＜prologue＞
I started a blog called “The Baby Boomer Generation’s Miscellaneous Blog”（Dankai-sedai no garakutatyou:団塊世代の我楽多（がらくた）帳） in July 2018, about a year before I fully retired. More than six years have passed since then, and the number of articles has increased considerably.
So, in order to make them accessible to people who don’t understand Japanese, I decided to translate my past articles into English and publish them.
It may sound a bit exaggerated, but I would like to make this my life’s work.
It should be noted that haiku and waka （Japanese short fixed form poems） are quite difficult to translate into English, so some parts are written in Japanese.
If you are interested in haiku or waka and would like to know more, please read introductory or specialized books on haiku or waka written in English.
I also write many articles about the Japanese language. I would be happy if these inspire more people to want to learn Japanese.
my blog’s URL:団塊世代の我楽多（がらくた）帳 | 団塊世代が雑学や面白い話を発信しています
my X’s URL:団塊世代の我楽多帳（@historia49）さん / X
In 2009, during the Democratic Party of Japan (DPJ) administration, the development of the supercomputer “Kei computer” was delayed due to the budget cut by Renho, a member of the Diet, who asked, “Wouldn’t it be okay if it came in second place? ” The development of the supercomputer “Kei computer” was delayed after its budget was cut due to the famous and impactful phrase, “Why can’t we be second? In June 2011, however, it became the world’s fastest supercomputer (10.5 peta, or more than 1 Kei computation per second), proving the high level of Japan’s scientific and technological capabilities and bringing it to the forefront.
The “Kei computer” will cease operation in August 2019, and the baton will be passed to its successor, “Fugaku. Fugaku” is scheduled to be in operation from around 2021.
The Kei computer started full-scale operation in 2012, which means that it had been in operation for seven years.
１．What is the Fugaku supercomputer?
The next-generation supercomputer “Fugaku” will be 100 times faster than the “Kei computer” with about three times the power consumption.
2. What about relocation or reuse of the previous model “Kei computer”?
The supercomputer “Kei computer” at “RIKEN” in Port Island, Kobe, will cease operation on August 30, and its dismantling will begin after September, with its successor “Fugaku” to be installed at the same location.
By the way, as a layman, it seems to me that the hardware of the “Kei computer,” which cost 79.3 billion yen for development and other costs, would be wasted if it is not relocated or reused, but this seems to be impossible as a practical matter.
The reasons why relocation or reuse is impossible are as follows.
(1) 10 billion yen in annual maintenance costs are required
The 10 billion yen needed for electricity, water for cooling, and maintenance costs means that this is not a facility that can be operated at a huge expense as it deteriorates over time.
(2) It is also impossible to divide the 864 racks in total and install them in universities across the country.
A new water-cooling facility would need to be constructed, which would cost 60 million yen to relocate and more than 100 million yen in annual maintenance costs, and there is no way to guarantee future maintenance and inspection work.
(3) It is impossible to disassemble and sell the “board” (system board) or “storage device” (memory)
The “board” and “CPU” are designed “exclusively for the Kei computer,” which means they are not versatile enough to be used in ordinary personal computers.
That is why, in the end, RIKEN is extracting metals and other materials from the parts and selling them to cover part of the removal costs, and is also working to have some of them stored in science museums across Japan as “historical materials for computational science.
3. the current status of the supercomputer development competition
By the way, I would like to take a look back at the first comment made by Renho, “Wouldn’t it be okay if it came in second place? ”
In the world of sports such as track and field and swimming, the “speed race” has become so heated that it is causing various adverse effects. In golf as well, I feel that the “distance race” has gone too far.
This is why I began to wonder whether it is right to pursue only “the world’s fastest computation speed” in the supercomputer development race as well.
According to Mr. Shin Ito, former Counselor of the Cabinet Office and “General Director of Conceptual Japan,” the true meaning of that statement was as follows.
① What is important is not only the speed, but also the ease of use for users (researchers)
② Is there really a need in industry for a supercomputer with a “10 peta” computing speed?
③ While the U.S. is trying to develop a 10 peta supercomputer by 2012, even if Japan’s supercomputer becomes the world’s fastest, it will not stay the world’s fastest forever.
④ The idea of “cost-effectiveness” does not necessarily fit in the world of science, but isn’t it a problem that we cannot see the results of more than 100 billion yen of taxpayer money being invested?
This makes me feel that I have a “point”.
As a result of this project screening, the Ministry of Education, Culture, Sports, Science and Technology (MEXT) has shifted to a “user’s perspective. They have linked the supercomputer to be developed with domestic supercomputers through a network for joint use, delayed the development timeframe, and reduced the total development cost.
In the “calculation speed ranking” of supercomputers in June 2019, the Kei computer was ranked 30th in the world and 3rd in Japan. The world’s fastest is the American “Summit” at 149 peta, and Japan’s fastest is AIST’s “ABCI” (ranked 8th in the world) at 19.9 peta.
Because of this, the “Fugaku,” the successor to the “Kei computer,” will “focus on user-friendliness and further improve practicality” (Mainichi Shimbun, 9/3/2019), and will not aim to be the world’s No. 1 in speed.
The “Kei computer,” which ceased operation at the end of August this year, has also greatly improved its characteristics other than speed, such as being ranked first (2016) in all four categories of the “HPC Challenge Award,” which measures superiority in memory access speed and performance, even after it ceased to be the world’s fastest.
Furthermore, in the “Graph500” performance ranking for “analysis of big data, which requires large-scale and complex data processing,” the company maintained the No. 1 position for nine consecutive terms (five years) from June 2015 to June 2019. This can truly be said to be the result of development that is not concerned with speed.
Looking back in this way, it seems that I misunderstood Councilor Renho’s statement. I suspect that many members of the public also misunderstood it because of their preconceived notion that “business sorting is a performance.