Separate names with a comma.
Discussion in 'Video Cards and Graphics' started by Hitman928, Jan 7, 2013.
Bought the Nexus 7 form my daughter for Christmas. We all use it, LOL.
I like the concept and i think it's too early to dismiss it.
On the other side, since Sony and Nintendo have similar devices and are well established in the marketplace, nVidia will have to rely on smart/solid advertisement and marketing to succeed. Anyway, only time will tell.
You dual boot Ubuntu and Win7, why? Are you a shill going balls out in defense of Direct X by chance? In a perfect world I'd like to see an industry standard win out. In the gaming market, I can't think of that ever happening. Instead what we get is vendor lock ins giving us the best progress. If you game on Windows- you are supporting a vendor lock in. Given that, is PhysX better then the run of mill physics we normally get? By miles. Its' edge is larger then DirectX's edge over OpenGL. If someone comes on lamenting DirectX I have no issue calling them out, we just don't get a ton of lunatic Linux fans screaming like the lunatic team red around here.
How do your propose they use the same browser? They did use the browser.apk for both devices, it happens that N10's is the Nexus build of Chrome.
Your monitor is a PC component, this is an ARM based gaming platform. Kind of not even close to the same thing
Android games, you previously stated Nexus 7 games which made me think you meant Nexus 7 games(something off of Tegra Zone).
Device functions as both, not to mention it is a killer emulator setup.
Install chrome on both devices. It's in the play store.
Ben, can you please clarify this statement? Are you saying that if Project Shield has to sell 50m units for nVidia to be happy with it? I don't think this can cross 10m to be quite honest.
I was so torn on whether to get the nexus 10 or the nexus 7. The nexus 7 form factor is great, but I just wish that the it had Tegra 4 in it when I was in the market. Ended up with the nexus 10 and still regret it because it's not as portable.
I was mainly using it as a point of reference for the person that didn't think it would be as popular as Eyefinity. The portable market has a *very* different scale of success as you know, and even half a million units, which would be *far* beyond the adoption rate of Eyefinity, would be considered a rather huge failure. I don't expect nVidia to get close to 50 million units for many different reasons, one of which being if Shield moves relatively well I see them updating it on a fairly regular basis(compared to Sony or Nintendo's hand helds).
If they move 3 million in the first year it will exceed my expectations by a long shot.
If they wanted this to go anywhere they'd be better off making it into a platform, as in get a Tegra device and it can take advantage of the features if you buy the controller and hook the device up to it. It might add some complexity in dealing with different resolutions, but the additional complexity it would add is well worth the additional number of sales.
At the CES keynote did they talk about PC hardware or just mobile stuff? I didn't get a chance to see it.
Software on the PC side, something that is likely to be very big for the casual crowd who buys prebuilts-
Well after further thought on the whole situation, I'm disappointed with Nvidia's entire CES event and, more specifically, Tegra 4's initial impressions.
1. Quadcore A15 on TSMC's HPL which is great news for power consumption, it will surely be much more efficient than samsung's exynos.
2. It's the first quadcore A15.
3. The new real-time HDR mode is cool and definitely a plus, but it's not a huge game changed when comparing Tegra 4 to it's upcoming competition.
4. Die size relative to the A6X. But that doesn't matter very much because Apple doesn't sell their chips to anyone.
1. It's entire demonstration was uninspiring and flaky. The only thing JHH could come up with as to why we need/want faster mobile processors is to shave a second off webpage loading? Seriously? That's it? No mention of phones becoming powerful enough to work as a desktop PC when connected to a large screen and blue tooth peripherals for control? No mention of it being as powerful as a console and being able to completely replace said consoles? No mention of more graphically intense games or better video calls?
2. The benchmark they ran wasn't an apples-to-apples comparison as the Nexus 10 ran chrome and the Tegra 4 device was running "browser" (i.e. not the same web browser).
The frame rate of Dead Trigger 2 was extremely low. I hope that was running on Tegra 3 or that it was highly unoptimized because it looked bad.
3. Shield (though not 100% tegra 4 specifically related) won't be a blockbuster selling device. It may sell OK, but it will remain a novelty relative to any other handheld gaming device. That is my opinion, but I think it's both solid and safe to say that Shield isn't going to sell millions upon millions of units.
4. Tegra 4's GPU is still not based on a unified architecture, meaning it's still pre-G80 tech and basically just Tegra 3 on steroids. It doesn't support DX11 or even DX10, and furthermore first benchmarks that leaked the night of the announcement shows that it's GPU is only 3-4x faster than Tegra 3, and more importantly, not even as fast as apple's A6X. http://forum.xda-developers.com/showthr ... ?t=2087129 I'm sure optimizations can and will be made, but 10% improvement is a realistic expectation, and that still puts Nvidia behind Apple's 32nm A6X in GPU power.
Coincidentally, by Nvidia's own admission HERE,
it pretty much sums up that even if the benchmark isn't real, it's actually very close to how it will end up performing - 3.5x faster than Tegra 3 but still noticeably slower than the A6X and barely faster than Samsung's Exynos 5 w/ Mali T604 graphics.
It's just a pure shame that the world's most premier graphics company cannot come out with world-class graphics performance in a market they have been competing in since Tegra 1 first came to market in 2009. I think it would have been a much better move to make a dual-core A15 w/ a power saver core and dedicate more die space to bringing the fastest possible graphics than settling for being second place at best upon release. Is Tegra grey going to be on 28nm or 40nm? Because at this point, if it's on 40nm, it's low end performance will be a tough sell vs. anything qualcomm will have. It desperately needs a die shrink to increase performance without using more power in order to maintain competitiveness.
The other wild card - a potential steam based PC (with an nvidia GPU) running steam big picture mode integrated into Linux (I know Linux is small now, but if given the support from major players like Valve and had a GOOD retail product to back it up, things would change fast) - could be good but only if Valve specifically releases it themselves and treats it more like how the mobile market treats their products - update the product every 12-18 months to improve the specs but maintain the same price and backwards compatibility.
Why would a processor for a non-x86/x64 environment not running windows support Direct X?
What's the problem with 3-4x the game performance? Tegra 4 has the same size as Tegra 3. Over the time it will cost them nearly the same to produce the chip. So you can expect that Tegra 4 will find its way into low cost products, too. How many Samsungs, Apple and Qualcomm low cost tablets you know have better gpu performance? iPad Mini has A5 - 20% faster than Tegra 3...
The steam box is nothing more than a receiver. They will encode the games on a home server and send it over to the box.
Not strictly true. This will be the basic level of Steam Box, with more advanced ones going up from their. And the higher levels of Steam Box will be able to act as a game server to the lower levels. There's an interesting interview with Gabe on the Verge: http://www.theverge.com/2013/1/8/3852144/gabe-newell-interview-steam-box-future-of-gaming
Great read thanks for the link!
This and the Steam box are going to be failures. It's an already crowded market and I don't see where these things fit in.
General heads up, don't trust GLBench. It is utter garbage. Oh, that's right, we are in VC&G, I don't have to dumb it down
Remember back when PVR was trying to push the different Kyro boards and they had benches that would show them going toe to toe and besting a GF2Ultra when in most games they lost pretty badly to a GeForce DDR- same type of bench. It is meant to make TBRs look good and IMRs look bad. Total garbage bench. If you actually watch it, the graphics on it would have looked bad in an Android game from 2009 and it is bringing 2013 hardware to its' knees, it really is an absolutely terrible bench. Futuremark will have their bench out for us very shortly, then we can have a decent mobile GPU bench to compare with. Right now, I go the audacious route of using games, and while I haven't gotten any decent hands on time with A6, A5X was supposedly much better then Tegra3 according to GLBench, but I couldn't find a single game that Tegra3 wasn't better in(in terms of actual visuals).
Shield *IS* a SteamBox. That is why it is so damn interesting too me. Yeah, it's a burly Android hand held whenever I am out. Hit a button and it becomes a SteamBox for the living room in a snap? Yeah, that is something I am *most* interested in. That is what I think makes Shield so interesting. The Android functionality of it is cool, but really we could get all of that using a Nexus 7 and an added controller. But add the SteamBox functionality in? That makes it very compelling to me.
I don't think a valve made PC will be a failure. It may not attract the top tier PC enthusiasts, but it isn't meant to. It will be AAA PC gaming, including mods, better frame rates, better graphics, and more robust browsers and usability than consoles, all with the simplicity of a console and controlling input the user wants ( keyboard, mouse, or controller as opposed to being pigeon holed into strictly just controller).
The steambox hasn't been unveiled yet, the "steam friendly PC" from Xi is *not* the official steambox.
Valve is being very flexible in allowing third parties to create steam friendly PCs, but this is not Valve's version - We don't have many details yet but i'm pretty confident that they will make a great device.
Windows dead on arrival edition? That one?
It was a good idea but the apps are still low in number and aren't good. Surface RT had some nice intentions, but ultimately will not be taken seriously. I think the Surface pro is going to be great, though - in every metric it is better and smaller than the macbook air or any ultrabook. Now, the price on the other hand...
I like what I did see with the I500 modem and Tegra 4!
Finally some good news on Tegra 4: http://www.anandtech.com/show/6666/the-tegra-4-gpu-nvidia-claims-better-performance-than-ipad-4
TL;DR the leaked GLbenchmark was from very early silicon, and Nvidia says Tegra 4 finalized silicon will be faster in games and in GLbenchmark than A6X. I still think JHH and Nvidia's presentation of Tegra 4 was nothing but awful and uninspiring, but this bit of news does carry some heavy weight and reversal of what was a string of bad news after T4's intro.
I have also come to the conclusion that SHIELD was thought up and created as an enticing way to "sell" T4 chips that otherwise won't yield good enough for phones or tablets. Profit margins won't be high, but it's certainly better than throwing functional T4 chips out.
I wasn't aware that RT used Direct X. Looking into it it appears to only support DX9. So it's I guess it's more microsoft's limitation than it is Nvidia not added dx10/11 support.