5 ms·

Are the neural engines actually used for anything, to anyone's knowledge? Edit: Apparently in iPhones, they are used for FaceID.

can16358p 5y ago
Probably object tracking in videos will be the best use of them. Or, there will be some new form of video generation (like the ones generating video from Deep Dream etc, but something aimed at studio production) using ML that wasn't practically usable before. It opens many doors, but it will take at least many months, if not years, to see some new "kind" of software to emerge that efficiently makes use of them.
acchow 5y ago
I thought it's also used to activate Siri by voice without any CPU usage
sharikous 5y ago
CoreML models may run on them to macOS's discretion. If you manage to get your neural network in CoreML's format you may use it.
daggersandscars 5y ago
Adobe uses them Lightroom / Photoshop for some functions. https://www.digitalcameraworld.com/news/apple-m1-chip-makes-adobe-apps-80-faster-and-some-gains-are-bigger-still https://www.digitalcameraworld.com/news/apple-m1-chip-makes-...
bm-rf 5y ago
Would something like huggingface transformers ever be able to support this? Or is it best fit to just use the GPU.
jazzyjackson 5y ago
Adobe Photoshop, Premiere etc make use of it for scene detection, content aware fill, "neural filters" and so on
- poyu 5y ago
  This[1] neural filter? [1] https://www.youtube.com/watch?v=hq8DgpgtSQQ https://www.youtube.com/watch?v=hq8DgpgtSQQ
xiphias2 5y ago
I think the most important use of the neural engines so far is for the internal camera postprocessing. Better camera postprocessing is the reason why people buy new iPhones.
- piyh 5y ago
  Throw in translation, on device image labeling, stuff like on body walking/biking detection, voice recognition.
sercand 5y ago
Neural Engines may be used in CoreML models. I don't know it can be used with Apple's BNNS library [1]. You can use with TensorFlow Lite with coreML delegate as well [2]. And some tried to reverse engineer it and used it for model training [3]. [1] https://developer.apple.com/documentation/accelerate/bnns https://developer.apple.com/documentation/accelerate/bnns [2] https://www.tensorflow.org/lite/performance/coreml_delegate https://www.tensorflow.org/lite/performance/coreml_delegate [3] https://github.com/geohot/tinygrad#ane-support-broken https://github.com/geohot/tinygrad#ane-support-broken
speed_spread 5y ago
It's to build a giant distributed Aleph in which a preserved digitized Steve Jobs can live once again.
iamgopal 5y ago
Or does any popular ML libraries support it ?

can16358p 5y ago

Probably object tracking in videos will be the best use of them. Or, there will be some new form of video generation (like the ones generating video from Deep Dream etc, but something aimed at studio production) using ML that wasn't practically usable before. It opens many doors, but it will take at least many months, if not years, to see some new "kind" of software to emerge that efficiently makes use of them.

acchow 5y ago

I thought it's also used to activate Siri by voice without any CPU usage

sharikous 5y ago

CoreML models may run on them to macOS's discretion. If you manage to get your neural network in CoreML's format you may use it.

daggersandscars 5y ago

Adobe uses them Lightroom / Photoshop for some functions. https://www.digitalcameraworld.com/news/apple-m1-chip-makes-adobe-apps-80-faster-and-some-gains-are-bigger-still https://www.digitalcameraworld.com/news/apple-m1-chip-makes-...

bm-rf 5y ago

Would something like huggingface transformers ever be able to support this? Or is it best fit to just use the GPU.

jazzyjackson 5y ago

Adobe Photoshop, Premiere etc make use of it for scene detection, content aware fill, "neural filters" and so on

poyu 5y ago
This[1] neural filter? [1] https://www.youtube.com/watch?v=hq8DgpgtSQQ https://www.youtube.com/watch?v=hq8DgpgtSQQ

xiphias2 5y ago

I think the most important use of the neural engines so far is for the internal camera postprocessing. Better camera postprocessing is the reason why people buy new iPhones.

piyh 5y ago
Throw in translation, on device image labeling, stuff like on body walking/biking detection, voice recognition.

sercand 5y ago

Neural Engines may be used in CoreML models. I don't know it can be used with Apple's BNNS library [1]. You can use with TensorFlow Lite with coreML delegate as well [2]. And some tried to reverse engineer it and used it for model training [3]. [1] https://developer.apple.com/documentation/accelerate/bnns https://developer.apple.com/documentation/accelerate/bnns [2] https://www.tensorflow.org/lite/performance/coreml_delegate https://www.tensorflow.org/lite/performance/coreml_delegate [3] https://github.com/geohot/tinygrad#ane-support-broken https://github.com/geohot/tinygrad#ane-support-broken

speed_spread 5y ago

It's to build a giant distributed Aleph in which a preserved digitized Steve Jobs can live once again.

iamgopal 5y ago

Or does any popular ML libraries support it ?