When I first learned about computer science at the age of 11 or so (and in 1982 or so) the first page of the text book put digital and analogue computers on what seemed to be an equal footing. And then proceeded to ignore the latter for the rest of the book. Apart from a few notable exceptions ( https://en.wikipedia.org/wiki/Phillips_Machine ) I've often wondered about analogue computing.
Quantum annealers (D-Wave machines) are basically analogue computers, with Josephson junctions as the primary component as opposed to oscillators. I wonder if they could render these images, too?
If you want to understand the issue with analog computers, design a SHA-256 circuit for one of them and consider the consequences of trying to push a megabyte of data through it. While that is an extreme example I chose precisely to make the issues clear, much real computing has many of the same characteristics, just distributed a bit more widely in time and space.
Or, to put it another way, you can make anything sound good if you consider only the positives and anything sound bad if you only consider the negatives. Analog computing sounds amazing when you read the brochure and consider only the positives. But when you bring the negatives back in, it makes sense why it is not frequently used. It is not a case of the mainstream keeping some great idea down because, uh, Big Digital or something, it's a case where digital computing turns out to be a stonking good idea and it's hard for the analog world to compete and it's virtually impossible for them to ever be anything but a niche.
Neural networks are an interesting possibility for a future successful niche, although even so, it would be neural networks specifically that may grow in importance and not analog computing in general. And I still wouldn't guarantee it'll be a good idea... we may have a lot of trouble keeping what would be very deeply nested analog circuitry stable in the real world and digital may still win out, e.g., an analog neural net that has a noticeable personality shift when it gets warmer may not be the best engineering solution. That's a question for 20 or 30 years from now.
Yes, I had the same question. I don’t think so, as currently designed. It trains to specific points / classes in an embedding space. They didn’t discuss how one might go to non-trained points in the paper as far as I could read, and they did show some visualization around the idea that the runs aim at / around set points in the space.
This method is cool and the post explains it well. It would, however, be good to get more detail on the energy efficiency they flag as their motivation: is this model actually more energy efficient than the comparators they highlight?
It seems like total parameter count is more or less on par with conventional approaches so any gains won't be from there.
We can implement coupled oscillators in hardware but are the couplings and frequencies programmable? If they're being streamed in I guess you'd still have a memory bandwidth bottleneck and associated energy usage. If not then the fair comparison is to a conventional model hardcoded in an ASIC which AFAIU is actually quite energy efficient.
Really interesting - if I understood the article correctly, they're simulating this on conventional hardware, so in order to get the proposed benefits, it would need to be implemented in some other electronic medium.
It’s not clear to me how this would ever be practical since it seems dependent on n^2 scaling.
You’ve got to wonder when you have an image generation demo why would you possibly have 64 x 64 pixel output as your demo?
If I’m understanding this properly to generate a 4K image, you need like 5 trillion point to point connections on the chip. Even if power use from the oscillators is zero that’s going to be an issue.
Yes I too am perplexed. I'm into audio synthesis so I feel I have somewhat better-than-average knowledge of oscillators, from the component or elementary mathematical level (depending on whether they're analog or digital) to complex interactions for fun and profit (frequency, phase, ring modulation).
These are cool results but I was disappointed not to find any discussion of where oscillator array technology stands today what the manufacturing challenges/opportunities might be. It seems like it would be prohibitively expensive for anything beyond minimal networks of a few hundred nodes that could be used in sensors. Even if you have perfectly consistent oscillators that synchronize to each other within very fine tolerances, wiring them up to each other is still a massive headache.
Think of the models making progress on CIFAR-10, ImageNet, CelebA, etc. 15 years ago. They had issues too and weren't just scaled-up as is to the architectures we have today.
I read through the article, and I'm not sure this is dependent on quadratic scaling.
Are they allowing all oscillators to influence all others, or are they picking modalities where the influences can be limited to some maximal fixed degree?
One would imagine that there'd be a variety of different topologies available to explore. Even if during training the treatment was fully connected, one could imagine the training itself biasing towards a maximal fixed degree per oscillator, and then inference later operating on a quantized version of that that drops the low-weight influences to zero.
Very cool work - refreshing to see a of different approach. I learned about Kuramoto oscillators many years ago from a book called Sync, by Steven Strogatz, which I highly recommend.
I love this ! Used to work at Rain AI on training neural networks in unconventional hardware - people often that computers don't necessarily have to be electronic digital - there is a whole domain dedicated to creating machines that can apply certain mathematical operations faster or more efficiently than their electronic counter parts. I created this site to try create a classification of that space:
Very cool. I’m reminded of Wolfram’s pitch that neural nets are a search through the very broad computational complexity of the function space they describe; he did a little work to show that you could find similar behavior in other function spaces. These oscillators are yet again a different function space, and its cool they can be harnessed in this way.
The question of what physical / electronic phenomena is the most efficient yet large enough function space to be used for inference is a really good one to think about. I have no suggestions.
The required compute seems a bit high: "We trained all CIFAR-10 models on 1xB200 GPU, and all ImageNet 64×64 models on 8xB200 GPUs. The largest CIFAR-10 model uses 20 B200 hours to train, and the largest ImageNet 64×64 model uses 640 B200 hours"
They're proposing a loose range of hardware that is thought to be more efficient when used with this thing, that's the entire point. What they did with B200s is just a simulation, it's not supposed to be GPU-efficient.
31 comments
[ 4.0 ms ] story [ 55.6 ms ] threadOr, to put it another way, you can make anything sound good if you consider only the positives and anything sound bad if you only consider the negatives. Analog computing sounds amazing when you read the brochure and consider only the positives. But when you bring the negatives back in, it makes sense why it is not frequently used. It is not a case of the mainstream keeping some great idea down because, uh, Big Digital or something, it's a case where digital computing turns out to be a stonking good idea and it's hard for the analog world to compete and it's virtually impossible for them to ever be anything but a niche.
Neural networks are an interesting possibility for a future successful niche, although even so, it would be neural networks specifically that may grow in importance and not analog computing in general. And I still wouldn't guarantee it'll be a good idea... we may have a lot of trouble keeping what would be very deeply nested analog circuitry stable in the real world and digital may still win out, e.g., an analog neural net that has a noticeable personality shift when it gets warmer may not be the best engineering solution. That's a question for 20 or 30 years from now.
We can implement coupled oscillators in hardware but are the couplings and frequencies programmable? If they're being streamed in I guess you'd still have a memory bandwidth bottleneck and associated energy usage. If not then the fair comparison is to a conventional model hardcoded in an ASIC which AFAIU is actually quite energy efficient.
You’ve got to wonder when you have an image generation demo why would you possibly have 64 x 64 pixel output as your demo?
If I’m understanding this properly to generate a 4K image, you need like 5 trillion point to point connections on the chip. Even if power use from the oscillators is zero that’s going to be an issue.
These are cool results but I was disappointed not to find any discussion of where oscillator array technology stands today what the manufacturing challenges/opportunities might be. It seems like it would be prohibitively expensive for anything beyond minimal networks of a few hundred nodes that could be used in sensors. Even if you have perfectly consistent oscillators that synchronize to each other within very fine tolerances, wiring them up to each other is still a massive headache.
Are they allowing all oscillators to influence all others, or are they picking modalities where the influences can be limited to some maximal fixed degree?
One would imagine that there'd be a variety of different topologies available to explore. Even if during training the treatment was fully connected, one could imagine the training itself biasing towards a maximal fixed degree per oscillator, and then inference later operating on a quantized version of that that drops the low-weight influences to zero.
https://computers.tugdual.fr/
The question of what physical / electronic phenomena is the most efficient yet large enough function space to be used for inference is a really good one to think about. I have no suggestions.
20 B200 hours for CIFAR-10 seems like a lot...