Chapter 371: Superconducting Chip
Robots laying photovoltaic panels are much more efficient than humans.
Because of wearing spacesuits, human astronauts during the laying process, when watching the live broadcast, whether from first-person or third-person perspective, you always feel that the movements are awkward, with an illusion of the body being out of control.
But robots are much smoother, especially the robots just shipped from Earth; new things are always the most satisfying to use.
The audience in the live broadcast room simply feels refreshed, similar to watching decompression videos.
One photovoltaic panel after another is planted like crops.
“So refreshing, our progress is too fast.”
“Watching this time, the progress feels lightning fast. When will Huawei manage to develop the superconducting chip?”
“It should be soon, right?”
“It should be soon. If it weren’t for the superconducting chip, there probably wouldn’t be need for such a massive photovoltaic array, right?”
In the live broadcast room, besides discussing robots installing photovoltaic modules, it’s all about when the superconducting chip will be ready.
This is a technological breakthrough unique to China.
Even though this technological breakthrough is still only at the concept blueprint stage, netizens on the Chinese Internet have already won to the point of numbness.
Not only Chinese netizens won to the point of numbness, foreign media also prematurely celebrated China’s win to the point of numbness.
Take India as an example; Indian media loves reflection the most, especially comparing with China.
European and American newspapers can ignore this matter, but India cannot.
From the appearance of Deep Red, Indian media has been reflecting:
“We are significantly behind China in research and development spending, research talent density, university rankings, and other aspects. China has national strategies like the Thousand Talents Plan and Made in China 2025; we lack such long-term planning. In the artificial intelligence field, China’s investment far exceeds India’s. This is the reason for the gap between us.
The Indian government has ambitions and talents in artificial intelligence, but our country’s venture capital is unwilling to support true technological innovation; they only want to be technology porters, bringing mature American technology back to India.
Such innovation is unsustainable, and this is the fundamental reason for the gap between us and China.”
When Deep Red came out, India’s reflections gave Chinese people a sense of familiarity, because China’s previous reflections on the gap with America sounded similar.
But after the superconducting chip plan was announced, Indian reflections started to depart from rationality, often venting emotions like Indians are just no good, India is just no good.
There is a Chinese descent person’s answer on Quora that hits the nail on the head:
“India has always thought they can compete with China and should catch up to China: Mumbai with Shanghai, New Delhi with Yanjing, Bangalore with Pengcheng. China’s technological progress particularly draws their attention.
And this time, India’s collective breakdown around the superconducting chip has triggered wider and more intense discussions in Indian public opinion than the appearance of Deep Red, which confuses people unfamiliar with India. The superconducting chip is only at the theoretical level with no physical prototype yet; Deep Red is a real product with weaker computing power but better effects. Why does the former trigger even more breakdown in India?
On this, I have many Indian friends around me. I think their thinking logic is that we are the same as China; although China is ahead of us now, there is no essential difference—we are both chasers.
No matter how powerful Deep Red is, it is an imitator of GPT. Even if it surpasses GPT, its appearance is not long after GPT; maybe it’s original, but in Indians’ perspective, you are imitating, plagiarizing.
So their discussions and reflections are still at a relatively peaceful and rational level, but the superconducting chip is a completely non-existent product, a concept first proposed by China with the possibility of implementation.
This makes India realize that maybe everyone is different; China may not be a chaser; China may be transforming into the role of innovator in the technology field, or even already has become one.
Digging deeper, India’s self-positioning is still a country developing by relying on Europe and America capital, Europe and America technology, and Europe and America market. They think China is the same, so everyone is in a competitive relationship: you get more, India gets less.
But the emergence of the superconducting chip makes India realize that we are also accepting capital, industry, and technology export from China, which breaks them down, because unknowingly China has achieved a transformation in identity and status, but India is still India.”
Of course, around the superconducting chip, not only India is paying attention; developed countries are also paying attention.
This may relate to the next generation of chip materials.
Practitioners in the chip field can intuitively feel that silicon-based chips have reached a limit; every step forward is extremely difficult, with rising costs, falling yield rates, and various negative factors emerging.
If not adopting more advanced 3D structures, silicon-based chips will reach the end in a few years.
So whether China’s low-temperature superconductivity route is feasible has become the focus of industry attention.
The Moon can guarantee constant low temperatures and utilize low temperatures to build superconducting chips, which naturally cannot be done on Earth. They are concerned about whether low-temperature superconductivity might exhibit some interesting characteristics, and whether these characteristics can guide the emergence of next-generation chip materials.
Superconductivity itself sparks endless imagination, so is room-temperature and atmospheric-pressure semi-superconductivity possible without needing superconductivity?
New environments and new conditions may give birth to new materials.
Therefore, the industry is particularly focused on the latest progress of China’s ultra-low temperature superconducting chip.
Of course, internally at Huawei, it’s even more valued; they mobilized troops, dispatching the most elite team from Songshan Lake to Shanghai.
Their main task in the first year is to verify technical feasibility. The technical path is already determined: using iron-based superconductor FeSe thin film, on SrTiO3 substrate via molecular beam epitaxy growth, achieving superconductivity at 100K temperature. Such a sample is theoretically feasible, but in practice?
How does it perform on the Moon? Not just computation itself, but stability, power consumption, etc., how about other states.
They need to produce a sample first.
With Apollo Technology’s capabilities, once they have a sample, they can immediately send it to the Moon for testing.
The environment on the Moon and such are all prepared; power is available, shadowed region exploration completed, ready for testing anytime.
It’s all set except for the crucial part.
“Engineer Wu, how’s the progress on your side?” Lin Ran is also concerned about this; he holds a meeting with the technical team about once a week. The technical team is jointly built by Huawei and Apollo Technology, with personnel ratio about 7:3.
Engineer Wu is the specific head of this technical team, a senior engineer in Huawei’s semiconductor line second only to Liang Mengsong.
First month: “Professor, we started with FeSe. Bulk FeSe is a semiconductor with Tc of only 8K, but single-layer thin film under interface effect can reach 109K.
The lunar vacuum environment perfectly matches MBE growth, avoiding oxidation.” Engineer Wu said.
The team’s researchers wearing goggles, operating equipment: first heat SrTiO3 substrate to 600°C to clean the surface; then control evaporation rates of iron source and selenium source, iron atomic beam intensity at 0.1 monolayer/min, selenium in excess to ensure stoichiometry.
During growth, Engineer Wu occasionally corrects parameters: “Watch the substrate temperature; too high will cause lattice mismatch, reducing electron-phonon coupling. Target thickness is about 0.5nm single atomic layer.”
After the first sample growth completed, they checked crystal structure with X-ray diffraction (XRD): peaks show good epitaxy, but resistance test in liquid nitrogen bath (77K) has Tc of only 50K, far below expectation.
Second month: “I think it’s selenium vacancy defects causing incomplete Fermi surface reconstruction. Engineer Wu, try adding post-annealing step, heat to 400°C in vacuum to promote interface charge transfer.” Lin Ran reminded, “I think interface effect will be key; SrTiO3’s polar layer will induce two-dimensional electron gas, boosting Tc.”
This relates to a 2014 Nature paper, which mentioned that the FeSe/SrTiO3 system can use interface effect to push Tc from 8K to over 100K.
The team iterated three times, adjusting Se/Fe ratio from 6:1 to 8:1, finally seeing progress on the fourth sample: XRD shows sharp peaks, indicating perfect lattice matching.
Third month, initial signs of dawn: using high-pressure oxygen doping, FeSe thin film’s lattice distortion, a-axis parameter from 3.76 to 3.78, enhancing electron-phonon coupling.
In simulation observations, Tc reaches 105K.
Lin Ran said: “I know everyone is excited, but this is not enough; we need to keep optimizing.
Because the lunar south pole’s radiation environment will interfere with Cooper pairs, but low temperature can suppress thermal noise.
We need to integrate radiation shielding layer, using boron-doped diamond as buffer. BDD’s Tc is only 11K, but its wide band gap can block cosmic rays.”
They started doping experiments: introduce oxygen beam in MBE chamber, pressure controlled at 10^-6 Torr, doping level 0.1-0.2 atomic%.
Testing used four-probe method to measure resistance-temperature curve: under helium cryocooler, cooling from 300K, resistance drops sharply to zero near 110K, magnetization test confirms Meissner effect, critical current density Jc reaches 10^5 A/cm².
“Professor, based on failure sample analysis, STM shows oxygen clusters causing phase separation.” Engineer Wu said.
Lin Ran thought for a moment and said: “Is adjusting oxygen beam energy feasible?”
They adjusted oxygen beam energy from 5eV to 3eV for uniformity optimization.
Fourth month, the team finally made the second sample: a 5cm square chip, surface shimmering with metallic luster, integrated BDD shielding layer 2μm thick.
Testing under liquid nitrogen simulation, resistance drops to zero, capable of running simple AI algorithm: chip processes 100×100 matrix multiplication, efficiency 500% higher than silicon-based, with no heat accumulation.
The entire team was unprecedentedly energized, because at least to this point, this path is feasible.
From the material path perspective, this can surpass silicon-based.
On Earth, we can’t surpass Nvidia in the short term, so we look to the stars.
When the team’s morale was boosted, Lin Ran reminded: “This is just Earth testing; Moon’s microgravity will affect thin film stress; we need to simulate vacuum outgassing.”
Sixth month, the team conducted final verification in the vacuum simulation chamber.
Experiment personnel put on gloves, carefully placing the sample into the test fixture.
All members held their breath, some waiting outside the laboratory, some in the office: this is the final step; if it passes, it can be sent to the Moon.
“Start the simulation!” Lin Ran commanded.
Chamber evacuated to 10^-7 Torr, temperature cooled to 100K via radiation, simulating lunar radiation with proton beam bombardment at 10^10 particles/cm²/s.
Chip connected to AI test circuitry: input a convolutional neural network model, processing simulated lunar image data.
Screen shows resistance remains zero, computation error rate <0.1%, Jc drops only 5% under radiation.
“General Manager Lin, it’s stable!
FeSe’s interface superconductivity perfectly maintained in vacuum, shielding layer absorbs 80% radiation, Cooper pairs undamaged.”
Experiment personnel shouted excitedly.
From failure to success, it only took half a year.
This lightning speed would make anyone proud.
Moreover, what they made is a chip capable of artificial intelligence algorithm computation, countless times stronger than a basic demo.
Lin Ran smiled and applauded; this is truly original in meaning, not following anyone, a path no one has walked before.
A side note here: like the photovoltaic breakthrough, researchers need concentrated management, but Lin Ran doesn’t; his freedom is much higher, simultaneously responsible for research on many lines.
From perspectives of different lines, the abilities Lin Ran shows are different. In the photovoltaic sector, those young scholars feel it’s mathematical ability, a kind of omnipotent mathematical brute-force cracking ability; anything that can be modeled mathematically, Lin Ran can find an exact solution for, thanks to cracking the NS equations.
While in the superconducting chip line, Engineer Wu’s feeling is erudition; any related paper, Lin Ran has read, can list one two three four five, and what he says may work, and finally proves effective, leading the entire team to trust unconditionally, with advancement speed far exceeding expectations.
In the subsequent half year, testing on this chip sample continues, applying magnetic field up to 5T, Tc still holds at 105K, matching Ginzburg-Landau theory predicted upper critical field Hc2 ~ Tc^1.5.
Under temperature fluctuation ±5K, chip stable, no degradation.
Integrating water ice sublimation heat dissipation, under heat flux <1W/cm², the chip still runs stably.
Lin Ran has accounts on various social platforms; backend private messages are all sorts, from borrowing money to finding wife or mistresses one two three four five, to claiming to have conquered the strong Goldbach Conjecture—countless private messages.
Lin Ran doesn’t check private messages and rarely posts content.
Netizens tease that every post causes a stir.
This time as before, Lin Ran’s latest video quietly went online.
The video has no background music, exceptionally quiet.
Lens pulls back from Earth’s blue arc, through pitch-black space, straight to the Shackleton Crater at lunar south pole.
In the image, the permanently shadowed region is pitch black, nearby thermometer shows “100K”.
A chip visible to the naked eye appears in the lens.
Then lens switches to laboratory equipment, MBE chamber humming, iron atoms and selenium atoms evaporating, layer by layer depositing on SrTiO3 substrate, forming single-layer FeSe thin film.
Image flashes through R&D process, with images but no parameters; professionals can at most get a rough idea.
Test footage: in vacuum simulation chamber, temperature down to 100K, 1MeV proton beam bombardment; data flowing on display.
Image switches to rocket launch, chip landing at Shackleton Crater with lunar rover.
After posting the video, Lin Ran followed with a post: “Superconducting chip experiment effect perfect, about to head to the Moon.”
Then another post: “Forgot to mention, this Moon mission is rich, total three astronaut slots; we plan to open one. Anyone interested? I will personally go up; those who want to go to the Moon with me to witness the miracle can contact us.
PS: Fee is not cheap, proceed according to your means.”
