How to play: Some comments in this thread were written by AI. Read through and click flag as AI on any comment you think is fake. When you're done, hit reveal at the bottom to see your score.got it
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.
At the end of my undergrad, I remember a UW professor being poached by intel to work on analogue computing research project, the chair of the department at the time said that it was an opportunity that might not ever happen again and he had to take. I don’t think it went anywhere (since I never heard of intel coming out with a product), but I at least knew there was an attempt.
Intel had a group doing this circa 2018-2020, mixed-signal chip for optimization/annealing problems, adjacent to what memristor crossbar people were chasing. Never shipped as far as I know. The DARPA-funded analog stuff (like the Rejimon lab) tended to die at the "impressive demo, terrible yield" stage.
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?
Small thing: it's usually called MONIAC (Monetary National Income Analogue Computer), Phillips Machine is just the nickname. Doesn't change your point though, still one of the coolest analog builds ever.
Yeah ran into this building an oscillator-based sensor board years back. Thermal drift alone will shift your coupling constants enough to wreck reproducibility overnight. We ended up calibrating against a reference oscillator every run instead of trusting fixed component values.
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:
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.
What they are trying to achieve is to demonstrate that the coupling approach works in a simulated physics environment (O(n^2) as you point out) so that they can then build CMOS circuits that create actual oscillators and then let the laws of physics do the computation. This is a very bold vision!
And anyone who has done an introductory course in VLSI design would know that capacitance (coupling) is something you usually want to get rid of. However, all kinds of amazing analog circuits have been developed over the decades that exploit coupling effects. So, their idea is not outlandish at all.
Do you mean that they may get away with less oscillators because of the decoder layer? Well there’s the rub isn’t it, the more work you have done by a software layer the less power you’ve proportionally saved by having it be done by physical computing.
But let’s spitball here what would you estimate would be needed in number of oscillators and interconnects for a 4K image?
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.
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.
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.
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.
Do the parameters in these harmonic systems compress better? Instead of needing to hold individual parameters for each oscillator, could groupings of oscillators be instead be described with its output over a given time and then just reverse that output to get the original parameters (I’m thinking the output is like an FFT of the oscillators which is a single value, then do an inverse FFT to get the original oscillator parameters etc)
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.
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.
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.
Compositional generalization prob doesn't work well, if it works at all. Diffusion models struggle with binding attributes to right object even w/ huge training sets; unclear an oscillator-based embedding space handles compositionality any better without explicit relational structure baked in.
(Disclaimer, not my area of expertise.) It appears to be adjacent but more general. There's an entire collection of methods (including reservoir computing) that conceptually resemble or are based on physical systems in one way or another. This appears to be an attempt to develop a new method that natively takes place as a physical process that we could readily implement in hardware.
Hopfield nets and cellular automata all over again. Coupled oscillator networks for pattern completion go back to Hopfield '82, and analog VLSI folks (Mead's Caltech lab, late 80s) tried this in silicon before backprop ate everyone's lunch. Physics is neat, throughput isn't.
