4 ms·

> If thresholds are applied to define discrete values of the quantity, then it’s not analog any more. I agree. It’s my fault I’m not describing this clearly, b

by dahart 3y ago

> If thresholds are applied to define discrete values of the quantity, then it’s not analog any more. I agree. It’s my fault I’m not describing this clearly, but I was referring to switching between digital and analog signals. One can do math using analog circuitry and the interpret the result as digital. The basic problem that I’ve been hinting at is that what we call “digital” has analog components, and what we call “analog” in computing usually turns into digital eventually. (Maybe not analog audio processing, but analog computing does.) It’s often a mix, and there’s a big gray area. > there are endless debates about whether class-D amplifiers are “digital” or not. Right! I believe there are debates about whether modern “analog” audio delay pedals are analog or digital too, right? They are now using more cheap digital components to simulate analog delay lines, which is really if you think about it switching between analog and digital multiple times before you ‘hear’ it. Does it even make sense to summarize it as either analog or digital? > “Noise” is ill defined by itself. It means an unwanted fluctuation of a signal. [..] The vagaries of floating point arithmetic [..] are interesting but basically sideshows. I guess it’s getting lost, but I’m making an analogy between noise and floating point, not claiming that FP rounding error is the same thing as noise. I’m perfectly familiar with noise, and I’m just saying that, in a way, floating point rounding isn’t entirely dissimilar from the net effects of noise: they both are accurate to within a certain tolerance, and they both grow larger errors with chained operations. Fp32 is way more accurate than a typical analog signal, but that’s because it has a lot of bits. Fp8 on the other hand might really have closer to the accuracy that a high quality analog signal has. Digital fp8 might be repeatable where analog isn’t, but that doesn’t necessarily make it more accurate, nor is it something we characterize as “exact” like @corethree was claiming. My only point in this thread was to say analog computations do have some uses, which is why they’re still being used and studied. @corethree said there’s no practical use because analog can be simulated with digital. I’m sure as an audiophile, you’re aware that claim isn’t true. It’s true that analog can be simulated with digital but that’s a slow simulation and has no bearing on whether analog components are useful. One of the analog computing applications I had in mind when I replied was the new Photonic chips. https://www.nature.com/articles/s41586-023-06558-8 https://www.nature.com/articles/s41586-023-06558-8 Another one is quantum computers, they’re quite analog, and definitely different from ‘digital’. A third is fast low-precision AI training chips. There are a whole bunch of startups trying to make these come to market right now, all of them full of smart people who’ve studied the problem and believe that analog computation for AI training solves a real problem.

corethree 3y ago
> what we call “analog” in computing usually turns into digital eventually. uh no. This isn't strictly a rule that's true. It depends on the device whether it's an analog device or a digital device. Not all devices are digital. A DC motor is an analog component, a light bulb, a CRT oscilloscope display, an analog meter or gauge, an H-bridge. All these devices can live on the output side of an analog computer, or even a digital one with the right ADCs or DACs. You're wrong here. >Digital fp8 might be repeatable where analog isn’t, but that doesn’t necessarily make it more accurate, nor is it something we characterize as “exact” like @corethree was claiming. Why are you replying to my statements in a comment to another person? Makes no sense. Anyway, repeatable and exact are sort of similar concepts. If I can repeat a number that means I can replicate it, exactly. Hence the term exact. I also never said digital fp8 is more accurate or exact. There's no such thing as "digital fp8" or "analog fp8" it makes no sense. Digital and analog are orthogonal concepts to floating point. I said electrical analog devices cannot precisely represent a ONE. Simply because you cannot remove all the noise from an analog signal to make it a perfect ONE. To help you understand, instead of floating point, think of an INT8. An an analog computer cannot represent the exact instance of an INT8 in the same way a digital one can. >@corethree said there’s no practical use because analog can be simulated with digital. I’m sure as an audiophile, you’re aware that claim isn’t true. This is probably the only area where you might be sort of right. Sounds like banging on a drum or the sound of plucking a string have origins from analog components. Even though these sounds can be replicated digitally the creation of unique sounds via analog synthesis is better suited to be done by an actual person banging on the actual thing and making the sound he wants. But I wouldn't call this analog computing. > It’s true that analog can be simulated with digital but that has no bearing on whether analog components are useful. Analog components are useful. Analog is what powers everyones home. Analog computing is the thing where analog has virtually no practical usage. Almost 100% of signal processing is now DSP. I'm sure there are really really niche areas where analog is still used, but in general it's not practical. I never said analog components are not useful. Let me just add to this. Your comment wasn't just about the usefulness of analog computing. It was about a "reinvention" of a digital computer that used a base-10 numeral system and parallel data lines with one line for each digit. You misinterpreted that "invention" as a "analog computer." Your responses are more about you not understanding things then it is about analog computers and their practical usefulness. >Another one is quantum computers, they’re quite analog, and definitely different from ‘digital’. No. Quantum computers are not analog. They have qbits, so an aspect of it is discretized into symbols. They aren't digital either. >One of the analog computing applications I had in mind when I replied was the new Photonic chips. https://www.nature.com/articles/s41586-023-06558-8 https://www.nature.com/articles/s41586-023-06558-8 This is Literally the area where I said analog could be better. Speed and energy. Read the paper that's where the improvements are. They reimplemented a Feed forward neural network (something usually done in digital computers) in an analog photonic chip and the main improvements were: Speed and energy. Literally.

by dahart 3y ago

corethree 3y ago

> what we call “analog” in computing usually turns into digital eventually. uh no. This isn't strictly a rule that's true. It depends on the device whether it's an analog device or a digital device. Not all devices are digital. A DC motor is an analog component, a light bulb, a CRT oscilloscope display, an analog meter or gauge, an H-bridge. All these devices can live on the output side of an analog computer, or even a digital one with the right ADCs or DACs. You're wrong here. >Digital fp8 might be repeatable where analog isn’t, but that doesn’t necessarily make it more accurate, nor is it something we characterize as “exact” like @corethree was claiming. Why are you replying to my statements in a comment to another person? Makes no sense. Anyway, repeatable and exact are sort of similar concepts. If I can repeat a number that means I can replicate it, exactly. Hence the term exact. I also never said digital fp8 is more accurate or exact. There's no such thing as "digital fp8" or "analog fp8" it makes no sense. Digital and analog are orthogonal concepts to floating point. I said electrical analog devices cannot precisely represent a ONE. Simply because you cannot remove all the noise from an analog signal to make it a perfect ONE. To help you understand, instead of floating point, think of an INT8. An an analog computer cannot represent the exact instance of an INT8 in the same way a digital one can. >@corethree said there’s no practical use because analog can be simulated with digital. I’m sure as an audiophile, you’re aware that claim isn’t true. This is probably the only area where you might be sort of right. Sounds like banging on a drum or the sound of plucking a string have origins from analog components. Even though these sounds can be replicated digitally the creation of unique sounds via analog synthesis is better suited to be done by an actual person banging on the actual thing and making the sound he wants. But I wouldn't call this analog computing. > It’s true that analog can be simulated with digital but that has no bearing on whether analog components are useful. Analog components are useful. Analog is what powers everyones home. Analog computing is the thing where analog has virtually no practical usage. Almost 100% of signal processing is now DSP. I'm sure there are really really niche areas where analog is still used, but in general it's not practical. I never said analog components are not useful. Let me just add to this. Your comment wasn't just about the usefulness of analog computing. It was about a "reinvention" of a digital computer that used a base-10 numeral system and parallel data lines with one line for each digit. You misinterpreted that "invention" as a "analog computer." Your responses are more about you not understanding things then it is about analog computers and their practical usefulness. >Another one is quantum computers, they’re quite analog, and definitely different from ‘digital’. No. Quantum computers are not analog. They have qbits, so an aspect of it is discretized into symbols. They aren't digital either. >One of the analog computing applications I had in mind when I replied was the new Photonic chips. https://www.nature.com/articles/s41586-023-06558-8 https://www.nature.com/articles/s41586-023-06558-8 This is Literally the area where I said analog could be better. Speed and energy. Read the paper that's where the improvements are. They reimplemented a Feed forward neural network (something usually done in digital computers) in an analog photonic chip and the main improvements were: Speed and energy. Literally.