Nvidia Brings Kepler to Mobile Devices

Nvidia has a long history in mobile of over-promising on performance and power consumption. Given the expected ship date of this product it would not be surprising for it to be only slightly faster (or quite possibly a bit slower) than products that will ship in the interim. See: Every single Tegra.

I am a total novice when it comes to chip design; but aren&#x27;t there obvious tradeoffs with any architecture? Isn&#x27;t it really hard to have a silver bullet that is super powerful &#x2F; wide in its processing path and efficient for power?<p>I would imagine that for mobile, the &quot;insert favorite gpu metric here&quot; per watt would be more important than the aggregate metric.

It would be fantastic if convergence devices like Ubuntu Edge could use this in both a mobile as well as a &quot;docked&quot; power state.<p>The way I&#x27;m envisioning it, the processor would detect when the device is docked in a full power desktop or notebook configuration. At that point it would switch to a higher power profile for PC gaming and work station class as computing tasks.

I would really like to be excited by this, but since the stretch goal for the mobile industry is building the next plants vs zombies, my expectations are minimal.

It&#x27;s telling how they mention incredible performance&#x2F;power ratio only in passing, but focus their pitch on the unified architecture across different classes of machine (mobile to supercomputer), rather like Fred Brooks&#x27; 360 project at IBM.

This needs to be in the next OUYA or version 2.0 of the rumored Google console. I&#x27;d pay $200 for such a console with a great controller. Not sure about SHIELD. Portable consoles are not really my thing, and that extra screen and battery adds a lot to the cost, but I imagine it would be a lot more compelling with it, too.

I&#x27;d say this was one of those cases where changing the original headline was perfectly acceptable.