Relax. I'm in agreement with you. I'm entitled to a bit of fun. This was going much better than the last time until you showed up, stinker.
That's because scientific and technological products and endeavors are a province of business (often including businesses that draw money from the public sector and the government for contracts which are sensitive to security). One industry that uses supercomputers and processing power is the film industry. This is an artistic industry that depends on tech-related apparatuses. Your call for us to "work on our science/tech" is really a call to grow our businesses.
Almost all of the Chinese supercomputers out there have Intel technology in them. You want to know the only microprocessor companies in the world that come close to rivaling Intel? AMD, Apple, Qualcomm. Right now the South Koreans are in a distant second with the exception of NVIDIA for GPU's (ex. Samsung who manufactures Exynos), and the Chinese are....I wouldn't even know where to put them. They're probably not even yet at the level of Texas Instruments in terms of technological sophistication; the difference is they're still working at it. It's just they are manufacturing like a motherfucker and have a massive stream of revenue.
This has NOTHING to do with the sophistication of their microprocessors and EVERYTHING to do with the value at which they sell them. This is why short-toothed producers like Mediatek and HiSilicon are rapidly advancing (also because the Chinese shamelessly steal IP with industrial espionage).
For those who don't understand what the second two paragraphs were communicating, it was an allusion to how supercomputers are no longer really a matter of architectural rivalry, in terms of the sophistication of the designs, as it was during the height of the Cray vs. IBM rivalry of the 90's, but a race of parallel processing, reflecting the cheaper throw-cores-at-the-wall strategy popular with Chinese semiconductor upstarts, where they link together as many possible CPU clusters as they can onto a single football field in order to claim the top spot. So it's a bit of a dick-measuring contest.
For that reason the supercomputer race in some ways reminds me of that major reveal of Elon's magical battery pack in his Tesla cars. Everyone was expecting some groundbreaking technology, but instead it was really a refinement of previously existing battery technology, just with a ton of these batteries linked together in the trunk.
The TOP500 really is a measurement of business power more than technological proficiency, because the revenue to build these things is always going to be drawn from a combination of sectors, but more typically public investment; of course public investment is predicated on tax revenue, and that is strongly correlated with a country's GDP, and directly with its business garnishments.
More saliently, almost all supercomputers are so expensive, and answer such a narrow need, that very few of them are privately owned at all. In fact, one of the questions I
don't know the answer to off the top of my head is the last time-- if
ever-- a privately owned supercomputer held the #1 position. Off the top of my head, the most powerful one I can think of that I know was purely owned privately was none other than Deep Blue, which beat Kasparov, and IIRC it was only the 16th most powerful in terms of FLOP power when it was created. That isn't what set it apart.
This is a sobering truth for any conservative. It challenges the heart of the philosophical conviction that the private sector is always best suited to answer a need. There is almost no private entity large enough to own one that would have a need for it. Interesting quandary, indeed.
While, unlike you, I did unironically support corporate tax reform, like you, I did not support the latest budget. I weary of this sudden zeitgeist among conservatives where we superstitiously attribute economic windfalls where there are none. When the economy was strong under Obama, around where I live, I often heard the pro-Trump voters echo the pro-Bernie voters with comments like, "That's Wall Street, not Main Street." Main street is definitely doing a bit better, at the moment, and kudos to Trump for this, but I can't remember the last time I heard a Hannity fan reference the Labor Force Participation Rate, or crow that the sky is falling because of the expanding debt or debt-to-GDP ratio. Apparently the worries of the inflation-to-wage ratio also suddenly disappeared merely because wages have gone up a bit.
While I do believe that conservative approaches to economics and government can also succeed in the technological race, I want to generate awareness and reflection on the opportunity cost of those uncollected billions, and perhaps remind Trump supporters of the benefit that government spending in scientific research and development can have. I am quite liberal when it comes to this issue. It's one of the only things I think is worthy of public spending. And guess what? We're spending a shitload. If we're going to run up that debt let's run it up on stuff like this.
Trump is the President who appointed Rick Perry, the man who in his own campaign promised to abolish the DoE just four years earlier (well, if he could have remembered), a few years prior to this Obama initiative. So the man who heads the body that financed this supercomputer, and will operate it for the American public good, is the same one who doesn't think that body should exist. Odd. If one frames the building of this computer positively as an accomplishment of Trump, when he wasn't even the President who ordered it, suddenly pro-Trump posters will whistle and holler their approval.
Isn't that ironic? I don't think it should matter.
/cdn.vox-cdn.com/uploads/chorus_asset/file/11525117/msingleton_180612_2663_0005.jpg)
