Elon Musk "slaps in the face" to Zeng Yiqun: 4680 has been successfully mass-produced, and MegaWatt Flash Charging is here

Elon Musk is still too all-round!

Long-term leadership in autonomous driving alone was already something—but now electrification has surged back into the top tier:

A North American regulatory document unexpectedly disclosed Tesla’s latest battery technology, directly revealing the secret:

Tesla’s version of megawatt fast charging has already entered mass production.

Still the latest second-generation 4680 batteries….

Back then, when he spoke face to face with Musk and insisted that the 4680 battery wouldn’t work, the “battery king” Zeng Yuqun was “slapped in the face” by old man Ma’s response—using technology and mass production.

Tesla’s megawatt fast charging unexpectedly exposed

The news source is a public document submitted to the California Air Resources Board (CARB).

Originally, the document was meant to disclose the battery capacity information of Tesla’s freshly mass-produced second-generation heavy truck, the Semitruck—but on closer inspection, Tesla’s latest battery technical parameters were included as well:

Tesla Semitruck comes in two battery versions:

• Long-range version: usable battery capacity of 822 kWh, estimated range of 500 miles (about 805 kilometers), peak power of 800 kW, supports 1.2 MW supercharging

• Standard-range version: usable battery capacity of 548 kWh, estimated range of 325 miles (about 523 kilometers), peak power of 525 kW, also supports 1.2 MW supercharging

For reference, the long-range all-wheel-drive battery capacity of Model 3 and Model Y is about 75-80 kWh—meaning the Semi’s long-range battery pack energy is roughly 10 times that of passenger vehicles.

But the curb weight of Model 3 and Model Y is still less than 1/20 of the Semitruck.

Behind this are the combined effects of aerodynamic optimization, efficiency curves of the three-motor system, and production design weight-reduction measures—so in terms of energy use, the Semi’s measured energy consumption is about 1.7 kWh/mile. That corresponds to roughly 0.6 miles per kWh, significantly better than the 0.4-0.5 miles/kWh that is common in the electric heavy-truck industry today.

In charging efficiency, peak power can reach 1.2 MW (i.e., 1200 kW)—Tesla’s version of megawatt fast charging.

Using the 822 kWh long-range battery as an example, with 1.2 MW peak power, it can theoretically replenish about 60% of the charge within 30 minutes—exactly the duration required by U.S. transportation regulations for drivers’ mandatory rest.

That means the Semi’s charging time perfectly matches the legal rest time. While the driver stops to rest, the vehicle completes recharging, without consuming extra operation time.

These points show that the Semi’s battery system is not a simple scale-up of a passenger-car solution—it is engineered specifically around the real operating conditions of three categories of Class 8 heavy trucks.

Comparing horizontally with Tesla’s own battery product line, the progress embodied in the Semi-mounted solution is quite obvious.

The first-generation 4680 battery (used in the Model Y produced at the Texas factory) has an energy density of 244 Wh/kg, with peak charging power of about 250 kW—corresponding to the V3 supercharger. The second-generation 4680 battery—Cybercell—raises energy density to 272 Wh/kg, an 11.5% increase. Supported charging power jumps to 1200 kW, corresponding to the V4 supercharger and megawatt charging stations.

When compared with industry competitors, BYD’s second-generation Blade Battery has been in mass production since 2025, used in models such as the Han L. Its system energy density is about 190 Wh/kg (cell-level data differs), and it supports peak charging of about 1500 kW using a dual-gun supercharging setup.

CATL’s newly released fourth-generation Shenxing battery claims a nominal energy density of about 260-280 Wh/kg, and also claims peak charging power of up to 1200 kW. But the key difference is that the fourth-generation Shenxing is expected to enter mass production only at the end of 2026, and is still in production-line commissioning—meaning it is essentially a future product.

The conclusion is clear: after years of dormancy, Tesla’s three-electrode technology has returned again, now competing in the same top tier as BYD, and in absolute numbers Tesla is somewhat more conservative than BYD.

What matters is that Tesla’s megawatt fast charging and the second-generation 4680 are not lab prototypes, nor are they future contracts announced at a press conference. They are already mass-production technologies installed in the Cybertruck and Semitruck.

From this perspective, Tesla is about one year ahead of CATL.

After all, CATL’s just-released third-generation Shenxing is still a future product, with mass production not expected until late 2026.

In a sense, this is also the strongest rebuttal after Musk once “taught Zeng Yuqun how to do things” face-to-face.

How did Tesla do it?

The breakthrough of the second-generation 4680 battery is not a single technical point, but the result of parallel progress across three fronts: physical design, electrochemical system, and manufacturing process.

A laboratory at the University of California, San Diego performed precise disassembly and electrochemical testing on the Cybercell, revealing the real source behind the performance leap.

First—and most directly—comes the “physical benefits” brought by casing weight reduction.

To ensure structural strength for the first-generation 4680’s 46 mm large-diameter cylindrical cells, the casing thickness was as high as 0.6 mm—an archetypal case of “over-engineering.” The second-generation cell directly thinned the casing to 0.35 mm, a reduction of about 42%.

This is extremely aggressive engineering. For 46 mm diameter batteries, the casing wall is only one-third of a millimeter thick, yet it must withstand winding stress and encapsulation pressure—pushing steel stamping processes to the limit.

But the gains are also very direct: thinning the casing frees up more internal space to load active materials, while the weight of non-active substances drops significantly. Just this one improvement contributes about a 20 Wh/kg increase in energy density.

In other words, without changing any chemical formulas, Tesla relied solely on improved manufacturing precision to achieve nearly a 10% performance gain.

But physical thinning alone is far from enough—the electrochemical system upgrade is the real core technical breakthrough.

The second-generation battery’s cathode material upgraded from the first-generation NMC 811 (nickel 81%, cobalt 12%, manganese 7%) to NMC 955 (nickel 91%, cobalt 5%, manganese 4%). Every 1 percentage point increase in nickel content yields a positive response in battery capacity, and 91% nickel content already reaches the scientific frontier of high-nickel cathodes in current mass production.

Meanwhile, lowering cobalt content to 5% reduces dependence on cobalt ore in places like the Congo and also spreads out material costs.

A key verification logic comes from changes in electrode thickness:

Measured results show the negative electrode thickness decreased only from 250 microns to 240 microns—a reduction of just 4%. But the positive electrode thickness dropped sharply from 180 microns to 150 microns—a reduction of 17%.

In lithium batteries, the lithium-ion capacity of the anode and cathode must be strictly matched. When the cathode thickness is substantially reduced yet it can still carry the same total lithium-ion amount, the only explanation is a qualitative leap in the active-material density of the cathode itself.

This chemical improvement additionally contributes about 10 Wh/kg to energy density. Combined, the two parts perfectly explain the jump from Tesla’s second-generation 4670 battery energy density of 244 to 272 (Wh/kg).

Beyond energy density, megawatt fast charging also depends on the entire battery pack structure and process innovation.

The biggest enemy of high-power charging is heat generated by internal resistance. The second-generation 4680 battery made multiple mechanical-structure optimizations to reduce resistance.

First, the biggest difference from the first-generation 4680 is that the anode copper foil is directly welded to the bottom cover, eliminating the traditional interfacial current collector.

Second, the aluminum positive current collector changed from a slotted design to a solid disk, increasing the electronic-overflow area. And by significantly thinning the electrode, it greatly reduces ion diffusion resistance in the solid phase.

These three combined improvements significantly reduce heat generation during high-rate charging and discharging. This is why Cybertruck’s current charging speed is limited by software to a “mid-tier in the industry” level, but the hardware reserves far higher potential. Once V4 supercharging stations are unlocked, lower internal resistance will support more aggressive charging curves.

At the manufacturing process level, currently only the anode of the second-generation 4680 uses a dry process; the cathode still follows the traditional wet coating method.

Let’s explain the dry process here: it is a “revolution” rather than an “evolution” of traditional battery manufacturing. It removes the most energy-intensive and most expensive intermediate steps, fundamentally reshaping battery manufacturing cost and speed.

You can roughly understand the dry process as “directly compress dry powder into electrodes,” like pressing dried flour into shape, rather than using traditional methods that add water to mix into dough, then put it into an oven to dry.

Conservatively, the manufacturing cost can be reduced by about 30%. Overall cost can be reduced by 10%-20%. Production efficiency is 7 times that of the wet process.

A full dry process means the second-generation 4680 is still far from reaching the performance limits of current lithium batteries—and Tesla has more technical exploration in reserve.

For example, silicon-based anodes could push energy density to 300 Wh/kg and shorten charging time, expected to be introduced within 1-2 years; asymmetric lamination technology could simultaneously improve energy density and charging speed, conservatively adding another 35 Wh/kg; lithium doping technology could theoretically reach 330 Wh/kg….

Ternary lithium batteries approaching 400 Wh/kg are already comparable to entry-level semi-solid batteries. But in terms of cost, 4680 has an overwhelming advantage.

4680 is not a dead end. Instead, for a fairly long period in the future, it is set to be one of the main routes for iteration of power batteries.

Musk “slaps Zeng Yuqun in the face”

This line can be traced back to a conversation three years ago.

At that time, CATL chairman Zeng Yuqun, during an industry exchange, pointed out in front of Musk that the 4680 large cylindrical battery has inherent structural design flaws. The overly large diameter makes it difficult to dissipate heat at the core. It is hard to balance casing strength with internal resistance, and mass-production feasibility is doubtful.

Judging from subsequent progress, Zeng Yuqun’s assessment at the time was not without merit.

Because Tesla’s first-generation 4680 performance was indeed rather ordinary: energy density was only 244 Wh/kg, not meaningfully better than 2170, and charging performance did not meet expectations—plus the dry process couldn’t be broken through for a long time.

It is said that in response to Zeng Yuqun’s “reasoned explanation,” Musk went silent directly.

Later, many outside observers also believed that Tesla’s self-developed 4680 battery route was heading into a dead end.

But just like Tesla’s self-developed chips, there can be setbacks and stagnation—yet Musk will ultimately respond to doubts with technological breakthroughs.

For example, thinning the 4680 casing plus structural optimization solves internal resistance and heat dissipation issues. High-nickel cathodes paired with matching anodes fill the energy-density shortfall. Dry anode processing is already underway, and a full dry process is on the way…

Looking back today at Zeng Yuqun’s “assertion,” it can’t really be said to be wrong. Based on the first-generation product’s performance, that judgment was reasonable. But the timeline of technology and mass production has already left him behind.

Tesla used three years to prove that the so-called “inherent flaws” of the 4680 battery are not a technological dead end—they are engineering problems, and engineering problems mean they can be solved.

Tesla has been leading in autonomous driving. With multimodal large models and a data-driven system, it has now become consensus across the L2 and L4 autonomous driving tracks.

After years of stagnation in electrification, it has now made an astonishing comeback.

Moreover, new technology exploration is proving that existing lithium-battery routes have not yet reached the physical limits—energy density can still be pushed higher, charging speed can still be increased, and manufacturing costs can still be lowered.

Compared with all-solid-state batteries still in the “early stages,” the incremental improvement route Tesla chose is the key killer feature for cost dominance in the coming years and for widening the gap with competitors.

Zeng Yuqun criticized Musk face-to-face and doubted that the 4680 was impossible to make work. And now Musk has given the best response with technology and mass production.

On one hand, Tesla is indeed too all-round.

On the other hand, CATL has been lying around making money over the past few years, so it was too easy.

Source of this article: Smart Car Reference

Risk Warning and Disclaimer

        There are risks in the market; investment is to be done cautiously. This article does not constitute personal investment advice, and it does not consider any specific investment objectives, financial conditions, or needs of individual users. Users should consider whether any opinions, viewpoints, or conclusions in this article are suitable for their specific circumstances. Invest accordingly at your own risk.