The Pixel 2's Dormant 'Visual Core' Chip Gets Activated In Latest Android Developer Preview (techcrunch.com)

The Google Pixel 2 and Pixel 2 XL both feature a custom Intel "Visual Core" co-processor, which is meant to improve speed and battery life when shooting photos with Google's HDR+ technology. The chip has been hanging out in the phone not really doing much of anything -- until now. TechCrunch reports of a new developer preview of Android 8.1 due out today that puts the chip to use. "The component is expected to further improve the handsets' cameras, which were already scoring good marks, production issues aside." From the report: According to the company, Pixel Visual Core has eight image processing unit (IPU) cores and 512 arithmetic logic units. Using machine learning, the company says it's able to speed things up by 5x, with one tenth of the energy. Access to the chip, combined with the Android Camera API means third-party photo apps will be able to take advantage of the system's speedy HDR+. Sounds swell, right? Of course, this is still just an early preview, only available to people who sign up for Google's Beta program. That means, among other things, dealing with potential bugs of an early build. Google wouldn't give us any more specific information with regards to when the feature will be unlocked for the public, but it's expected to arrive along with the 8.1 public beta in December.

Comments

Shark jumping tipping point

[Anonymous Coward]

I figure there's two types in Google right now, a group that can barely code a "hello world" app, and a group that knows their subject in real detail, and they're in balance making Google at the tipping point.

Look at the Android phones and tablets. Google's Pixel 2: The plastic section starts at the same place as the volume button making it mechanically weak. That passed both design and testing phases. Or tablets that turns portrait when running a portrait app, forcing you to pick up the device, keyboard and all, when the portrait app runs on many devices normally a lot smaller. Why wouldn't you simply give it a phone sized portrait window? Why treat these as non-resizable in Chrome when they really support lots of different sized phones so are clearly resizeable!...... again dumbos in there. Or material design man, the radio button and checkbox that are both circles and you only know its action by pressing it to see what happens.... dumb design.

And yet here is a group that has been working with machine learning and design a chip to speed up the processing of their phone cameras.

None of the Google devices sell in any volume, they'll never match even a few % of iPhoneX's sales because there are too many dumbos and not enough geniuses there.

It's like they're at a tipping point.

Re:Shark jumping tipping point

[jellomizer]

Google thinks in software. Apple thinks in hardware.
This isn’t a bad thing but different approaches to problems.
Most facial recognition systems rely on smart software to look at the face and determine if it is you or not. Apple puts in some hardware to put IR dots on your face to make the software simpler.
The pixel camera has a lot of software features to counteract hardware limitations. Apple will try to put more hardware to the problem.
The hardware thinking is often leads to a more expensive product, that can be improved by better software in the future . While software thinking often come to peak ability much quicker requiring a hardware upgrade. However the product is cheaper.

Re:Shark jumping tipping point

[AmiMoJo]

Due to the form factor of a phone and the limited options available, everyone uses the same Sony sensors in their high end phones. Lens technology peaked years ago.

Most manufacturers use optical image stabilization, while Google uses digital. Results seem about the same for photos and for video digital is clearly superior, so some are starting to add digital to their optical systems when in video mode.

Photo quality wise it's all about software processing, since everyone has basically identical hardware. Some phones now have two cameras that combine images, but none produce results as good as Google's single camera+software. Google's HDR performance in particular is way ahead of anyone else, especially at night. Night shots are probably Apple's biggest weakness.

The other use for dual cameras "portrait mode", where the background is automatically blurred. Again, Google does it entirely in software and it seems to be better than any dual camera solution. Most of them, including Apple, struggle to find the edges of your head. It's a hard task, especially given the variety of hair and how thin/transparent it can get at the margins. Google claims to use AI, presumably Tensorflow.

Re:Cameraphones still take poor pictures

[im_thatoneguy]

HDR+ in a cameraphone is like saying a 4K screen is better than a 720p one for playing 720p video.

No it's like saying a 4k screen is better than a 720p one for 4k video.

HDR+ increases the dynamic range of the sensor. The 3 most important metrics of an image are: Dynamic range\Noise (check), Color Rendition and Sharpness. Cameraphones are already plenty sharp, nearly all of the top camera phones shoot DNG uncompressed RAW and now they're implementing improved HDR modes to extended the dynamic range.

I own an 8k RED camera but almost all the photos I take are on my smartphone because the image is reasonably good. If I had one request for smartphone sensors it would be "more dynamic range" aka HDR+.

Intel ME?

[mrwireless]

To what extent is this SOC comparable to Intel's management engine? Isn't Google basically putting a second computer in my computer? With all the risks that entails?