Vietnam Defense and News forum

[Yellow is Okay]

The holders and

Vietnam edge closer to the #AFCU23 Final!
Who will join them and complete the semi-finalists cast?

Interesting. Definitely pulling for China and South Korea today. Hoping the semifinals are pure rivalry, hatred, and honor, the full set.

 

[Viet]

Rumours are saying Viettel has already picked out a mostly Japanese vendor lineup for the fab. TEL for deposition, clean, etch, and probe, plus Nikon for lithography. And node-wise, regardless of whether the scanner ends up being Nikon or ASML, the key is the litho class: if it's 193i immersion, you can still push down toward 7nm-class with a ton of multiple patterning. If it's DUV dry like ArF dry or KrF, that's not a realistic route for 7nm-class patterning.

And people also forget Huawei has been accused for years of ties to the Chinese state and military, while Viettel is literally state-owned and even has a military arm. That's why the real question isn't just "who supplies the tools" but whether they can actually move the litho capability down toward more modern nodes, and whether any of that output can be exported competitively without running into restrictions, excluding yield capability.

Making a fab with wafer production, semiconductor, custom chips in 32nm. That’s a big leap for Viettel and Vietnam in general, no matter it’s a private or a state or military company.
Globally there are 3 countries that can make lithography machines. China, Japan and the Netherlands. The Chinese will never deliver. Yes very possible the lithography comes from Japan. But unlikely. So that’s why I assumed it’s ASML.

 

[Yellow is Okay]

Making a fab with wafer production, semiconductor, custom chips in 32nm. That’s a big leap for Viettel and Vietnam in general, no matter it’s a private or a state or military company.

If you look at the market share upstream for silicon wafers, a lot of the semiconductor powerhouses still import wafers/materials from a small set of qualified suppliers. Nobody's going to give you an easy ride on tech transfer like that, because the real moat isn't just the tools, it's the process integration know-how, yield learning, and the whole ecosystem. And 32nm-class basically doesn't make much sense as an export play, know why? It's a mature-node, commodity-priced game, and China can crank out 28nm and above at scale the way Vietnam produces shoes: high volume, cost-optimized, hard to beat on margin.

 

[Viet]

If you look at the market share upstream for silicon wafers, a lot of the semiconductor powerhouses still import wafers/materials from a small set of qualified suppliers. Nobody's going to give you an easy ride on tech transfer like that, because the real moat isn't just the tools, it's the process integration know-how, yield learning, and the whole ecosystem. And 32nm-class basically doesn't make much sense as an export play, know why? It's a mature-node, commodity-priced game, and China can crank out 28nm and above at scale the way Vietnam produces shoes: high volume, cost-optimized, hard to beat on margin.

You have to start from somewhere. As beginner everything is difficult. Lesser jumping into making let say 10nm chips.
Not to forget Making semiconductor is one of the most difficult things ever. only few countries in the world can make it. Very expensive, very difficult. I read the process requires 1,000 steps to complete a chip. Just 1 error in the process then the chip becomes a plastic waste.

You may be right I have to correct myself.
Viettel most likely buys 32nm lithography machines from Japan (Nikon or Canon). The US likely will not give its approval yet to sell ASML machines to Vietnam.

A comment from FPT chairman on Viettel making chips. His company most likely will be the second company to construct fabs.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

 

[ChineseTiger1986]

Making a fab with wafer production, semiconductor, custom chips in 32nm. That’s a big leap for Viettel and Vietnam in general, no matter it’s a private or a state or military company.
Globally there are 3 countries that can make lithography machines. China, Japan and the Netherlands. The Chinese will never deliver. Yes very possible the lithography comes from Japan. But unlikely. So that’s why I assumed it’s ASML.

China wants to compete with ASML and Nikon/Canon for the market share, and if Vietnam has the needs for DUV lithography machine, then why not?

 

[Viet]

China wants to compete with ASML and Nikon/Canon for the market share, and if Vietnam has the needs for DUV lithography machine, then why not?

I think not realistic. Possibly Vietnam will buy from China, BUT there are political concerns AND
can SMEE deliver? Will Ccp allow it?
Considering COMAC struggles to produce aircraft. very doubtful SMEE can deliver lithography to other countries.

 

[Yellow is Okay]

You have to start from somewhere. As beginner everything is difficult. Lesser jumping into making let say 10nm chips.
Not to forget Making semiconductor is one of the most difficult things ever. only few countries in the world can make it. Very expensive, very difficult. I read the process requires 1,000 steps to complete a chip. Just 1 error in the process then the chip becomes a plastic waste.

You may be right I have to correct myself.
Viettel most likely buys 32nm lithography machines from Japan (Nikon or Canon). The US likely will not give its approval yet to sell ASML machines to Vietnam.

A comment from FPT chairman on Viettel making chips. His company most likely will be the second company to construct fabs.

To view this content we will need your consent to set third party cookies.
For more detailed information, see our cookies page.

Accept third party cookies

I never said you shouldn't start from somewhere. And the whole "Nikon vs ASML" lithography thing is still rumor-level anyway. What actually matters is the litho class, because that basically sets the ceiling on what your process can realistically do and, by extension, what business model even makes sense.

If Vietnam can legally and practically move toward more modern nodes with real manufacturing scale, real yield learning, and a proper integration flow, then sure, that's a great trajectory. But if the constraint forces you to park at 32nm, the economics get ugly fast. That's a mature-node game with commodity pricing and brutal margin pressure, and you're competing against suppliers who can run 28nm+ at massive scale with fully amortized toolsets and supply chains.

It's like VinFast buying an old BMW engine line. You can say you have got "the tech", but you're still inheriting last-cycle capability while the seller walks away with a nice check. BMW and the German state probably laughed all the way to the bank on that one.

 

[Viet]

I never said you shouldn't start from somewhere. And the whole "Nikon vs ASML" lithography thing is still rumor-level anyway. What actually matters is the litho class, because that basically sets the ceiling on what your process can realistically do and, by extension, what business model even makes sense.

If Vietnam can legally and practically move toward more modern nodes with real manufacturing scale, real yield learning, and a proper integration flow, then sure, that's a great trajectory. But if the constraint forces you to park at 32nm, the economics get ugly fast. That's a mature-node game with commodity pricing and brutal margin pressure, and you're competing against suppliers who can run 28nm+ at massive scale with fully amortized toolsets and supply chains.

It's like VinFast buying an old BMW engine line. You can say you have got "the tech", but you're still inheriting last-cycle capability while the seller walks away with a nice check. BMW and the German state probably laughed all the way to the bank on that one.

Viettel making fabs is similar to Vinfast making cars, or Thaco making subways. It’s not driven by money. It’s not economics. At least the primary motive. The companies would not make profits for years.

It’s purely driven by politics. All about self sufficiency, making stuffs in Vietnam.

 

[Yellow is Okay]

Viettel making fabs is similar to Vinfast making cars, or Thaco making subways. It’s not driven by money. It’s not economics. At least the primary motive. The companies would not make profits for years.

It’s purely driven by politics. All about self sufficiency, making stuffs in Vietnam.

Vietnam doesn't really get a say, or have the leverage to dictate semiconductor supply chains, so your definition of "self-sufficiency" is a stretch. But yes, at least now you're starting to sound more realistic.

 

[Viet]

Vietnam doesn't really get a say, or have the leverage to dictate semiconductor supply chains, so your definition of "self-sufficiency" is a stretch. But yes, at least now you're starting to sound more realistic.

Not dictate or anything. Take carmaker Kim Long Motor, it buys diesel engine tech from China then produces the engines in its own factories. Sufficiency here means the company no longer needs to import the engines from China.

 

[Yellow is Okay]

Not dictate or anything. Take carmaker Kim Long Motor, it buys diesel engine tech from China then produces the engines in its own factories. Sufficiency here means the company no longer needs to import the engines from China.

Kim Long and Yuchai had a tech-transfer agreement for the engine. In Viettel's case, Vietnam is mostly just buying equipment from semiconductor vendors, not securing process or specific integration know-how through a true tech transfer. Those are fundamentally different situations. And if anything, this route will bind Vietnam even more tightly to the Western, especially US-led, semiconductor ecosystem and supply chain.

 

[Viet]

Kim Long and Yuchai had a tech-transfer agreement for the engine. In Viettel's case, Vietnam is mostly just buying equipment from semiconductor vendors, not securing process or specific integration know-how through a true tech transfer. Those are fundamentally different situations. And if anything, this route will bind Vietnam even more tightly to the Western, especially US-led, semiconductor ecosystem and supply chain.

Why is bad if integrated into western led supply chain? Should we integrate into eastern led system instead?
We should make chips, not ideology.
I asked OpenAI “how Viettel builds the first fab?” Here the answer

“
Here’s a clear, step-by-step explanation of how Viettel is able to build the first semiconductor fab in Hanoi — including who helped, what makes it possible, and why it’s realistic even for a company without prior fabs:

1.


Government Support & Strategic Priority

Vietnam’s government has made semiconductor production a national priority because:

• Chips are vital for telecom, defense, and digital economy
• Vietnam wants to reduce reliance on imports
• The national strategy (Vietnam Semiconductor Industry Plan) encourages domestic production

Viettel’s fab is supported by a government directive and state planning, which helps with:

• Land allocation (Hoa Lac Hi-Tech Park)
• Licensing and regulatory approval
• State investment and incentives

Without this political backing, such a capital-intensive project wouldn’t be possible.


2.


Viettel Already Has Strong Capabilities in Electronics


Viettel isn’t a beginner — it’s one of Vietnam’s largest tech and telecom companies with experience in:

• Designing electronic products (routers, IoT, military communications)
• Complex manufacturing ecosystems
• Supply chain management

They have engineering talent, R&D units, and a long corporate history in tech.

So even though they haven’t built a wafer fab before, they do have:

• Engineering culture
• R&D experience
• Enterprise scale required for such a project

3.


Funding & Investment


Fab construction costs are huge — typically billions of dollars.

Viettel taps multiple funding sources:

• Internal capital
• Government financing and support
• Bank loans
• Strategic partnerships

Many countries subsidize fabs because they:

• Increase national security
• Create high-paying jobs
• Expand technology ecosystems

In Vietnam’s case, the state is treating this like critical infrastructure.


🛠 4.


Partnerships with Foreign Technology Providers

Viettel is not building every part of the fab themselves. They contract with experienced global semiconductor suppliers for key technologies:

Equipment & Tools

• Lithography → most likely from Japan (Nikon / Canon)
• Etchers, deposition tools → US, Europe, Japan
• Test & inspection equipment → global suppliers

Design & IP

• Viettel can license standard cell libraries, IP blocks, and design tools from companies like Cadence, Synopsys, Siemens EDA

Construction & Cleanroom Expertise

• International engineering firms help build the cleanroom and facilities.

So Viettel integrates global technology, rather than reinventing these systems.


5.


Targeting a Mature Production Node (~32 nm)


The first fab is not aiming for bleeding-edge chips like 5 nm or 3 nm — these require:

• EUV lithography (extremely expensive)
• Highly advanced ecosystem
• Massive R&D infrastructure

Instead, Viettel’s fab focuses on:


~32 nm process


Industrial, IoT, wireless, automotive, defense chips


Specialty chips, not high-volume mobile/AI CPUs

Approaching a mature node makes the project feasible because:

• Tools are cheaper
• Equipment is widely available
• Intellectual property requirements are more straightforward
• Yields (production quality) are easier to manage

6. Training, Workforce & Knowledge Transfer

Building a fab requires highly trained engineers. Vietnam, through:

• Universities
• Training programs
• International hires
• Technician training with partners

is building a workforce capable of design, operations, and fab maintenance.

This doesn’t happen overnight, but focused investment in people makes the project real.

7. Phased Implementation

The fab isn’t going from zero to fully automatic overnight. It will be done in phases:

Phase 1: Construction & Infrastructure

• Groundwork
• Cleanrooms
• Utilities

Phase 2: Tool installation & calibration

• Lithography
• Etchers
• Deposition tools

Phase 3: Pilot production

• Yield optimization
• Test wafer

Phase 4: Mass production

• Full capacity

This staged approach reduces technical risk.

8. Learning from Other Countries

Countries with new fabs (e.g., Malaysia, Philippines, Vietnam, India) typically start with:

✔ Mature nodes
✔ Foreign partnerships
✔ Export-oriented manufacturing
✔ Government incentives


Vietnam is adopting the same model.

Summary — Why It’s Possible

Viettel’s fab is succeeding because:

1. Government backing & national strategy
2. Viettel’s existing tech and engineering base
3. Capital investment & financing structures
4. Global equipment and design partnerships
5. Targeting an achievable process node (~32 nm)
6. Training local workforce
7. Phased implementation

If you want, I can also explain:

✔ The specific steps in a semiconductor fab build

✔ Why cutting-edge nodes require EUV (and why that’s hard)

✔ What types of chips Viettel will likely make first


Just tell me which one you want!

 

[Yellow is Okay]

Why is bad if integrated into western led supply chain? Should we integrate into eastern led system instead?
I asked OpenAI how Viettel builds the first fab. Here the answer

“
Here’s a clear, step-by-step explanation of how Viettel is able to build the first semiconductor fab in Hanoi — including who helped, what makes it possible, and why it’s realistic even for a company without prior fabs:

1.


Government Support & Strategic Priority

Vietnam’s government has made semiconductor production a national priority because:

• Chips are vital for telecom, defense, and digital economy
• Vietnam wants to reduce reliance on imports
• The national strategy (Vietnam Semiconductor Industry Plan) encourages domestic production

Viettel’s fab is supported by a government directive and state planning, which helps with:

• Land allocation (Hoa Lac Hi-Tech Park)
• Licensing and regulatory approval
• State investment and incentives

Without this political backing, such a capital-intensive project wouldn’t be possible.


2.


Viettel Already Has Strong Capabilities in Electronics


Viettel isn’t a beginner — it’s one of Vietnam’s largest tech and telecom companies with experience in:

• Designing electronic products (routers, IoT, military communications)
• Complex manufacturing ecosystems
• Supply chain management

They have engineering talent, R&D units, and a long corporate history in tech.

So even though they haven’t built a wafer fab before, they do have:

• Engineering culture
• R&D experience
• Enterprise scale required for such a project

3.


Funding & Investment


Fab construction costs are huge — typically billions of dollars.

Viettel taps multiple funding sources:

• Internal capital
• Government financing and support
• Bank loans
• Strategic partnerships

Many countries subsidize fabs because they:

• Increase national security
• Create high-paying jobs
• Expand technology ecosystems

In Vietnam’s case, the state is treating this like critical infrastructure.


🛠 4.


Partnerships with Foreign Technology Providers

Viettel is not building every part of the fab themselves. They contract with experienced global semiconductor suppliers for key technologies:

Equipment & Tools

• Lithography → most likely from Japan (Nikon / Canon)
• Etchers, deposition tools → US, Europe, Japan
• Test & inspection equipment → global suppliers

Design & IP

• Viettel can license standard cell libraries, IP blocks, and design tools from companies like Cadence, Synopsys, Siemens EDA

Construction & Cleanroom Expertise

• International engineering firms help build the cleanroom and facilities.

So Viettel integrates global technology, rather than reinventing these systems.


5.


Targeting a Mature Production Node (~32 nm)


The first fab is not aiming for bleeding-edge chips like 5 nm or 3 nm — these require:

• EUV lithography (extremely expensive)
• Highly advanced ecosystem
• Massive R&D infrastructure

Instead, Viettel’s fab focuses on:


~32 nm process


Industrial, IoT, wireless, automotive, defense chips


Specialty chips, not high-volume mobile/AI CPUs

Approaching a mature node makes the project feasible because:

• Tools are cheaper
• Equipment is widely available
• Intellectual property requirements are more straightforward
• Yields (production quality) are easier to manage

6. Training, Workforce & Knowledge Transfer

Building a fab requires highly trained engineers. Vietnam, through:

• Universities
• Training programs
• International hires
• Technician training with partners

is building a workforce capable of design, operations, and fab maintenance.

This doesn’t happen overnight, but focused investment in people makes the project real.

7. Phased Implementation

The fab isn’t going from zero to fully automatic overnight. It will be done in phases:

Phase 1: Construction & Infrastructure

• Groundwork
• Cleanrooms
• Utilities

Phase 2: Tool installation & calibration

• Lithography
• Etchers
• Deposition tools

Phase 3: Pilot production

• Yield optimization
• Test wafer

Phase 4: Mass production

• Full capacity

This staged approach reduces technical risk.

8. Learning from Other Countries

Countries with new fabs (e.g., Malaysia, Philippines, Vietnam, India) typically start with:

✔ Mature nodes
✔ Foreign partnerships
✔ Export-oriented manufacturing
✔ Government incentives


Vietnam is adopting the same model.

Summary — Why It’s Possible

Viettel’s fab is succeeding because:

1. Government backing & national strategy
2. Viettel’s existing tech and engineering base
3. Capital investment & financing structures
4. Global equipment and design partnerships
5. Targeting an achievable process node (~32 nm)
6. Training local workforce
7. Phased implementation

If you want, I can also explain:

✔ The specific steps in a semiconductor fab build

✔ Why cutting-edge nodes require EUV (and why that’s hard)

✔ What types of chips Viettel will likely make first


Just tell me which one you want!

I never said it's bad. I'm saying your "self-sufficiency" claim is a stretch. And I'm not asking how Viettel will build its first fab either.

My point is whether they can legally move beyond that baseline. If they can't, or if U.S. policy shifts overnight, they risk ending up like VinFast buying an old BMW engine line: you "have the tech" on paper, but you're stuck with last-cycle capability while everyone else has already moved on.

 

[Viet]

I never said it's bad. I'm saying your "self-sufficiency" claim is a stretch. And I'm not asking how Viettel will build its first fab either.

My point is whether they can legally move beyond that baseline. If they can't, or if U.S. policy shifts overnight, they risk ending up like VinFast buying an old BMW engine line: you "have the tech" on paper, but you're stuck with last-cycle capability while everyone else has already moved on.

So what’s the alternative?
Vinfast not buying old BMW tech? Viettel not buying old stuff from Nikon? Nothing? Wait and see until the sky turns green? Germany and Japan will never sell latest technology to Vietnam.
Self sufficiency does not mean you buy nothing more. nobody lives in a vacuum.

 

[Yellow is Okay]

So what’s the alternative? Vinfast not buying BMW tech? Viettel not buying from Nikon?

VinFast already ditched their entire gasoline lineup. And on semiconductors, even within SEA, Singapore shut down its own fab back in the 2000s, and Malaysia, with far more experience and a deeper industrial base than Vietnam, still hasn't seriously gone down the path of building their own domestic fab.

So I don't see any guarantee Viettel can LEGALLY move past 32nm, given how export controls and tool access actually work. And if you're stuck at 32nm, "competing" is basically a knife fight in a mature-node, commodity-priced market that only makes sense if you don't care at all about profits or running costs.