[HN Gopher] M1076 Analog Matrix Processor
___________________________________________________________________
M1076 Analog Matrix Processor
Author : tosh
Score : 78 points
Date : 2023-11-15 15:22 UTC (7 hours ago)
(HTM) web link (mythic.ai)
(TXT) w3m dump (mythic.ai)
| liquidify wrote:
| What makes it analog?
| ganzuul wrote:
| > In practice it is possible, in a flash memory made on a
| mature process such as 40nm, to store reliably a range of
| charges that correspond to a digital resolution of 8 bits.
|
| - https://mythic.ai/wp-
| content/uploads/2022/02/MythicWhitepape...
| strbean wrote:
| Another interesting excerpt:
|
| > When a charge is programmed into a flash memory device, its
| electric field has an effect on any signal passing through
| it. In the Mythic architecture, the flash transistor acts as
| a variable resistor that reduces the signal level passing to
| the output. That reduction is proportional to the analog
| value stored in the memory. This simple effect implements the
| multiplication stage found in DNN calculations. The
| accumulation process, in which the output from each of those
| calculations is summed, is handled by aggregating the output
| of an entire column of memory cells. Thanks to these two
| properties, the Mythic architecture can process an entire
| input vector in a single step rather than iterating at high
| speed as in a digital processor
| teruakohatu wrote:
| If you are wondering about the analog part:
|
| > Analog computing provides the ultimate compute-in-memory
| processing element. The term compute-in-memory is used very
| broadly and can mean many things. Our analog compute takes
| compute-in-memory to an extreme, where we compute directly inside
| the memory array itself. This is possible by using the memory
| elements as tunable resistors, supplying the inputs as voltages,
| and collecting the outputs as currents. We use analog computing
| for our core neural network matrix operations, where we are
| multiplying an input vector by a weight matrix.
|
| > Analog computing provides several key advantages. First, it is
| amazingly efficient; it eliminates memory movement for the neural
| network weights since they are used in place as resistors.
| Second, it is high performance; there are hundreds of thousands
| of multiply-accumulate operations occurring in parallel when we
| perform one of these vector operations. Given these two
| properties, analog computing is the core of our high-performance
| yet highly-efficient system.
|
| https://mythic.ai/technology/analog-computing/
| nabla9 wrote:
| I suspect that this is where all inference processors are
| heading eventually. The benefits are just too great when exact
| computation is not required.
|
| Training might be harder to implement.
| pclmulqdq wrote:
| This is a very specific technology that has serious scaling
| issues, and the neural networks coming out today are huge in
| comparison to YOLOv5 and ResNet. The company has already
| failed once. This will probably have its niche in some
| computer vision stuff for a little while, but models have
| already outgrown it.
| nabla9 wrote:
| This particular implementation of the idea of analog matrix
| processor implemented at low level may fail, but some
| different implementation will succeed.
| pclmulqdq wrote:
| There are some fundamental limits to analog computing
| that start to really hurt at the small nodes, which you
| need to scale up. There's a very good reason they are
| stuck at 40 nm.
| mwbajor wrote:
| I have an analog circuits and a bit of an analog computing
| background and I will tell you this: analog computers have
| been around for awhile in commercial applications, still to
| this day theyre used to great success in some niche
| applications. They're limited by their flexibility as they
| are not really reprogrammable. Analog circuits need to be
| designed for each specific application. There are papers
| and prototypes built to create "reconfigurable" analog
| versions of FPGAs but they are limited by physical
| scalability issues, noise, routing, etc.
| inasio wrote:
| It's basically Ohm's law V = RI, or in this case I = CV
| (conductance times voltage), on a lattice of wires. On many
| analog devices setting the values (write) can be very time
| consuming, although the read operation can indeed be fast and
| super low energy. Unclear if this is the case here.
| dumbo-octopus wrote:
| Veritasium did a decent explainer/interview with this company,
| about this product. https://www.youtube.com/watch?v=GVsUOuSjvcg
| moffkalast wrote:
| I don't suppose these will be retail purchasable any time soon?
| Would be a useful thing to plug into the Pi 5's new PCIe port,
| not unlike a NCS2. Although with probably very spotty software
| support...
| ThinkBeat wrote:
| Hmm
|
| This chip is said to perform 25 TOPS. A A100 80GB SXM is said to
| perform 1248 TOPS. Which is an entire card not a single CPU.
|
| You could theoretically achieve the same with 50 M1076
| processors.
|
| Would the benefit be any of:
|
| * a lot cheaper * more energy efficient. * small size * easier to
| mass produce?
| blovescoffee wrote:
| But for small / embedded models, you'd much much prefer to use
| this chip from mythic than a huge (and possibly more expensive)
| a100
___________________________________________________________________
(page generated 2023-11-15 23:00 UTC)