200 dollars, and run Windows rather than Android games, its improved performance (3X+ GPU, 20X+ GPU) would make for vastly better value and longevity.
Price is not a valid comparison.
The most expensive parts of a phone are not part of the Ouya. No screen, no battery, no cell radio.
The Ouya console has chosen a lack-luster SoC. If they chose a PowerVR554, an Adreno225/320 or even a Mati-T604 SoC I doubt it would have been more than a few dollars difference.
Now, many years later, we can see just how badly Nvidia has failed
I find ARM incredibly confusing. I understood PC CPU/GPU, and I am on the whole pretty knowledgeable. As informed as your post maybe. In my mind I can only name two really recognisable ARM brands!? Snapdragon[because it was everywhere] and Tegra...and I only associate one with graphics performance. I have to say brand goes a long way.
but if you think single platform games, on a vapourware machine costing twice as much as this console is somehow a threat to Nvidia Arm, its not. It might bring more affordable Windows Tablets to the masses...if anyone still wants one, or a cheap steam box if they can get it out of the gate fast enough. You have to remember Android is set to become the dominant platform this year.
Did they use Gamecube quality hardware in the NES?
Give Ouya a break, it's a brand new console and it's only on its first generation.
Give the makers time to soak up some feedback on Ouya's weak points and the next version will probably be beefed up a bit.
and as proof, I give you Fatal Fury Special running on the Sega Master System (Game Gear technically, but the hardware is identical) WITHOUT tonnes of custom mappers chips.
As for your SNES, well it was kinda slow. Not like it ran at the same speed as a Colecovision but.. oh wait. It did. And Ranger X broke the color barrier. Heck, most SNES games didn't run base hardware. It's why cart loaders don't work well.
You might be right about the N64, but it didn't help much when the carts where $70 a pop a year after a game launched. As for the gamecube, it was the most powerful system of it's generation and the PS2's library trounced it (Persona 4 anyone? Godhand? Okami was a better Zelda than Zelda).
Hi! I make Firefox Plug-ins. Check 'em out @ https://addons.mozilla.org/en-US/firefox/addon/youtube-mp3-podcaster/
The SMS had more processing power, better graphics and sound.
Processing power was a wash, as Z80 is less efficient clock for clock than 6502. SMS had more color depth, more RAM, and ability to update name. The NES won on sound with an extra bass octave, more timbres for pulse wave instruments, and a digital sample playback channel, and it won on graphics with the ability to scroll a vertical split screen area.
I don't know if it will run well; but running 'better' than a single-core CPU built on an older design and running at less than half the clock speed, with half the RAM, should be essentially automatic.
How well XBMC supports your platform is highly relevant. In theory the RK3066 is a great place to run XBMC but in practice the video acceleration is almost nonexistent and your best hope is that your vendor will release a version which launches the external player.
"You're right," Fisheye says. "I should have set it on 'whip' or 'chop.'"
...no that is just passing the buck.
Passing what buck? To whom? We see always-on DRM on PCs because it's the publishers, not the hardware makers, the console makers are only doing this to appease the content publishers. We see DRM on desktop Linux, OSX, iOS, Android, Playstation and Nintendo platforms as well so 'blame microsoft' is pretty ignorant.
Sadly, the rule with games available on Android...is that they're almost all terrible. There are very few exceptions.
Except that is not even remotely true, having owned a Android game console for over 18 months, its my primary source of gaming, and the costs are cheap too. Android is becoming the primary gaming platform.
Pointing out that the Ouya doesn't compare well with other ARM devices is like complaining that the Cadillac CTS doesn't keep pace with a Ferrarri - the Oya's not supposed to be a work horse. It wasn't designed to be one and it wasn't a priority. The bigger problem it has is that its controller, from what I've read, feels "mushy" and suffers from high latency. That's an actual problem. You can still make some pretty cool games with the console and it's pretty wide open so at least there'll be some enthusiasm by the developer community. For heaven's sake, the homebrew community still writes for the Intellivision! ;) Let's give Ouya some breathing space, I think it's a pretty ambitious little project and everyone's going to learn something from it. I'll pick it up when it's released.
Face it, you can't have great, immersive, polished, professional-quality games for $2.99
Ignoring the fact that you have not looked at Google Play recently :) Lets spend a little time looking at costs.
http://latimesblogs.latimes.com/entertainmentnewsbuzz/2010/02/anatomy-of-a-60-dollar-video-game.html
These figures are rough and back in 2010 by Steve Perlman, founder of OnLive That bring the cost of a video game down to $27. For your $2.99 Andoird game the developers pay $25 for registration to distribute on the Google Play Store. Application developers receive 70 percent of the application price...leaving you with $2.09
A quick look at the console market http://www.vgchartz.com/ and consoles average about 80M potential customers at the end of a consoles useful life. Android is Heading towards 1Billion activations, and continue to grow [currently only 12.5x larger than Consoles].
I am making no claims that more money can be made from Android games than tradition console gaming, but comparing on total selling price alone is foolish when Android market is massive and continues to growl; there is no second hand market; risks are smaller; development costs cheaper; Customers buy more games; Alternative revenue streams.
That is ignoring the fact that your favourite engine spits out binaries that will work on a plethora of platforms...Look at Unity http://en.wikipedia.org/wiki/Unity_(game_engine)
However, people who want to play Android games will play them on their Android phones (if they have one)
Absolutely, and their tablets too. Ignoring the fact that they in themselves are pretty good game platforms. I have owned a Android console from Sony over 18Months http://en.wikipedia.org/wiki/Xperia_Play. Right now the Ouya is not the only successful kick-starter Android console http://gamestick.tv/ or that there are gaming tablets from Archos http://www.archos.com/products/themed/gamepad/index.html?country=us&lang=en#a Wikipad’s and 7-inch Android gaming tablet called Wikipad http://www.wikipad.com/...or even that Sony have introduced native DUALSHOCK 3 controller support for Xperia phones http://techcrunch.com/2013/04/09/sony-adding-dualshock-3-controller-support-to-xperia-devices/ Android gaming as you can see is taking off right now...even with traditional controllers.
You can even do the event handling without touching Java:
http://developer.android.com/reference/android/app/NativeActivity.html