Talk:Supercomputing in China

Suggestion for future expansion and update

It would be good if those who read Chinese fluently and write English well could expand this article from the viewpoint of the relationship to the PHPC100 personal supercomputer. While Tianhe-1 is a fast machine, unlike most other supercomputers, it did not have a specific application in mind, and its main goal was development of the knowledge for building a really fast machine. That knowledge is likely to work its way down to new models of the PHPC100 type computer, pretty much the way knowledge gained through Formula One racing eventually works its way into passenger cars. So it would be good to set an example where information from Chinese sources gets updated into English Wikipedia over time. Much of the information on the PHPC100 is in Chinese, so a section on that in this article will be a good start, for this has the potential of changing the game for "personal supercomputing" where the impact will be far greater than the top levels of supercomputing. History2007 (talk) 16:47, 10 November 2010 (UTC)

See edits. Problem is understanding the technical content... In ictu oculi (talk) 05:17, 26 March 2011 (UTC)

WP:BRD and Relevance of Item

History2007, I don't understand why irrelevant information should be presented in this article - I have never seen Non-US top position super computers ever been given such treatment as to put future speculative "China beating" super computer in the article. I can only speculate its sour grape mentality because currently China sits on No.1 Position. If this is the case, then its perfectly okay to put any future speculative "American beating" super computers on the IBM Sequoia page as well ?? As I see it, those future speculative 'ranking' or super computer should be left out of the article. THIS IS WIKIPEDIA. We are here for the FACTS, not some speculation from the future. Spunking (talk) 14:20, 30 December 2010 (UTC)

I am sorry Spunking but I can not agree with you. As for the use of the term "sour grape mentality" that is not only speculative but also incorrect in my view. I see the information as relevant, given that the field is inherently fluid and there is no dispute about the correctess of the information. Your statement that it is "irrelevant" is your opinion and since we do not agree, your obvious remedy is to ask for a WP:Third opinion to obtain an independent view regarding its relevance - and I will hereby agree to respect said 3rd opinion. In the meantime, please respect the WP:BRD process and avoid a revert cycle. When two users do not agree, the remedy is a 3rd opinion, not a revert cycle. Please avoid reverts since that is the WP:BRD process. Thanks. History2007 (talk) 18:01, 30 December 2010 (UTC)

3rd Opinion

Viewpoint by History2007

The dispute is about this deletion. The material is called irrelevant. Some of teh material was carefully added by other users, not myself, and gives a perspective. The factual accuracy of the information in question is not being disputed, and I am seeking an opinion regarding relevance. I see "no injury" done to Wikipedia readers by knowing about the new supercomputer, and given that it relates to high performance systems, I see it as relevant, and factually valid. The reader will just get a better perspective by having this information. History2007 (talk) 22:18, 30 December 2010 (UTC)

Response to third opinion request (Disagreement about relevance of final sentence in the article):

I'm assuming your dispute is over the last sentence in the article, namely: "It should also be noted that the "top spot" in supercomputing is not a permanent ranking and continues to change, for instance, the IBM Sequoia supercomputer scheduled for 2011 has an anticipated speed of 20 petaflops, well above current supercomputer levels." This sentence is sourced to an article in the Guardian. While this sentence is accurate, I also do not see the relevance of the statement in the context of this article. Firstly, the fact that the "top spot" in supercomputing is not a permanent ranking is exceedingly obvious; I think this is something we can safely assume the reader already knows. The fact that there are top 100 and top 500 lists implies that this is a rapidly changing field. Secondly, this is supposed to be an article on supercomputing in China, so we should only be mentioning supercomputers in other countries if there is a good reason to (i.e. if one of the Chinese supercomputers was based on a particular supercomputer in another country, or inspired by it, or was noted for competing with it, etc). Just mentioning a supercomputer out of the blue (which doesn't exist yet, bordering on WP:CRYSTAL) doesn't seem relevant to me. The article needs quite a bit of cleanup work, and I don't see this single sentence being a serious issue with the article (certainly not worth edit warring over), and whether it stays, gets deleted, or gets modified won't make much of a difference.—SnottyWong soliloquize 22:30, 30 December 2010 (UTC)

Ok, but what about this paragraph:

Although China has the fastest supercomputer in the world and the second most number of supercomputer in the TOP500, the US still has far more supercomputers in the TOP500 overall. As of November 2010, there were 42 Chinese supercomputers, compared to 275 supercomputers in the United States among the top 500. Apart from China and the United States, Japan, France and Germany are the other countries with a supercomputer ranked in the top 10 positions.

Please clarify that too. My suggestion: Please edit these two paragraphs as you see fit and let us be done. Thanks. History2007 (talk) 22:37, 30 December 2010 (UTC)

I think that paragraph is a bit more relevant, as it is a direct comparison of China to other countries. I think it could be stated in a better way though. I'll see if I can come up with something. SnottyWong ^talk 22:47, 30 December 2010 (UTC)

Thanks, that was a pretty efficient 3rd opinion. History2007 (talk) 22:55, 30 December 2010 (UTC)

I've made some changes. Let me know if you have any comments. Thanks. SnottyWong ^comment 23:20, 30 December 2010 (UTC)

Pretty good fixes actually. And the table is worth 1,000 words - it highlights facts, e.g. about the UK. Now if you can dig up PCP 100 info, as discussed at the top of talk page, that would be great. Thanks. History2007 (talk) 23:37, 30 December 2010 (UTC)

By the way Snottywong two of the three articles on the specific centers need help. If you feel like making similar fixes there it will be good. Thanks. History2007 (talk) 12:21, 31 December 2010 (UTC)

References

Article needs updating

This page hasn't been updated in something like 4 years. "by 2012" Someone who knows something about supercomputing please share your knowledge on China's stuff. All I know is that Tianhe-2 is now ranked first in the world as of Nov 2013. 14.53.186.168 (talk) 12:33, 24 February 2014 (UTC)

Wiki Education assignment: Decentering Computer Science_ Transpacific Computing History

This article was the subject of a Wiki Education Foundation-supported course assignment, between 14 January 2025 and 7 May 2025. Further details are available on the course page. Student editor(s): Deszha, Hiiamhungry, Arcticrobot0612 (article contribs). Peer reviewers: Mangoes400, HoshiTurtle22, StudentEZH42B.

— Assignment last updated by Ihghjffhk (talk) 14:30, 31 March 2025 (UTC)

NPOV failure: Article reads as a pro China propaganda piece

This article is a disgrace. It reads like a propaganda piece. There are numerous absurd claims, such as that Intel CPUs are Chinese and that a 105 qubit quantum computer is a "supercomputer". No serious CS person would say that a 105 qubit quantum computer is a supercomputer of any kind. It is obviously just put there in an attempt to make a list of Chinese achievements look as impressive as possible.

Since this article is clearly in the context of national prestige it is paramount to point out how Chinese supercomputers for a long time were entirely based around non-Chinese hardware, even the Sunway is just a derivative of the Alpha, a RISC design, which is also non-Chinese. The entire tone of the article appears to try to put China in the best light possible, without even mentioning this until I started editing. It needs a rewrite to be neutral, please see Wikipedia:Neutral point of view. Amygdala and hippocampus (talk) 22:58, 27 May 2026 (UTC)

Has been somewhat improved now. The bits about restrictions probably need to go as well. There is no connection between the restrictions and supercomputing when China is able to produce its own processors. Amygdala and hippocampus (talk) 01:59, 28 May 2026 (UTC)

The article has the same issue as before but in the other direction now. It is not NPOV for the article to primarily focus on how "strong" or "weak" China is at supercomputing relative to the United States - compare supercomputing in Europe, supercomputing in India, supercomputing in Taiwan, etc. The whole framing of this article seems to revolve around a competition with the US and claim to the top spot, with normative statements like "China has been overall weak in the supercomputing domain." that are not in the sources cited. If the article has a purpose it should be to cover how, where and when supercomputing has developed in China with both factual information and perspectives provided by the sources, not an essay on if China is winning or losing a supercomputers contest. Gryfia (talk) 23:45, 4 June 2026 (UTC)

You appear to be missing a few things.

The bigger picture:

How is this article about "Chinese" (they are mostly not) supercomputers used?

Here is a very important example, from the China article on Wikipedia:

"Chinese supercomputers are ranked among the fastest in the world."

Full stop. The link in "supercomputers" is to this article.

Do you see the problem?

They falsely claim, in a context that is both historical and present (it is ambiguous) that China is a "world contender" or "superpower" within supercomputing.

See, the problem here is that:

Reality: China cannot even produce a competitive commercial home use desktop CPU.

Implied claim: China can produce supercomputers that are "ranked among the fastest in the world".

The reality is very different. You cannot buy a BMW and slap BYD on it and then be allowed to say "this is a BYD". Maybe if you have the paperwork right, you can, but you certainly cannot claim this as a national prestige achievement.

NPOV:

Why is that not NPOV (to say China has overall always been weak)? It is merely a summery of what the available references claim.

This is not a matter about culture and tastes. This is about hard numbers that can be directly compared, and it is about national prestige and national technological mastery.

The claim that China has always been weak in supercomputing is both true and is directly reflected in the already cited source: the authority TOP500.

Would you like me to rephrase it purely in terms of citing TOP500, year by year? I can do that. It will show that China has been weak overall. But there will be one citation for every year.

China is widely considered a superpower. China is also widely known for engaging in industrial scale plagiarism and IP theft.

This is not an article about culture.

Do you think it is completely irrelevant to even touch on, for an article about a topic as sensitive as supercomputers, which are directly tied into national pride and technical capability?

I hope you see the problem with China buying American components and slapping a Chinese label on them and then proudly claiming that they now have a Chinese supercomputer?

This would be akin to Toyota claiming they have the fastest car on the Nordschleife track, it is just that, the "Toyota" car, is actually a car using a Porsche chassis and suspension, Porsche brakes, and Porsche engine.

For people who are not intimately familiar with what "Intel" really is, for all they know, China might be producing these "Intel" CPUs locally. But that's not how it works. China is incapable of producing any competitive CPU. Amygdala and hippocampus (talk) 12:11, 5 June 2026 (UTC)

Uh, okay, I'm going to try and take this point by point because there's a lot going on here.

Firstly, if you have a problem with the language used on another page, take it up with the page in question rather than litigating it here. Wikipedia is not an R:S and this page is not the source for that claim, whatever it may be.

Secondly, the article is about supercomputing in China, not supercomputers built using only domestically sourced components. Supercomputing is a practice that covers the assembly, operation and usage of high-power computers. The source of the components used is a relevant piece of information, but supercomputers typically use essentially the same basic components as non-high power computers. There are articles on the semiconductor industry, and related CPUs, GPUs, etc. to talk about that. This article is about supercomputing in China.

Thirdly, as I already stated, that China is "overall weak" is a normative claim, and not one made by the sources. If the sources actually said that, then it might be reasonable to include it, at least with attribution, but as you admit the source for this claim is... Your own opinion. That opinion may be based on your reading of the information, but not one that it can be assumed another person would share based on the same information.

Fourthly, China's geopolitical status has no apparent direct relevance to this article, so I don't know why you even raise it, and whether "China" engages in industrial scale plagiarism and IP theft" is in itself a meta-topic. It is only relevant to the article if a source highlights IP theft/espionage as a particular source of supercomputing ability in China.

Fifthly - and I hope finally - this article does not inherantly have anything to do with "national pride and technical capability". Supercomputing may be a source of such, but as a wikipedia article its purpose is not to validate or disprove any such normative claims. It is, once again, an article on supercomputing in China, which should provide information and perspectives about the topic based on the sources, not an arena to litigate the weakness or strength of "China" as such. If you want to do that, please do it somewhere else. Gryfia (talk) 22:32, 5 June 2026 (UTC)

I don't believe you are acting in good faith any longer. The article is back to reading as a propaganda piece on China. I know exactly what you're doing. I can read between the lines.

Don't worry though, I'll find more sources and rewrite it again.

You're just wrong if you believe that supercomputers are treated in isolation in China or any country. What defines a supercomputer is its contextual and relative position to other computers.

WP:LEAD which is basically "establish context": The lead should identify the topic, establish context, and summarize the most important points according to reliable sources. For a national-supercomputing article, "relative to the U.S." or "global frontier" is obviously contextual when the article discusses performance or technological position.

As for the rest: This is simply invalid reasoning.

As I said, supercomputers are not defined by China. It is very much relevant for the article

The current version reads as if China is an impressive, world-leading authority in supercomputers, and have been so since the 1980s.

No serious scholar would do anything but roll their eyes at yet more propaganda.

You assert that I make a normative claim. Do you know what a normative claim is?

A normative claim is a statement attempting to assert how things should be, not how they are.

How things actually are: The source of the claim is TOP500. TOP500 is the authority on the rankings of supercomputers, something you should know since you seem very interested in supercomputers.

And I think you do know, but you're playing a game here.

Again, how is this a normative statement? Because you don't agree with my reading of TOP500 (which, if so, is very interesting, please tell me why you don't agree)? You want more sources? This is a very different argument than attempting to assert that the claim itself is normative in nature. Amygdala and hippocampus (talk) 01:28, 6 June 2026 (UTC)

Why exactly is it that you do not think I am acting in good faith? That you disagree with my position does not mean I am acting in bad faith; as I've expressed repeatedly, I'm trying to create a good article on supercomputing in China.

Of course supercomputing is contextual in every country and "HPC" is in a sense relative to the notion of "non-HPC" at any given time, but that is not in dispute here, and does not change the fact that supercomputing is about considerably more than the origin of components inside a supercomputer, which seems to be your sole concern. This weight is simply not reflected in the sources.

I actually agree with you on the lead and would like to have polished the previous one, but the version you have now fails on your own criteria. You've moved a one-sentence paragraph to the top on the origin of components in 2015, followed by another one-sentence paragraph that exclusively draws on the weaknesses slide of a 51-slide presentation from 2015. That is not "appropriate weight", per WP:LEAD.

I do think it is worth referring to the global position of supercomputing in China for context, but I don't think it should overwhelm the lead and article - because the article is about supercomputing in China specifically.

I don't know what you mean by "supercomputers are not defined by China"? Of course they aren't, as far as I'm aware that is not in dispute. I also think that is a very skewed reading, frankly. The article did not make any such claim, this is something you are projecting onto it because it did not make an explicit claim - entirely without expression in the sources, as yours did - that China is "weak" in supercomputing.

You are right on normative, though. The word I was looking for was "subjective". With that said, Top500 is indeed a significant source, but you were not quoting the authors or creators of the Top500, you were giving your opinion on what the rankings signified.

In any case, I'd like to actually make progress on the article, but that would mean being able to discuss it in good faith without these vague and, frankly, bizarre intimations about me (what "game" is it that you think I'm playing here?) Otherwise, if this is just going to be an endless edit war, we should seek outside input before continuing, e.g. by requesting a third opinion? Gryfia (talk) 16:19, 8 June 2026 (UTC)

You appear to be selectively ignoring what I'm writing.

There are two absolutely critical concerns:

1. How often has China actually been at the top since the first supercomputer was introduced? And how many supercomputers did they have, in relation to the U.S.? If you think this is not important in general and not a matter of national prestige then you're fooling yourself. This is the thing many people care about and this is very much a competition.

2. Of the mostly few, rarely many (ratio), supercomputers that China has fielded, how many were actually using original, domestic components?

The combination of these two are critical to point out in an article like this, and yes, they very much deserve to be the highlight. Simply because in combination, they are absolutely devastating in terms of what China achieves in the supercomputing domain.

You are simply twisting the truth if you write it in any other way. No, you shouldn't write an original paper, Wikipedia is an encyclopedia, nor did I.

Again, here is the point, with a befitting analogy:

Who cares if China purchases five hundred thousand Ferrari sports cars and puts a different label on them, and then claims to be a "superpower" in the sports cars domain? Don't you see how absurd that would be?

No one would care in terms of national prestige, that's for sure. Some might notice, and point out the blatant plagiarism, if no credit was given to Ferrari.

This is the reality of the situation. And no, I didn't point it out nearly this harshly in the actual article. I simply mentioned it, humbly, and with a reference to TOP500, which is what's required.

Now there are a couple more references that highlight the problem with supercomputing in China, respectfully.

You seem like you're acting in good faith now, but:

I believe you are not acting in good faith when you start to introduce your own, original (implied!) definition of what a supercomputer is by writing the article in such a way that the notion of a Chinese supercomputer becomes completely meaningless.

I believe you are not acting in good faith when you claim I make a normative (now subjective) claim, when I have made no such claim: I am simply citing TOP500. There is no need to have a separate, additional source confirming TOP500 as long as you don't use that information in an original way. Which I did not.

I wrote:

"China has had periods of TOP500 leadership, but this has not amounted to sustained indigenous dominance."

This is not only correct, but an incredibly important piece of information. If you simply write that "China started in year X and rose to the top of TOP500.", which was the tone of the article both before I started editing and after my edits were reverted, then an average reader will very likely read this as: China is very possibly world leading in supercomputing. Wow! This is impressive stuff!

This is an outright lie and sheer propaganda.

You cannot claim that you can match Ferrari sports cars in track performance, nor should you imply that you are matching them, when you are making ordinary family sedans. Using engines, brakes, and suspension you didn't even design nor manufacture.

"but you were not quoting the authors or creators of the Top500, you were giving your opinion on what the rankings signified."

I don't need to. I can simply cite the facts stated at TOP500.

The numbers are what matter, and the origin of the parts.

Note that I am not introducing any original research. I am simply stating that yes, they had TOP500 leadership, but, not many times, i.e. not sustained. Where is the OR here? This is a direct reading of the numbers, nothing more.

"I don't know what you mean by "supercomputers are not defined by China"? Of course they aren't, as far as I'm aware that is not in dispute."

The way the article was written it treated them basically in isolation in China. The article focused entirely on how China had improved without making appropriate comparisons where it is painfully needed.

From the very start to the end, proudly mentioning the Yinhe-1, being introduced decades after other supercomputers, and painfully underperforming, yet, this is hailed as an impressive achievement? It is not. That has to be pointed out. Then it goes on to talk about how China has "woken up" and basically outsmarted the rest of the world, when the reality is that China does this almost entirely buying non-Chinese components. Finally, the article would mention how China would have done even better, had it not been for U.S. sanctions---which ironically makes China have its cake and eat it too: The article pretends to ignore the fact that China uses U.S. technology, then points out afterwards how they are dependent on it, while not admitting that this very dependency actually undermines Chinese achievements in the supercomputing domain. Amygdala and hippocampus (talk) 23:29, 9 June 2026 (UTC)

Our disagreement is founded on the fact that you regard those two points as significantly more important than I do. Again, just look at the article's on supercomputing in Europe, India, etc. I believe it is important to situate the content of the article, which is about supercomputing in China, in a wider context - but as context, not the most important item of information.

I'll refer to a couple of definitions of "supercomputing" from what I hope you'll agree are reasonable sources. To quote from Deloitte, "Supercomputing is the use of a computer to solve a problem that requires a lot of computational power, a lot of data, or both. Comparatively, high-performance computing is the integration of all the knowledge and tools necessary to construct parallel supercomputers." Or from Hewlett Packard, "The term "supercomputing" refers to the processing of massively complex or data-laden problems using the concentrated compute resources of multiple computer systems working in parallel (i.e. a "supercomputer"). Supercomputing involves a system working at the maximum potential performance of any computer, typically measured in Petaflops. Sample use cases include weather, energy, life sciences, and manufacturing."

The source of components within these machines does matter, but the ability to produce given components does not equate to supercomputing proficiency, and vice versa. I also think you are heavily projecting what people might read from your own biases; the article never contained any claim that China is a "superpower" in supercomputing. Similarly, it never drew any conclusion about China's relative supercomputing capabilities from stating the fact that supercomputing in China began in 1981 and, the fact that machines operated in China have at times led the TOP500 rankings. I don't really know why the lead needs to bend around your particular obsession with making sure nobody thinks that China has had any achievements in supercomputing.

In any case, it seems that one of the "Supercomputing Centres" over there has decided to get back on the board with TOP500, and submitted a machine which, per TOP500 themselves, is leading the list using "a custom Chinese processor" (more info on DCD and HPCWire). This would seem to make your framing here a bit moot. Gryfia (talk) 01:17, 24 June 2026 (UTC)

I don't agree with anything here and I question not only your command of this field but also whether or not you've even read the article which you are criticizing.

You claim that I have an "obsession" and that "my point" is now "moot" because China has fielded a machine using custom processors.

This is not a unique event. China has already done this before, and I believe I was even the one who wrote the current sentence:

"Sunway TaihuLight used domestically produced manycore processors"

And I'm sure it was also present in the article previously too, but I'm not going to waste time to check because that is moot.

In brief, when you point out how I am "obsessed" and how "my point is moot" it reads as a thinly veiled personal attack and undermines any serious, credible points you might have.

You're using more than just some accusatory terms here: Before you were telling me I was making a normative claim, which I explained I am not. I think you eventually agreed I wasn't.

Now I'll explain that I am not biased.

What is "biased" anyway?

Typical good enough definition from any dictionary:

"a disproportionate weight in favor of or against an idea, thing, or group, usually in a way that is closed-minded, prejudicial, or unfair"

How am I "close-minded"?

How am I "prejudicial"?

How am I "unfair"?

There is no "bias". I am simply pointing out the reality, and I have TOP500 as well as other non-biased sources to support the interpretation, such as the already cited Qian. The situation is simple, please read carefully: If you integrate over time China's achievements and normalize for the size of China as a country, number of citizens, resources, etc. then China is doing extremely badly in supercomputing. Full stop. This is simply what the data indicates.

What would be unfair, is to take a country e.g. the size of the Netherlands, a mere 18 or so million people, and claim that they are "doing poorly" in computing, because, look at what USA manages. Why isn't the Netherlands at the same level, or even better?

Problem: The Netherlands is a small country. Why would you expect them to perform so well given this constraint?

Plot twist: They do. They do very well. They have ASML.

If they were 1.4 billion, with only something that amounts to SMIC (this is what China has, in computing), then they would be doing "extremely badly".

Because of normalization. You need to normalize in order to be fair. And that's exactly what I am doing.

You cannot take e.g. Latvia and claim they must be "able to compete with USA". Latvia has an excuse: They are a nation of some 1.8 million people. They are not a superpower, with enormous natural resources, enormous access to talent and people, and they haven't been trying for many decades to do well in computing.

And yes, none of this direct reasoning has any place in the actual article, it is OR by definition without citations. There is no need to point this out to me, but this is the truth and the correct interpretation and anything that steers away from this actively in the opposite direction, i.e. pretending that China is better than the United States, I will not accept any version of. Nor should you. It is an outright propaganda lie, and a pathetic one at that.

The disagreement is entirely the tone of the article and what impression the reader is left with after reading it. The rest is really just details.

This definition you quoted, while simplified, is good enough:

"Supercomputing is the use of a computer to solve a problem that requires a lot of computational power, a lot of data, or both."

And what is "a lot of computational power"?

And does it matter if this "a lot of computational power" is actually Chinese, in an article about Chinese supercomputing? Even you yourself said it does matter.

Indeed, "The source of components within these machines does matter", and while you're right that "supercomputing proficiency" is an entirely different matter---anyone could in theory own the world's fastest supercomputer, it is what you do with it that matters.

So how well does China do with its computational power and research in general?

How many Nobel prizes, how many inventions does China have? If you'd like, I could go into great depth about that too.

Would you like me to cite: https://people.idsia.ch/~juergen/med.html ? Where is China here? Can you even see them?

Or would you like me to start citing the lack of Chinese inventions, which there is already a nice summary of, here: https://en.wikipedia.org/wiki/List_of_Chinese_inventions#Modern_(1912%E2%80%93present) (and yes, I know I cannot cite Wikipedia within Wikipedia, but I'm sure I can work around that by citing the actual sources).

So you see, what you're trying to do, by pointing out that we need to separate the lack of hardware from the lack of achievements, that would if anything only make China seem less impressive. Not more. Just producing thousands of random frivolous research papers is just the game of publish or perish in the world of academia.

I'm not sure if this is the place of all of this though, to start directly pointing out China's lack of achievements? I think it makes a lot more sense to just stick to the hardware and to point out what China does with its supercomputers. Describe what supercomputing centers exist, etc.

But to just ignore the fact that China has heavily relied on non-Chinese supercomputers in an article about Chinese supercomputers? I think not. To just ignore the fact that China used MIPS without a license? I think not. To just ignore the fact that China was 20 years late to the party with its first supercomputer? I think not.

And no, this doesn't mean that I think we should just ignore China's actual achievements.

You're free to add whatever success stories you want that you think are relevant. I'll do the same, if I ever come across something noteworthy, but I'll also reserve the right to add peer reviewed sources that point out lack of achievements, or failures.

As for the latest bit of news, sure, this isn't the first time China made a supercomputer with their own hardware. But this is the thing, what is really "Chinese" hardware, here?

Is an AMD CPU really just "the work of AMD", in an article about AMD's CPUs? Or is it noteworthy that AMD is implementing their own version of what Intel made? Is this something that should be pointed out? This doesn't mean that AMD doesn't get credit where credit is due. I am a fair person, obviously AMD gets credit for e.g. x86-64, I won't go into more details, the analogy is sufficiently explained.

If you do look at Chinese "custom" hardware, what is really Chinese? MIPS? The bus? The memory architecture? What about the OS? What about the interconnect? Is a fair question to ask: Did China invent or make anything? The interconnect is what many would say almost defines a supercomputer, e.g. the old Cray supercomputers were basically defined by the proprietary Gray Gemini interconnect.

What, if any of this, did China invent, or make? Is this maybe something to bring up in an article about Chinese supercomputers? Amygdala and hippocampus (talk) 23:23, 25 June 2026 (UTC)

AI tag

I have restored the AI tag, as the text added here has not been fully addressed and much of it remains verbatim. This is probably the root cause of the whole debate above, both of you are most likely arguing about a chatbot. Gnomingstuff (talk) 20:15, 10 June 2026 (UTC)

Thanks for the heads up. I'll get around to it.

But the abovementioned discussion has little or nothing to do with this. The problem is the misrepresentation of China's achievements in a domain where the very definition of what is "super" is only relative to other computers, etc. I won't repeat the whole argument here. Amygdala and hippocampus (talk) 21:52, 22 June 2026 (UTC)