Nexus 9 To Be First Device With Android L And Nvidia Denver CPU
The date has been leaked several times in the past few weeks, but it seems tomorrow, October 15, is the real date for the launch of the Nexus 9 tablet, according to a source close to Google.
Google had initially planned to hold an official event to unveil the tablet (and possibly a few other products) but decided against a "big reveal," because Android L apparently still needs a few more tweaks.
Google chose HTC as its manufacturing partner to make the Nexus 9. The last time HTC made a Nexus device was early 2010, when it built the first Nexus phone ever. The collaboration with Google may give HTC another image boost, much like the company got when it created the Nexus One. A Nexus partnership also proved favorable to Asus and LG because of all the media attention the companies received from making Nexus devices.
The Nexus 9 specs seem to be unchanged from the leaks we've seen so far. The tablet will have an 8.9" screen with a 2048 x 1440 resolution, which is a slightly strange screen ratio that's somewhere between 4:3 and 3:2. To be 4:3, like the iPad, the tablet would need a 1920 x 1440 resolution.
Although it has been rumored for months, there haven't been too many other leaks about Denver-based devices coming out soon, and thus the rumor about the Nexus 9 being the first to use a 64-bit Denver CPU seems a little too good to be true. The fact that Google has recently added support only for 64-bit Intel chips to its Android emulator, but not for ARMv8, doesn't help.
According to the Google source, the Nexus 9 tablet will indeed come with an ARMv8-based Denver CPU from Nvidia and a Kepler GPU. The chip could potentially make the Nexus 9 the fastest mobile device on the market, even when pitted against the soon-to-be-launched iPad Air 2 and its A8X processor, despite the fact that Denver will be built on a 28nm processor while A8X will be built on a 20nm process.
Of course, this is based on Nvidia's own initial benchmarks, which show the dual-core Denver to be about as powerful as a 22nm FinFET Haswell Celeron and roughly twice as powerful as a high-end Snapdragon or even the previous Tegra K1 processor. Whether this prediction holds true remains to be seen, but chances are that it will be the fastest mobile chip around.
The Nexus 9 will also arrive with an 8MP rear camera and a "brushed aluminum frame," and it will weigh 480g. "BoomSound" fans will be happy to hear that HTC's excellent dual-speakers will be available on the tablet as well. An origami-style cover will also be available for the device and will apparently bend into any shape to hold the tablet in any orientation you want.
The Nexus 9 will be unveiled tomorrow in a blog post, and it will be available for pre-order on October 17 and for purchase on November 3. There will be a $400 version with 16 GB of internal storage and a $500 model with 32 GB of storage and LTE support. Both models will apparently support external storage, too, which would be a first for Nexus devices.
Follow us @tomshardware, on Facebook and on Google+.

I might leave Samsung for a Nexus 9.
That said benchmarks on a tablet are pretty useless to begin with. Tablets are normally single use devices where you're doing only one interactive program at a time. So long as it will play video smoothly, display readable text, and replay music without chop a tablet is no different now from the first generation android tablets.
The entire question isn't how fast it is, rather it's how stable it's software will be. That is a complete unknown with a new generation processor AND a next generation software stack.
I might leave Samsung for a Nexus 9.
I might leave Samsung for a Nexus 9.
8inch tablets now have 2560*1600 screen. The nexus 10 had that 2 years ago.
Going lower than an ipad iwould be a big dissapointment.
If they want a 4:3 screen, then they should go just like a chromebook pixel with 2560*1700.
8inch tablets now have 2560*1600 screen. The nexus 10 had that 2 years ago.
Going lower than an ipad iwould be a big dissapointment.
If they want a 4:3 screen, then they should go just like a chromebook pixel with 2560*1700.
It actually has the same resolution as the iPad. And 2560x1700 is not 4:3...
You must have really big pockets.
That said benchmarks on a tablet are pretty useless to begin with. Tablets are normally single use devices where you're doing only one interactive program at a time. So long as it will play video smoothly, display readable text, and replay music without chop a tablet is no different now from the first generation android tablets.
The entire question isn't how fast it is, rather it's how stable it's software will be. That is a complete unknown with a new generation processor AND a next generation software stack.
I might leave Samsung for a Nexus 9.
I might leave Samsung for a Nexus 9.
You apparently missed all the shield tablet scores. K1's scores are not an unknown, and Denver will probably just match it+better battery life due to in house cpu. That's not a jab, that would still be an awesome soc, just saying what I expect. The best part is nobody has to retool as it is pin compatible, so anyone making a K1 design now, should be able to just slot in the Denver version and boom, selling 64bit shortly after. Very wise on NVs part doing this. Regarding the problems, you're forgetting everyone will have the same issues with new software/stack etc and everyone will be moving to 64bit (hardware and software).
Since most of Android is simply Java running on top of Dalvik/ART, there is not much "64-bitness" for software beyond the core OS and runtime to worry about.
Since most of Android is simply Java running on top of Dalvik/ART, there is not much "64-bitness" for software beyond the core OS and runtime to worry about.
I didn't say it was difficult (or not), just that everyone deals with the same thing here. Not sure why your post says dragonsqrrl said it, I did
Evil comment editing: every time I quote a message but do decide to not post a reply, the text from all the unsent quoted messages get prepended to the next messages I reply to until I post a reply. Since the quote blocks only contain numeric message IDs, quotes get screwed up if I remove the wrong parts while cleaning up that mess of stray quotes.
Not sure if that is a feature or a bug. I'd have to try it from Chrome or IE to see if this is a server-side or browser-side thing.
Evil comment editing: every time I quote a message but do decide to not post a reply, the text from all the unsent quoted messages get prepended to the next messages I reply to until I post a reply. Since the quote blocks only contain numeric message IDs, quotes get screwed up if I remove the wrong parts while cleaning up that mess of stray quotes.
Not sure if that is a feature or a bug. I'd have to try it from Chrome or IE to see if this is a server-side or browser-side thing.
ROFL...It's a feature...Honest...
Anyway, I get your point
*grits teeth tries not to swear*