kilotaras 2 hours ago

Alibaba Cloud claims to reduce Nvidia GPU used for serving unpopular models by 82% (emphasis mine)

> 17.7 per cent of GPUs allocated to serve only 1.35 per cent of requests in Alibaba Cloud’s marketplace, the researchers found

Instead of 1192 GPUs they now use 213 for serving those requests.

  • yorwba 2 hours ago

    Not really, Figure 1(a) of the paper says that the 17.7% are relative to a total of 30k GPUs (i.e. 5310 GPUs for handling those 1.35% of requests) and the reduction is measured in a smaller beta deployment with only 47 different models (vs. the 733 "cold" models overall.) Naïve extrapolation by model count suggests they would need 3321 GPUs to serve all cold models, a 37.5% reduction to before. (Or 6.6% reduction of the full 30k-GPU cluster.)

djoldman 3 hours ago

Key paragraph:

> However, a small handful of models such as Alibaba’s Qwen and DeepSeek are most popular for inference, with most other models only sporadically called upon. This leads to resource inefficiency, with 17.7 per cent of GPUs allocated to serve only 1.35 per cent of requests in Alibaba Cloud’s marketplace, the researchers found.

checker659 21 minutes ago

They are working with tiny models. Not sure how well it'd scale to bigger models (if at all).

  • CaptainOfCoit 9 minutes ago

    They're all LLMs, so no, not tiny, but not exactly huge either:

    > Our current deployment runs in a cross-region cluster comprising 213 H20 GPUs, serving twenty-eight 1.8–7B models (TP=1) and nineteen 32–72B models (TP=4).

hunglee2 4 hours ago

The US attempt to slow down China's technological development succeeds on the basis of preventing China from directly following the same path, but may backfire in the sense it forces innovation by China in a different direction. The overall outcome for us all may be increase efficiency as a result of this forced innovation, especially if Chinese companies continue to open source their advances, so we may in the end have reason to thank the US for their civilisational gate keeping

  • dlisboa 3 hours ago

    History has shown that withholding technology from China does not significantly stop them and they'll achieve it (or better) in a small number of years.

    In many senses there's hubris in the western* view of China accomplishments: most of what western companies have created has had significant contribution by Chinese scientists or manufacturing, without which those companies would have nothing. If you look at the names of AI researchers there's a strong pattern even if some are currently plying their trade in the west.

    ---

    * I hate the term "western" because some "westeners" use it to separated what they think are "civilized" from "uncivilized", hence for them LATAM is not "western" even though everything about LATAM countries is western.

    • raincole 32 minutes ago

      > look at the names

      Why would I do that tho? If we look at the names of scientists/researchers/engineers/businessmen, the conclusion would be that the US has contributed nothing to the world. Europeans did all the hard work!

      • thesmtsolver 11 minutes ago

        Another equivalent way to look at that:

        Historically, top scientists/researchers/engineers/businessmen migrate from rest of the world to the US rather than to Europe or China.

        Imagine if Europe or China were a bit more open with immigration and equally attractive, we would see the same pattern there too.

    • zawaideh 3 hours ago

      Re: Western. A similar thing plays out when the term "international community" is used in news. It refers to the US and its major allies which means US, Canada, Western Europe, Japan, Australia and New Zealand more or less.

      • nicoburns 10 minutes ago

        > A similar thing plays out when the term "international community" is used in news. It refers to the US and its major allies which means US, Canada, Western Europe, Japan, Australia and New Zealand more or less.

        Wait, really? I thought "international community" meant all countries.

      • newyankee 2 hours ago

        Essentially countries that were developed prior to 1990 or so , although South Korea is a tricky case today going by this definition, as are Taiwan, Hongkong and Singapore

      • tsunamifury 2 hours ago

        Yes community refers to whose who participate in community.

        How is this hard to understand?

        Broadly speaking coast de ivory and the like is not a participant in the international community.

    • achierius 3 hours ago

      > most of what western companies have created has had significant contribution by Chinese scientists or manufacturing, without which those companies would have nothing. If you look at the names of AI researchers there's a strong pattern even if some are currently plying their trade in the west.

      While I don't disagree with your overall point, it's important to recognize that this is only a phenomenon of the last ~30 years, and to avoid falling into the trapn of Han racial chauvinism. E.g. there were ~no Chinese scientists in Germany in the 70s but they were heavily innovating nevertheless.

      • dlisboa 3 hours ago

        Absolutely. China obviously has a longer history with innovation but they like to make it seem everything was invented by them at some point in the past. I'd say newer technology is where China has had a bigger impact.

        Consequently newer tech is precisely where global cooperation is most required so no country can really do it by themselves. We could even say no country, western or otherwise, has been doing it on their own for the past 500 years or so but alas...

        • tsunamifury 2 hours ago

          It’s more helpful to think of China as an accelerant rather than an innovator in this position.

    • onlyrealcuzzo 3 hours ago

      > History has shown that withholding technology from China does not significantly stop them and they'll achieve it (or better) in a small number of years.

      It's worked for a very long time for aircraft.

      China has been pushing to build its own aircraft for >23 years. It took 14 years for COMAC to get its first regional jet flying commercial flights on a Chinese airline, and 21 years to get a narrow-body plane flying a commercial flight on a Chinese airline.

      If for no technical reasons and purely political, COMAC may still be decades away from being able to fly to most of the world.

      Likewise, in ~5 years, China may be able to build Chips that are as good as Nvidia after Nvidia's 90% profit margin - i.e. they are 1/10th as good for the price - but since they can buy them for cost - they're they same price for performance and good enough.

      If for purely political reasons, China may never be able to export these chips to most of the world - which limits their scale - which makes it harder to make them cost effective compared to Western chips.

      • wood_spirit 13 minutes ago

        But at the same time they are fielding multiple new stealth aircraft and their jets and missiles outperformed western aircraft in the recent Pakistan India flare-up.

        • ambicapter 6 minutes ago

          So you'd think they'd be able to build a commercial jet liner, no?

      • Yoric 2 hours ago

        > If for purely political reasons, China may never be able to export these chips to most of the world - which limits their scale - which makes it harder to make them cost effective compared to Western chips.

        Note that this happens at the same time the US is breaking up its own alliances, so as of this writing, there's no such thing as certainty about politics.

      • sofixa 2 hours ago

        > China has been pushing to build its own aircraft for >23 years. It took 14 years for COMAC to get its first regional jet flying commercial flights on a Chinese airline, and 21 years to get a narrow-body plane flying a commercial flight on a Chinese airline

        And both those planes have a strong dependency on "western" components that won't be overcome before the 2030s, and even then, they're around a generation behind.

    • caycep 2 hours ago

      this is true for anyone - create challenges, and you optimize efficiency elsewhere.

      Also, isn't this the usual path to better computer science? Reducing computation needs by making better/more efficient algorithms? The whole "trillions of dollars of brute force GPU strength" proposed by Altman, Nadella, Musk et al just seems to reinforce that these are business people at heart, not engineers/computer scientists...

    • rayiner 3 hours ago

      Nobody thinks the Japanese aren’t “civilized.” “Western” is just a euphemism for “rich and orderly.”

      • switchbak a minute ago

        It is an odd category, and Japan is often considered to be "Western" - these days at least. That certainly wasn't the case even a few generations ago.

        I think it's ostensibly supposed to be more about shared cultural values, but even that is a pretty weak way to divide countries. Perhaps "an ally of the United States" is a little more accurate?

        Any societal dividing line like this is bound to hit on problems once subjected to the real world.

      • bad_haircut72 2 hours ago

        Its more about democracy and adhering to the global (set up by America post WW2) system of laws and trade.

        • rayiner 2 hours ago

          I think most people considered Spain a "western" country even in 1970 when it was controlled by Franco.

    • notepad0x90 3 hours ago

      western is a cultural term derived from a geographic one. The US is also not 'western' strictly geographically as it is not in western europe, neither is australia. But they both originated from Britain's empire and share in it's cultural ancestry. It means "western europe and it's cultural derivatives". Spain and Portugal's empire fell away long before britain and france's and they don't have similar geopolitical relations like NATO, so it's hard to consider their former colonies/upstarts part of the same sphere of cultural influence.

      China for sure will catch up, the question is what they will do with it. They're not ambitious like the US/West. The US wanted influence all over the world as an extension of the cold war and to keep economic interests safeguarded. But China just doesn't operate that way. They're more hands-off. They could be opening up alibaba cloud datacenters all over the US, offering it as an AWS/Azure alternative, funding tons of startups all over europe, the US,etc... to exert their influence, but they won't. They have a more long-term low-and-slow approach to global domination. The "100 year marathon" as they called it, which they'll win for sure.

      China's greatest weakness is not just their lack of ambition,but their command-economy. They're doing capitalism but with central control of the economy. It intertwines government policy with corporate policy, making it harder to do business overseas (like with bytedance/tiktok).

      • tsunamifury 2 hours ago

        False.

        Westernism is broadly an extension of the academic notion of classicism, starting in Egypt and then Greece Rome and into Europe and the Americas.

    • tsunamifury 2 hours ago

      It’s helpful to think of westernism as a platonic ideal. Individually derived reason and virtue, superior to state and sometimes ‘gods’ as a tradition to drive up the total survivability, richness, and stability of the community.

      Concepts that enable the individual should empower a chosen configuration of society not the other way around.

      Contrast this with non westernism where either education of the individual is not valued or the state is the primary goal over the individual.

      I’ve worked with states governments and individuals around the world for 20 years and find this very useful definition. What’s confusing is the nations who have half adopted westernism but don’t fully due to either caste systems or government dominated thinking.

      It’s an arrow towards rationalism over tradition, individualism over collectivism, flatness over hierarchy, and future over past. But only the limit of the resources any given society has.

    • nextworddev 33 minutes ago

      Name one thing China has invented first in LLMs that the “west” adopted as a standard

      • nextworddev 3 minutes ago

        Your silence is deafening, qwen bots

    • chuckadams 2 hours ago

      I find "western" is often used to disparage "western thought", as in it can't grasp the deep wisdom of those mysterious orientals that transcends normal logic and reason. Declaring such a split is the underpinning of a whole lot of woo-woo beliefs.

      • hbarka 9 minutes ago

        Disparage or exalt? It can also be used in an objective sense without conjuring insult.

  • sspiff 7 minutes ago

    Their are signs that China is not open sourcing their SOTA models anymore. Both Huawei and Qwen (Qwen-Max, WAN 2.5) and have launched flagship models which are yet to be opensourced.

    • camel_Snake a minute ago

      Small counterpoint but there are also 2 new players putting out SOTA open source models (Moonshots Kimi and zhipus GLM) so we're still seeing the same number of models overall, just via newer entrants.

  • notepad0x90 3 hours ago

    I think anti-immigrant rhetoric will have the most impact against the US. A lot of the people innovating on this stuff are being maligned and leaving in droves.

    Aside from geography, attracting talent from all over the world is the one edge the US has a nation over countries like China. But now the US is trying to be xenophobic like China, restrict tech import/export like China but compete against 10x population and lack of similar levels of internal strife and fissures.

    The world, even Europe is looking for a new country to take on a leader/superpower role. China isn't there yet, but it might get there in a few years after their next-gen fighter jets and catching up to ASML.

    But, China's greatest weakness is their lack of ambition and focus on regional matters like Taiwan and south china sea, instead of winning over western europe and india.

    • dlisboa 3 hours ago

      > But, China's greatest weakness is their lack of ambition and focus on regional matters like Taiwan and south china sea, instead of winning over western europe and india.

      That's a strength. Them not having interest in global domination and regime change other than their backyard is what allows them to easily make partners in Africa and LATAM, the most important regions for raw materials.

      • notepad0x90 2 hours ago

        You would think so, but historically that's why they never became more than a regional power. Empires for millennia craved trade with China but only the mongols from that region made it all the way to western europe in their invasions.

        It is a strength, if their goal is to have a stable and prosperous country long term, and that seems to be what they want. good for them. But nature abhors a vacuum, so there will always be an empire at the top of the food chain. Such empires want to maximize wealth for their people and secure them against threats, that's why invasions and exploitation of weaker countries happens. That game hasn't changed. Friendly relations work, until you need a lot of resources from a country that doesn't want to give it up. Or, like with the US, when they're opening up military bases next to your borders and you need a buffer state. Or, when naval blockades and sanctions are being enforced against your country for not complying with extra-sovereign demands.

        History shows that countries content with what they have collapse or weaken very quickly.

        China will have a population crisis in a few decades for example, and it won't have the large manufacturing base and its people will be too used to luxuries to go back to slaving for western countries for pennies. Keep in mind that the current china itself is so great and prosperous because of all the invasions it did against western china and satellite states like Vietnam and north Korea (the US isn't special in this regard).

        • lossolo 2 hours ago

          > But nature abhors a vacuum, so there will always be an empire at the top of the food chain

          The world has been bipolar and multipolar before in history, and it can be again. The unipolar period of American dominance is ending.

      • ikidd an hour ago

        >Them not having interest in global domination and regime change

        I don't even know where to begin with that one.

      • OrvalWintermute 2 hours ago

        if you've been tracking the shark deals they give countries for loans, I think you'd recant what you just said.

        "while the CCP accuses the West of predatory interest rates, the average Chinese rescue loan carries an interest rate of about 5 percent, more than double the IMF’s standard 2 percent. As of Oct. 1, 2025, despite higher U.S. interest rates, the IMF’s Special Drawing Rights lending rate stands at only 3.41 percent, still significantly lower than what China charges struggling nations for so-called relief."

        These countries paying these loans are the ones least able to pay them back, and at more than double IMF loans, they are really putting them in a vise.

    • bkandel 2 hours ago

      China's greatest weakness is that their working-age population has already peaked and is in the process of plummeting, which will continue over the coming decades.

      • notepad0x90 2 hours ago

        Yes, and being content and lacking ambition isn't good. Expansionism and immigration can solve that, but they're culturally stagnant in that regard.

        Without immigration, the US would have faced the same problems.

    • rayiner an hour ago

      > But now the US is trying to … compete against 10x population and lack of similar levels of internal strife and fissures.

      I can’t tell whether you think the anti-immigration stance is a good thing or bad thing.

    • csomar 2 hours ago

      > But, China's greatest weakness is their lack of ambition and focus on regional matters like Taiwan and south china sea, instead of winning over western europe and india.

      How can they have international hegemony before they clear their regional order? China is more interested in aligning Taiwan than invading; though it’ll probably invade if it can’t align it diplomatically.

      China is probably not interested in continuing the current Western-style order but to implement their own sino-stuff. At least with the CCP at the helm.

    • drittel 2 hours ago

      [flagged]

      • ksynwa 2 hours ago

        Wow

        • dotnet00 2 hours ago

          Unsurprising that someone with such beliefs would be too much of a coward to use their real account.

          • drittel 2 hours ago

            This is my only account. I'm very proud of my beliefs.

        • drittel 2 hours ago

          Extremely low T reply

  • lesuorac 35 minutes ago

    The US isn't slowing China anymore.

    China has an import ban on chips [1] so its irrelevant what the US does.

    [1]: https://www.cnbc.com/2025/09/17/nvidia-ceo-disappointed-afte...

    • overfeed 14 minutes ago

      > China has an import ban on chips

      Only in response to the US bamming the export of cards China wamted. The import ban is the Chinese government burning the the landing ships, it clearly communicates to everyone involved that there is no going back.

    • xadhominemx 18 minutes ago

      The US is certainly slowing down China considerably. China would certainly not have an import ban on Blackwell GPUs if they were made available. And upstream, the ban on EUV and other high end semiconductor production equipment has severely limited china’s capacity to produce logic and DRAM (including HBM).

    • unethical_ban 22 minutes ago

      Would they have done that if the US had been more "reliable" in providing the chips and didn't cut them off in the first place?

      The point still stands that the US instigated the split.

  • segmondy 4 hours ago

    may backfire? it's a bit too late for that.

    go to 2024, western labs were crushing it.

    it's now 2025, and from china, we have deepseek, qwen, kimi, glm, ernie and many more capable models keeping up with western labs. there are actually now more chinese labs releasing sota models than western labs.

    • hunglee2 3 hours ago

      too early to call a winner, though it is disappointing to see US withdrawal from open source. Still the main outcome of open source is distribution / diffusion of the idea, so it will inevitably mean US open source will come back, hopefully via some grass roots maniac, there will be a Linus-like character emerge at some point

      • mixologist 2 hours ago

        user growth has slowed. the technology that should help users is only being pushed from the top, while users refuse to use it. openai pivoted to porn.

        does it really feel like they have a chance to recover all the expenses in the future?

        crypto grifters pivoted to ai and, same as last time, normal people don’t want to have anything to do with them.

        considering the amount of money burned on this garbage, i think we can at least declare a looser.

      • segmondy 3 hours ago

        i'm not calling a winner, i'm just saying that the chinese have caught up despite the embargo. google, openai & anthrophic have phenomenal models. i stopped using openai & anthropic after they called for open weight/source regulation. i use google because they offer gemma and i got a year gemini-pro subscription for free, use openai gpt-oss-120b since i can run it at home, and the only model i currently pay for is a chinese model.

  • rzerowan 4 hours ago

    Fingers crossed for convergence rather than divergence in the technical standards.Although the way hings are going it looks like the 2 stacks will diverge sooner rather than later , with the US+ banning the use of CHN models while simultaneosly banning the export of it quasi-open models. We may very well end up in a situation like the old PAL vs NTSC video standard where the PAL(EU/Asia/AFrica) and NTSC(America's/Japan) gradually converged with the adoption of digital formats. Instead here would be a divergence based on geopolitical considerations.

    • hunglee2 3 hours ago

      positive take: a bifurcated tech tree might give us (humanity) a better chance of faster advancement, as it would be a persistent A/B test in live environment. Where I would join you in the crossing of fingers is to ensure such A/B testing is competitive but not destructive. We may even evolve to a situation of complementarity, an American Ying vs the Chinese Yang. Lets hope so!

  • reliabilityguy 4 hours ago

    Tbh this whole situation reminds of how Japan excelled in making a lot more with a lot less after WW2, e.g., fuel-efficient engines, light cars, etc. these constraints were not present in the US (and to some extent in Europe), and resulted in US cars being completely not competitive in non-US markets.

    • dataviz1000 4 hours ago

      I've been in Chile, Peru, Colombia, Panama, and Costa Rica.

      The streets are flooded with cheap Chinese cars and I see more BYD than American cars. If the car wasn't made in Japan or Korea which probably account for most of the cars, it was likely made in China. Moreover, I haven't been in countries with the closest ties to China.

      • sofixa 2 hours ago

        > The streets are flooded with cheap Chinese cars and I see more BYD than American cars

        This isn't surprising in any way, American "cars" (quotes because the vast majority of what American manufacturers pump out isn't cars, it's trucks) haven't been competitive in decades. The only globally competitive vehicles were developed in Europe by GM Europe (Opel, since sold to PSA now Stellantis) or Ford Europe (which axed all models bar the Puma). The rest is too big, expensive and inefficient from the vast majority of uses. Tariffs and good marketing keep American car manufacturers in business in the US, but those don't work in most other markets.

        The more appropriate comparison is with European automakers such as VW Group, Stellantis (Peugeot, Citroën, DS, Fiat, Chrysler, Dodge, Ram), Renault. And there too BYD is winning as well in mosy countries, but at least there's a comparison possible.

    • tsunamifury 2 hours ago

      The premature optimizer is never the innovator.

      Japan eventually stopped that role and their products improved greatly.

  • knowitnone3 11 minutes ago

    It's much easier to copy what others are doing instead of spending the time and money for research and engineering. It's also much easier if you steal the tech. I could never have invented a bicycle but I can sure make a copy of one.

  • archerx an hour ago

    I want China to release GPUs with a ton of VRAM, 128gb - 256gb. It doesn’t matter if they are half as fast as Nvidia because having a big model at a reasonable speed is better than not being to run them at all. AMD could have done this and have had a massive impact on nvidia’s market share but they choose not to because reasons.

  • amelius 2 hours ago

    Another outcome may be that we now have to learn Chinese to understand their datasheets ...

    • anonzzzies 16 minutes ago

      I was doing this in the 70-80s with electronics from Hong Kong and Japan. The nice cheap stuff ( I was very young ) was all sheets in things I basically had to pattern match against notes of others on BBS and meetups.

  • IT4MD 2 hours ago

    I believe this is an Pollyanna take on AI. There is nothing about humans that tells us humans will bring AI to fruition for the other humans and a mountain of evidence showing how it will be used to abuse humans instead....for profits/power/whatever horse shit the masters of the universe have decided upon.

  • myth_drannon 2 hours ago

    China's innovation relies on the stolen western IP, without it, China is nothing. Also USSR/Russia is no longer a scientific powerhouse that can supply China with some military innovation. A dictatorship combined with cheap labour it 100% guarantees that the country's innovation is stunted, no matter what the Chinese propaganda claims.

  • belter 2 hours ago

    China is a nation of engineers...The US has been relying in on H-1B immigrants. Science is under attack. The truth is the US already lost: https://youtu.be/whVlI6H4d-4

braza 3 hours ago

Does someone know if there's some equivalent of those engineering/research blogs for Chinese companies?

I used to follow the ones from Western companies, but honestly, after some point in time, I would like to see some cases from what I consider is a good benchmark for everyone that does not work in FAANG in terms of engineering.

  • supriyo-biswas 13 minutes ago

    The company blogs of Chinese companies will often do articles like this[1] talking about a new innovation or optimization that they did, but this will be often just mixed in with marketing articles too.

    I would also assume there's a lot of content in the native Chinese forums, which unfortunately, as an English-speaking person, I wouldn't be able to easily refer to :(

    [1] https://www.alibabacloud.com/blog/how-does-alibaba-ensure-th...

throwaway48476 2 hours ago

Its easy enough for a a well resourced entity to take a pre trained model and deploy it on new hardware to save on the NVDA tax. It's far less likely for research and model training to happen outside the mature NVDA ecosystem.

muddi900 2 hours ago

[flagged]

  • dotnet00 2 hours ago

    This is such a popular coping tactic from Americans when it comes to facing actual competition, especially from China. Everything they do must either be a lie or just stolen American technology, as if there's something inherently special about Americans that no one else has.

    • muddi900 an hour ago

      That is a great respomse to something I did not say.

    • throwacct an hour ago

      Interesting. So, we're going to deny that most of the IP theft from China up to this moment? Do you even think China is this advanced just because of chinese innovation? C'mon man.

  • larus_canus 2 hours ago

    Ah, yes, the American media environment, which is internationally famous for not lying.

    • muddi900 an hour ago

      So we are in agreement, we should not take every thing at face value.

      • larus_canus 42 minutes ago

        I just hope you extend this skepticism consistently.

  • throawayonthe 2 hours ago

    scmp is kinda the opposite of state media lol