Short term they might have less compute, but long term they will most definitely have more compute. They don't have to worry about energy, they don't have to worry about people blocking them building data centers the only thing stopping them is there no Chinese manufacturer that can produce a chip as good as Nvidia but I would bet that solved in a year or so.
They don't have to worry about people blocking DC construction in the US either. All new AI datacenters are designated "dual-use" so the federal government is already letting local councils know to fuck off.
they can also put them in better places, vs trying to arbitrage expensive electricity costs and various US corruption thats built more around paying people off than getting things done
Yeah, but that brings up another problem that China doesn't have. Their federal government is always on the same page. The US federal government changes every two years and it doesn't seem like one side is going to be allowing building of AI data centers anymore and in fact might just ban AI in general.
It's hard enough to build them in the US that multiple companies are unironically spending 10s of millions of dollars to try to build them in space
Which cynics could say is marketing hype, but I tend to believe it's extremely difficult to build anything land/energy intensive in the modern United States
There are places that aren’t in the US that also aren’t in space. I still don’t get why it makes any sense to put this stuff in space, but if it does, it’s not the political difficulty of building stuff on earth. So it’s hard in the US. You’re telling me Mexico wouldn’t allow it? Argentina? Mongolia? Build a data center on a French nuclear test site in the Algerian desert and it’s still easier than space.
I'm actually the most innovative, I've written thousands of papers advancing the state of human knowledge. They're just in my basement and I don't show anyone.
if it's agentic stuff, they likely aren't hammering a core constantly and they will maybe sit idle quite often between model requests, so it makes sense. I do wonder how much memory they allocate to each one though.
it's just very efficient use of shared cores that is required to make these kinds of workloads cost efficient
A place I worked back around 2020 was running a Grafana instance per customer that got embedded on the web dashboard. We had 110 pods per GKE (kubernetes on gcp) 4 CPU node because that was a network imposed pod limit at the time. The nodes were usually idle--could have shoved a lot more on if not for the IP limit.
I think around that time Grafana changed their license tho so you couldn't host OSS Grafana as part of your service.
Even without that particularly special case, high-energy physics collaborations have long since broken the idea of authors. There are now many collaborations with hundreds of authors publishing regularly and quite a few that are into the thousands.
After parsing that article to get the paper link, that is insane. I am not in the field, but find it curious that each every person of that list of authors had direct significant contribution. Not my field though so just seems odd.
The LHC is a massive experiment. Just to cite Wikipedia:
> The Large Hadron Collider (LHC) is the world's largest and highest-energy particle accelerator. It was built by the European Organization for Nuclear Research (CERN) between 1998 and 2008, in collaboration with over 10,000 scientists, and hundreds of universities and laboratories across more than 100 countries. It lies in a tunnel 27 kilometres (17 mi) in circumference and as deep as 175 metres (574 ft) beneath the France–Switzerland border near Geneva.
They give credit to everyone who built the instrument, wrote the software pipelines to analyze the data, etc. That's thousands of people. That's how large experimental projects work.
I don't understand why half the comments here are about the author list, this is very common practice in e.g. large-scale physics experiments and biology, and every new GPT release from OpenAI equally had papers with tons of authors
Why are people so surprised by this? This is pretty common among all areas of science right now. This particular topic would have required a lot of hands to get it going.
It seems every DeepSeek paper/patent has a huge number of authors, and this one is no exception. They couldn't even fit everyone on the page, there are 31 others not shown. This could be an asset protection strategy (i.e., human assets). Imagine if there were only 3 authors. Those authors may get hired away by competitors. If you list every employee on every paper then competitors don't know who to lure away.
Corresponding author is the head of the group, usually the one who coordinates the study, the one who knows everyone else. Is the "if you have any question about anything in this paper, contact me" person, even years after publication.
You need to tighten up your tinfoil hat. When HN features all the discussions on OpenAI's shenanigans on Navier-Stokes, you weren't seeing claims this was anti-US.
I don't know the original intentions, but I don't feel it's anti Chinese just by itself.
It could be pro-Chinese if you frame it like they give merit to anyone involved, or that they take care of their most valuable employees (maybe they pay them another way), or whatever.
Or just observant of how Chinese labs publish their stuff. (Not sure if it's always like that I haven't take a look at the number of authors in this type of publications. Or at least I can't remember it.)
Because if they get you to dismiss or not care about what nonAmerican AI companies are doing maybe people won't see that they are innovating in public. So they can continue smear campaigns that China is simply stealing everything and that's all they do. So the average person will be relieved when the govt moves to regulate away competition from us AI companies and labs. Horray
> Why is this the top comment? Many of the comments, as well as this one, have no relation to content and only mention a triviality
This is hardly a triviality. If you are interested in a topic and you find a paper interesting, more often than not you are interested in reaching out to the author.
If a paper features a list of authors that outnumber the paper's pages 10-to-1, it's a major red flag, and it's quite plausible and expectable that 99%of the names in that list had zero input and might even have zero contribution to give to the topic. I'd even argue it's a kin to academic fraud.
By the way, the same goes for those researchers who work on paper mills, and manage to rake in production metrics that go well beyond an article per day.
If you're confused about who to contact to follow up on the ideas of a particular paper then write to the first author, or whichever author is listed as the contact person. This is not hard and is an orthogonal consideration to the length of the author list.
> If you're confused about who to contact to follow up on the ideas of a particular paper then write to the first author, or whichever author is listed as the contact person.
That just goes to show how much you understand the topic. The leading author is usually the institutional leader, and in paper mills it's frequently the guy who stands to benefit the most from the scam. Not the guy who actually did the work.
So now that you missed that one option, what's your plan? To walkt through the hundred names in the list?
I miss the YOLO (CNN computer vision model) days where one dude can publish a paper, completely disregard academic conventions, and yet push the field forward by leap and bounds.
Yea those days were cool. Shame he quit doing research once he realized the biggest application for his work was killing people with robots I guess. It was a nice period of time. Before everyone in industry realized the cool stuff they were building was likely only going to be used to surveil the public, target people, remove skilled labor, and probably develop new weapons (whether people call them that now or not isn't relevant). Good old days.
Or without the cynical lens they just operate like CERN where they acknowledge everyone who without them could not complete the research. Hyperauthorship is common in particle physics for example.
I wonder if they are signalling that if they can do this for training, then they can create an style agent swarm to hack anyone with 380k concurrent agents.
>"Within one scale unit, the platform spans nearly 160 CPU nodes with 30K cores and ∼250 TB of DRAM. It manages petabytes of layers and images. On a typical day, a single scale unit serves about 3 M sandbox instances, with peak concurrency reaching ∼380K and a creation rate exceeding 5,000 instances per second."
Impressive numbers!
Whoever would have thought (in prior years) that in 2026 AI Agents (not people or corporations) seem to be (or seem to be rapidly becoming) the biggest consumers of cloud computing resources...
12 sandboxes per code is insane, I wonder how many of these sandboxes are idle at a time. Depending on the tasks assigned the resource requirements are different. Compare an agent doing pdf conversion and one responding to a simple question. One is cpu bound the other is mostly network wait.
This is an interesting problem from infra perspective since you cannot predict the workload. On a bigger scale you may get away with forecasts.
Im waiting for tech that elastically allocates cpu/mem without restarting a container.
As models get better, safe-and-secure sandboxes/environments are going to be the way forward. With the recent rise in cases where models can somehow gain access to the internet and blast past the sandbox, it's very important to have all the resources contained within the sandbox with no access to the outside world.
Ax/agent substrate is focused on enterprise inference. This is squarely targeting the training and reinforcement learning patterns (it's not to say these technologies don't overlap, but perhaps closer to kube with knative & kata container support (and some level of session persistence across all 3). Or nomad... Or any orchestrator that supports multiple underlying isolation techniques.
I can't work out from the paper if the session can cross all 3 types of isolation (between persistence) but I'm assuming it can if it needs to evolve tool calling and security requirements.
In any case it feels like they're just composing and mangling technologies into a platform and not bringing much to the table (from what I've read so far; ie still using well established isolation tech and not something unexpected like a microkernel) but the outcomes are impressive
89 comments
[ 4.2 ms ] story [ 119 ms ] threadWhich cynics could say is marketing hype, but I tend to believe it's extremely difficult to build anything land/energy intensive in the modern United States
Not sure they'd be the same without the constraints.
it's just very efficient use of shared cores that is required to make these kinds of workloads cost efficient
I think around that time Grafana changed their license tho so you couldn't host OSS Grafana as part of your service.
[0] https://arxiv.org/abs/1207.7214
> The Large Hadron Collider (LHC) is the world's largest and highest-energy particle accelerator. It was built by the European Organization for Nuclear Research (CERN) between 1998 and 2008, in collaboration with over 10,000 scientists, and hundreds of universities and laboratories across more than 100 countries. It lies in a tunnel 27 kilometres (17 mi) in circumference and as deep as 175 metres (574 ft) beneath the France–Switzerland border near Geneva.
They give credit to everyone who built the instrument, wrote the software pipelines to analyze the data, etc. That's thousands of people. That's how large experimental projects work.
[1] https://cds.cern.ch/record/1471031
[2] https://impact.ornl.gov/en/publications/observation-of-a-new...
Or they are just rushing to say anything, and it's much easier to comment on that than the content of the paper.
I wonder how those few that were left out feel :)
all authors are listed. there is no conspiracy to hide authorship.
arxiv simply want to keep the page short not too long.
just click the link and it will show the others, this seems to be a limitation/UI feature of arxiv. The paper itself contains the full list
You need to tighten up your tinfoil hat. When HN features all the discussions on OpenAI's shenanigans on Navier-Stokes, you weren't seeing claims this was anti-US.
It could be pro-Chinese if you frame it like they give merit to anyone involved, or that they take care of their most valuable employees (maybe they pay them another way), or whatever.
Or just observant of how Chinese labs publish their stuff. (Not sure if it's always like that I haven't take a look at the number of authors in this type of publications. Or at least I can't remember it.)
This is hardly a triviality. If you are interested in a topic and you find a paper interesting, more often than not you are interested in reaching out to the author.
If a paper features a list of authors that outnumber the paper's pages 10-to-1, it's a major red flag, and it's quite plausible and expectable that 99%of the names in that list had zero input and might even have zero contribution to give to the topic. I'd even argue it's a kin to academic fraud.
By the way, the same goes for those researchers who work on paper mills, and manage to rake in production metrics that go well beyond an article per day.
Which itself not necessarily a sign of fraud: https://www.science.org/content/article/physics-paper-sets-r...
That just goes to show how much you understand the topic. The leading author is usually the institutional leader, and in paper mills it's frequently the guy who stands to benefit the most from the scam. Not the guy who actually did the work.
So now that you missed that one option, what's your plan? To walkt through the hundred names in the list?
christ
Impressive numbers!
Whoever would have thought (in prior years) that in 2026 AI Agents (not people or corporations) seem to be (or seem to be rapidly becoming) the biggest consumers of cloud computing resources...
This is an interesting problem from infra perspective since you cannot predict the workload. On a bigger scale you may get away with forecasts.
Im waiting for tech that elastically allocates cpu/mem without restarting a container.
DSec is a good step in this direction!
I can't work out from the paper if the session can cross all 3 types of isolation (between persistence) but I'm assuming it can if it needs to evolve tool calling and security requirements.
In any case it feels like they're just composing and mangling technologies into a platform and not bringing much to the table (from what I've read so far; ie still using well established isolation tech and not something unexpected like a microkernel) but the outcomes are impressive