Posted on 06/23/2026 8:14:18 PM PDT by SeekAndFind
China took back a coveted computing crown from the United States on Tuesday, ratcheting up a fierce technological competition that has implications for science, national security and geopolitics.
LineShine, a massive computing system in Shenzhen, China, was declared the world’s fastest by a group of researchers using a set of standard tests for supercomputers. Besides raw speed, the system stood out because it uses only standard microprocessors and not the special-purpose chips called graphics processing units, which most high-end supercomputers rely on for heavy number crunching.
That underlying design could point to a better way to blend artificial intelligence with traditional scientific tasks, said Jack Dongarra, an organizer of the so-called Top500 list of the world’s most powerful supercomputers.
Dr. Dongarra, a professor of computer science and electrical engineering at the University of Tennessee, recently inspected the new machine, at the Shenzhen Cloud Computing Center. LineShine’s test results were more than 20 percent faster than those of El Capitan, a system at Lawrence Livermore National Laboratory in California that has topped a twice-yearly ranking of supercomputer performance since November 2024. China had not placed a machine at the top of the list since 2017.
“It’s an impressive system,” Dr. Dongarra said of LineShine. “They upped us by developing a system that is not reliant on GPUs.”
The new supercomputer adds to the race between China and the United States for technological supremacy. U.S. tech giants like OpenAI, Anthropic and Google have developed leading A.I. models, while another American company, Nvidia, has become the world’s dominant supplier of A.I. chips. China has tried to innovate in different ways, with the Chinese start-up DeepSeek releasing a cutting-edge A.I. model
(Excerpt) Read more at nytimes.com ...
“because it uses only standard microprocessors and not the special-purpose chips called graphics processing units”
which is like hitching a metric shitton of Model Ts together to pull a train, instead of using a modern diesel-electric locomotive ... using standard microprocessors will use MASSIVE amounts of power [and cooling] compared with equivalent sophisticated GPUs ...
Like watching Andy Griffith make a spaceship out of junkyard parts.
Horrors.
without even reading the daily Mail, I know it’s trump’s fault.
>> like hitching a metric shitton of Model Ts together to pull a train, instead of using a modern diesel-electric locomotive ...
ROFL! Good analogy.
>> I know it’s trump’s fault.
Of course it’s trump’s fault! Because Trump. :-)
how many super computers does it take to equal 1 AI data center?
More CCP propaganda! Don’t believe it!
Yah but half an hour later you have to compute again.🙄😏
Salvage I,fun show.🤔
China reports....
I dont believe it.
I bet none of the tests involved cryptocurrency mining or “AI anything”.
The supercomputer is able to generate 1 trillion fortune cookie sayings per second, as well as generate winning lottery numbers.
And WHO’S chips did they use? Their own?
With stories like this they may sound convincing, but there always seems to be a few truths not completely explored.
Their supercomputer love you long time.
RE: how many super computers does it take to equal 1 AI data center?
Well, I’m doing some simple calculations…
What counts as a “supercomputer” today?
The world’s top traditional HPC supercomputers (e.g., DOE’s El Capitan mentioned in the article ) typically have 40,000–50,000 GPU-equivalent accelerators.
What counts as an “AI data center”?
Modern AI data centers—sometimes called AI factories—are built around 100,000–200,000+ GPUs per cluster.
Just as example… Elon Musk’s xAI Colossus Phase 2 has 200,000 H100-equivalents (largest known).
These are just single clusters, not entire campuses. A full AI data center campus may host multiple clusters.
So how many supercomputers equal one AI data center?
Using GPU-count comparison (most direct) and
Using El Capitan (~44k GPUs) as the baseline, A single modern AI cluster = 2 to 5 of the world’s largest traditional supercomputers.
And note …. AI data centers often contain multiple clusters, plus:
massive power infrastructure (50–140 kW per rack vs. 8–15 kW in legacy DCs)
unified GPU fabrics behaving like one giant machine (supercomputer-like)
liquid cooling and high-speed interconnects at supercomputing scale
So, A full AI data center campus may host 3–10 clusters, depending on design.
Therefore, a full AI data center campus can equal 10–50+ traditional supercomputers in aggregate compute.
“Israel’s IC manufacturing primarily focuses on specialty analog, mixed-signal, and silicon photonics foundries rather than high-volume, leading-edge digital logic. The ecosystem is heavily defined by a robust fabless and R&D sector, alongside two major domestic manufacturing facilities”
so, probably the advanced specialty ones are their own: “photonics” ICs are the tip-off regarding focused beam weapons ...
To my understanding, leading photonic IC’s are currently being manufactured by Intel, IBM, and Cisco.
...I suppose my point is more, China didn’t develop the processors, neither the schematic nor the manufacturing. They’re putting existing, non-Chinese, processors together in a proprietary system architecture.
Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.