First, Virtual AI Compute and Now a Virtual Supercomputation Complex
December 19, 2025
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
Do you remember the good old days at AT&T? No Judge Green. No Baby Bells. Just the Ma Bell. Devices were boxes or plastic gizmos. Western Electric paid people to throw handsets out of a multi story building to make sure the stuff was tough. That was the old Ma Bell. Today one has virtual switches, virtual exchanges, and virtual systems. Software has replaced quite a bit of the fungible.
A few days ago, Pavel Durov rolled out his Cocoon. This is a virtual AI complex or VAIC. Skip that build out of data centers. Telegram is using software to provide an AI compute service to anyone with a mobile device. I learned today (December 6, 2025) that Stephen Wolfram has rolled out “instant supercompute.”
When those business plans don’t work out, the buggy whip boys decide to rent out their factory and machines. Too bad about those new fangled horseless carriages. Will the AI data center business work out? Stephen Wolfram and Pavel Durov seem to think that excess capacity is a business opportunity. Thanks, Venice.ai. Good enough.
A Mathematica user wants to run a computation at scale. According to “Instant Supercompute: Launching Wolfram Compute Services”:
Well, today we’ve released an extremely streamlined way to do that. Just wrap the scaled up computation in RemoteBatchSubmit and off it’ll go to our new Wolfram Compute Services system. Then—in a minute, an hour, a day, or whatever—it’ll let you know it’s finished, and you can get its results. For decades I’ve often needed to do big, crunchy calculations (usually for science). With large volumes of data, millions of cases, rampant computational irreducibility, etc. I probably have more compute lying around my house than most people—these days about 200 cores worth. But many nights I’ll leave all of that compute running, all night—and I still want much more. Well, as of today, there’s an easy solution—for everyone: just seamlessly send your computation off to Wolfram Compute Services to be done, at basically any scale.
And the payoff to those using Mathematica for big jobs:
One of the great strengths of Wolfram Compute Services is that it makes it easy to use large-scale parallelism. You want to run your computation in parallel on hundreds of cores? Well, just use Wolfram Compute Services!
One major point in the announcement is:
Wolfram Compute Services is going to be very useful to many people. But actually it’s just part of a much larger constellation of capabilities aimed at broadening the ways Wolfram Language can be used…. An important direction is the forthcoming Wolfram HPCKit—for organizations with their own large-scale compute facilities to set up their own back ends to RemoteBatchSubmit, etc. RemoteBatchSubmit is built in a very general way, that allows different “batch computation providers” to be plugged in.
Does this suggest that Supercompute is walking down the same innovation path as Pavel and Nikolai Durov? I seem some similarities, but there are important differences. Telegram’s reputation is enhanced with some features of considerable value to a certain demographic. Wolfram Computer Services is closely associated with heavy duty math. Pavel Durov awaits trial in France on more than a dozen charges of untoward online activities. Stephen Wolfram collects awards and gives enthusiastic if often incomprehensible talks on esoteric subjects.
But the technology path is similar in my opinion. Both of these organizations want to use available compute resources; they are not too keen on buying GPUs, building data centers, and spending time in meetings about real estate.
The cost of running a job on the Supercompute system depends on a number of factors. A user buys “credits” and pays for a job with those. No specific pricing details are available to me at this time: 0800 US Eastern on December 6, 2025.
Net net: Two very intelligent people — Stephen Wolfram and Pavel Durov — seem to think that the folks with giant data centers will want to earn some money. Messrs. Wolfram and Durov are resellers of excess computing capacity. Will Amazon, Google, Microsoft, et al be signing up if the AI demand does not meet the somewhat robust expectations of big AI tech companies?
Stephen E Arnold, December 19, 2025
Comments
Got something to say?
-
- Subscribe to Beyond Search
Feature archive
News archive
-
