In December 2025 people started talking about a rumor. The rumor was that ASUS might start making memory for computers. They might do this early as the second part of 2026. The reason for this is because there is not memory for computers in the world.

A lot of websites that talk about technology said this was going to happen.. It turns out that this information came from people who are not directly related to ASUS. ASUS said that this is not true. They do not plan to make their memory, for computers. Instead ASUS will work closely with the companies that already make memory for computers. People who watch the market are saying the thing: it is getting harder to get memory because companies that make memory are using most of their ability to make things for artificial intelligence and data centers. This means that the price of memory has gone up a lot. This is a problem for companies that make computers and smartphones. The story is talking about an issue with memory supply and a rumor that it is unlikely that ASUS will start making all of their own memory called DRAM in their own factories by 2026. ASUS is having a time, with memory supply just like other computer and smartphone makers. That said, ASUS could pursue more feasible forms of vertical integration (module assembly, partnerships, co-investment or buying up module-level supply) that would help reduce exposure to short-term RAM shortages.

1) So I want to know where this rumor actually started. Let us look at the timeline and the people who reported it. Where did the rumor come from is a question. We have to look at the timeline of the rumor and the reporting of the rumor to understand what really happened with the rumor.

In late December 2025 several outlets including Wccftech, Gadgets360 and other tech sites published stories suggesting ASUS might enter the DRAM business and establish dedicated DRAM production lines by the end of Q2 2026. These pieces typically referenced a leak or non-English sources and suggested the move would be a defensive response to tightening memory markets driven by AI/data-centre demand.

Almost immediately, other outlets picked up the story and some added context or local-language sourcing. But within a few days a stronger counterpoint emerged: reporting that ASUS itself had told Taiwan’s Central News Agency (CNA) that it had “no plans to invest in a memory wafer fab” and that the company would instead deepen partnerships with existing memory suppliers and optimize product lifecycles. Outlets such as Tom’s Hardware and Indian Express published updates that pushed back on the idea ASUS would build full wafer fabs.

So: the rumor -> rapid amplification -> company denial. That pattern is common in supply-chain panic moments.

2) Why the rumor seemed believable is because of the market context

The background makes the rumor seem believable. The rumor is more likely to be believed because of the background. The background and the rumor just seem to fit in a way that makes the rumor intuitively plausible or, in other words the rumor seems true because of the background.

• Memory shortages are real and widespread in 2025–2026. Demand for DRAM and other memory types accelerated as cloud providers and hyperscalers poured capital into AI training and inference infrastructure; that redirected capacity towards high-margin HBM and server DRAM and away from commodity PC/smartphone DRAM. Analysts (IDC, Digitimes and others) reported supply tightness and price hikes that materially affect OEMs.

• Prices rose sharply. Several industry trackers reported large year-over-year increases in DRAM module pricing in late-2025; with foundry priorities skewed toward cutting-edge AI chips and HBM, commodity DRAM suppliers have been under pressure to allocate capacity to the most profitable segments. This creates real short-term insecurity for companies that rely on third-party suppliers.

• Companies that make things have done this before. When they have trouble getting the things they need to make their products they think about getting control over how they get these things. They might invest in companies that make these things make deals with them or even start making these things themselves. This has happened before so it is possible that a company that makes things might want to make its memory supply chain.

Given the situation it is easy to understand why people believed the rumor that ASUS might start making memory. Companies like ASUS that make a lot of computer parts have a problem when the price of memory goes up and it is hard to get. This is because memory is an important part that they need to make their products. When the price of memory increases it affects companies like ASUS that make a lot of hardware. The rumor about ASUS making memory got a lot of attention because people know that companies, like ASUS do not like it when the price of memory goes up.

3) What “making RAM” actually means — multiple levels of complexity

The phrase “make RAM” can mean different things to people. It is really important to understand what “make RAM” means in order to figure out if something is possible or not, with “make RAM”.

When people think about making RAM they usually think about Wafer-level DRAM fabrication, which is also known as fabs. This is the process of building factories to make silicon wafers that produce DRAM chips also called DRAM dies. To do this you need to use things like lithography and process engineering. It also requires a lot of money, which is called capex. It takes a long time to get everything ready. We are talking about years of design and qualification. It costs billions of dollars.

There are a few big companies in the world that make most of the DRAM wafers like Samsung, SK Hynix and Micron. They have factories just for this. If someone new wants to get into this business they need to have a lot of money and know a lot, about the process. Wafer-level DRAM fabrication is a deal and it is hard to get into. Building a new DRAM wafer fab and getting to volume production typically takes multiple years and tens of billions (including ecosystem support). For a consumer-electronics OEM to do this inside ~6 months (the rumor’s Q2 2026 target) is essentially impossible.

Module assembly, which includes memory modules, sticks or packaged modules is a basic step. This is where companies buy the actual memory parts from the places that make them and put them together into things called DIMMs or SODIMMs. They then test these things sort them out and give them a brand name. Sometimes they even add software or cooling systems. A lot of companies and memory vendors do business this way. They work with memory modules. That is what they focus on these memory modules. It is a lot easier for a company that makes things like an Original Equipment Manufacturer to set up or add lines to put and test modules really fast in just a few months instead of years. They can then say these modules are theirs even though they still need to get the parts like the DRAM die from other people first. Some companies have done this they make the modules and then use their name on them so they can get what they need or make their own brand really fast. The Original Equipment Manufacturer can make their modules and sell them under their own name this way the Original Equipment Manufacturer has more control, over the Original Equipment Manufacturer brand.

When it comes to design contracts, joint ventures or co-investment an Original Equipment Manufacturer can work with foundries, memory designers or module makers. The Original Equipment Manufacturer does this to make sure they have the capacity they need through offtake agreements, equity stakes or joint design projects.

These are things that companies often do when there are problems with the supply of something. This helps to reduce the risk, for the Original Equipment Manufacturer without them having to pay the cost of building their own factories, which is a big expense.

Design Manufacturers and Electronic Manufacturing Services companies have partnerships and they store extra supplies. The companies that make the products can talk to them to get the things they need first buy a lot of things at the time or change the products they make to use different kinds of memory like using DDR4 memory when it is a good choice. This way they can get things done faster. It will cost them less money.

When the press says “ASUS will make RAM” what they probably mean is that ASUS will put together RAM modules or work, with companies to make RAM. They are not saying that ASUS is going to build a new factory to make RAM from scratch.

4) What ASUS. Why it is important to us. The statement from ASUS is something that we should think about because it can affect the things that ASUS does. ASUS is a company that makes a lot of things that people use every day. So when ASUS says something people listen because it can change the way we use the things that ASUS makes. The statement from ASUS is about the things that ASUS wants to do in the future. ASUS wants to make better things that will help people. This is why the statement from ASUS matters, to people who use the things that ASUS makes.

ASUS reportedly told Taiwan’s Central News Agency that it had no plans to invest in a memory wafer fab, and that the company would deepen cooperation with memory suppliers and adjust product strategies accordingly. That is an important corrective statement from the company — it denies the most dramatic interpretation (building wafer fabs) and points to more standard supplier management tactics. In short, ASUS appears to be disavowing the high-capex fab narrative while acknowledging the problem and promising supplier coordination.

Why does that matter? That is because only a small number of firms have the scale and the intellectual property and the capital to make money from running Dynamic Random Access Memory wafer factories. For a company, like ASUS that makes equipment it usually makes sense to partner with companies or make deals to buy products from them or put together modules or work on designing products. This is cheaper and faster for ASUS and it is what ASUS said they would do.

5) Technical & economic barriers to ASUS building DRAM fabs quickly

ASUS probably does not want to do this. Even if ASUS did want to there are a lot of problems that would get in the way of ASUS doing it.

• Building a place that makes computer chips like DRAM is really expensive. It costs billions of dollars to get started and a lot of money each year to keep up with the technology. New companies that want to make these chips usually need help from the government or a big partner with a lot of money like billions of dollars to be successful, with DRAM.

• Making something like this is very complicated and needs knowledge: the technology to make DRAM getting more of them to work properly and the brain of the operation the controller, are all very specialized. To make good DRAM wafers that work well you need a team of experts who know a lot, about the process and they have to keep making small changes over many years. The DRAM process technology is a part of this. The DRAM yield ramping is also important. The DRAM controller IP is very specialized too.

• Supply chain and foundry capacity: It takes a lot of time to build the things that support the supply chain and foundry capacity even when we spend money on them. We need to get materials and special equipment like EUV or immersion steppers for advanced nodes and also set up lines, for testing and packaging. The supply chain and foundry capacity are very important. It is not easy to get everything ready quickly.

• The time it takes to go from planning to making a lot of something is usually several years, not just a few months. This is because the memory chips, called DRAM dies need to be really good and work well with the systems they are used in. Getting these memory chips to meet the standards of the companies that make the equipment and making sure they work with the systems is not easy. The memory chips have to be reliable and perform well. They also have to pass tests to make sure they are compatible, with the systems. This whole process takes a time and it is a big challenge to get everything right.

• Economies of scale: DRAM is a scale business; the major players have massive volumes that make their cost per bit much lower. A newcomer would face price competition with incumbents unless targeting niche, high-margin segments (which reduces the immediate supply benefit to mainstream PC SKUs). Several industry trackers emphasize that new capacity isn’t likely to appear quickly enough to fix 2026 tightness.

Building factories that make computer chips, known as fabs is really expensive. It takes a long time. It is also very tricky to do this kind of work. This is why it makes sense when the company says they do not have any plans to build a wafer fab. The company is being honest about their plans for a wafer fab because building a wafer fab is not something that companies usually do without a lot of thought. The companys statement about their plans, for a wafer fab is believable because it is based on what’s really happening in the industry. Building a wafer fab is just not something that the company wants to do now.

6) Things that ASUS can really do instead (. What to look out for)

If the main goal is to protect the product roadmaps of ASUS, from memory shortages then there are ways that ASUS can do this. These ways are more realistic. They are also faster and cost less money. ASUS talked about some of these ideas when they said they would work closely with the people who supply them with memory.

ASUS needs to figure out how to get module capacity. One way to do this is to buy a company that makes modules or invest in one. These companies put together DIMMs and SODIMMs. If ASUS does this it can get the modules it needs fast much faster than if it were to build its own factory. This can all happen in a few months or less, than a year if the right company is out there and it is not too hard to deal with all the regulations. ASUS could. Invest in companies that assemble DIMMs and SODIMMs which would help the company secure a steady supply of finished modules.

When we talk about long-term supply agreements this is what we mean. We are talking about things like prepayment and offtake. So what does this really do for us? It helps us make sure we have the memory we need. We do this by making deals with companies that make memory like Samsung, SK Hynix and Micron. We can also make deals with the companies that put the memory together which are called module houses.

These deals are like contracts that last for years. They help us get the memory we need even when it is hard to find. The bad thing is that we have to pay more for it.. At least we know we will have the memory we need. This is important for us because we do not want to run out of memory. We want to make sure we have enough so we can keep doing what we do. Long-term supply agreements, like prepayment and offtake are important for this. They help us get the memory we need from companies, like Samsung, SK Hynix and Micron.

Strategic partnerships / joint ventures: Co-invest in manufacturing capacity at the packaging, assembly, test (OSAT) layers or in module assembly plants with partners in Taiwan, Vietnam, or elsewhere.

SKU engineering: Redesign product SKUs to use more available memory types (e.g., pushing DDR4 variants where acceptable), tuning memory configurations across product lines to reduce peak demand.

Stockpiling and demand smoothing: Increase inventory on critical SKUs or smooth production schedules to reduce peaks.

When it comes to labelling of modules one thing we can do is buy DRAM dies and then outsource the assembly to companies that already make modules. After that we can sell these modules under the name. This way we have control over the brand. We still have to rely on the supply of DRAM dies from private labelling of modules. Private labelling of modules is important because it allows us to use the branding, on these modules.

Buying smaller memory vendors: Acquire smaller or regional memory makers (module-level) to expand control.

If you want to know that ASUS is really going down these roads you should look out for them saying they are working with companies or that they are putting money into smaller companies that make parts or that they are buying these companies or that they have big agreements to buy a lot of products from someone else. These things happen quickly. They make sense because they do not cost as much as building big factories to make computer chips. ASUS is likely to choose these options because they are faster and more believable, than building their wafer factories, which can be very expensive and time consuming so you should watch for these signs that ASUS is pursuing these paths.

7) What would happen to the market if ASUS or other companies that make computers started to make memory modules

If ASUS and other companies that make computers were to get into making memory at the module level it would be a change.

The market for memory modules would be very different if ASUS and other computer makers were selling them.

This would affect a lot of people who buy and sell memory modules, including companies that already make them.

There would be a lot of competition, between ASUS and other computer makers and the companies that already make memory modules.

ASUS and other computer makers getting into the memory module business would change the market in ways.

It would be interesting to see what happens to the market if ASUS and other companies that make computers start making memory modules.

If big companies, like ASUS make sure they have a supply of secure modules or build their own modules then some things that could happen soon are:

• Greater OEM control over module design and binning — allowing better integration and possibly OEM-specific testing or firmware optimizations for reliability or gaming overclocks.

• This will give ASUS a break with making their own things but it does not mean the whole market will get better. If one big company buys a lot of modules and keeps them it will not make more modules available, in the world it will just change who has them the total number of modules that ASUS and other companies can use will still be the same.

• When a company buys a lot of memory and stores it this can increase their costs in the term. It also means they need money to run their business. The company that makes the products can either pass these costs on to the people who buy their products or they can absorb the costs themselves which means their profits will be lower. This is related to the price of the memory and the margin of profit that the company makes on their products the memory price and margin effects.

• Downstream competition: If the companies that make computers start selling their memory modules in stores it will be harder for other memory brands like Corsair and G.Skill and Kingston to sell their products.. A lot of people still like to buy memory from these other brands instead of the company that made their computer. So it will not be a problem for these brands for a long time. People, like Corsair and G.Skill and Kingston because they are brands.

• This is a thing for companies that make modules, like Adata and TeamGroup and the companies that work with Kingston because they might see that people want more of their products and they can make them faster, than the people who make the tiny parts that go into these products so they can increase how many they make.

So module-level steps can help make an Original Equipment Manufacturer or an OEM, a little safer but they cannot suddenly make dies all over the world. That is still the problem. The main issue is that we cannot make dies quickly and that is what is holding everything back the global die production is the real bottleneck, for an OEM.

8) Broader implications for the PC, smartphone, and AI ecosystems

For PC and smartphone OEMs: the short-term memory crunch forces SKU trade-offs (lower RAM options on entry models, delayed high-RAM SKUs), price increases, and potential product delays. Analysts warned of shipment declines and higher consumer prices into 2026.

For consumers: less choice at given price points, higher prices for high-memory configurations, and possibly a slower refresh cycle for older devices.

For the intelligence and cloud sector the big companies that provide a lot of computing power are taking up most of the advanced memory and high bandwidth memory that is available. This means that the companies that make memory want to focus on selling to these companies because they can make a lot of money from them. At the time the companies that make equipment for artificial intelligence and cloud sector have to find new ways to use memory, such, as compressing data or using different types of systems because the artificial intelligence and cloud sector companies are using up so much of the available memory and high bandwidth memory. The artificial intelligence and cloud sector is driving this change and the memory producers and equipment makers are responding to the demands of the intelligence and cloud sector.

For the memory industry price surges can be good for suppliers because they can make money in the short term. This might also make companies that are already in the business want to spend money on new projects. However building factories that make wafers takes a long time. So it is likely that the people who already make a lot of memory the producers will be the ones to help balance the supply of memory by 2027 or 2028 rather than new companies the OEM entrants that want to get into the memory business. The memory industry will have to wait for these large producers to expand what they do. The memory industry is important. These large producers of memory are the key to making sure there is enough memory, for everyone.

9) Probability assessment — short, medium, long term

• Short term (by Q2 2026): Low probability that ASUS will build a new DRAM wafer fab and start producing chips. Company denial plus practical realities make that outcome highly unlikely. More plausible are module-level actions or contracts.

• In the few years say one to three years it is possible that ASUS will work with other companies. This could be by partnering with them investing money in them or making more of their parts. It seems likely that ASUS might buy a part of companies that make parts or sign long-term deals with companies that produce computer memory like DRAM. ASUS is trying to make sure they have a supply of the things they need. ASUS will probably make some smart moves, like investing in module makers or signing contracts with DRAM producers to get what they need.

• When we think about the term like three to seven years it seems unlikely that we will see a lot of progress in making things work together at a very basic level without some big partners and a lot of money being spent.. If we were to team up with a company that makes computer chips or a big memory company that could make things very different. In the past new companies that try to make wafer DRAM have had a tough time.

10) How to read future developments (what to watch)

If you want to see if ASUS really does something, about the issues of just saying things to the press you should look out for:

Official ASUS filings or press releases announcing investments, M&A (especially purchases of module makers, OSATs or assembly houses), or joint ventures.

If a company has contracts with Samsung, SK Hynix, Micron or major OSATs that would be a good way for them to make sure they have a steady supply of what they need. These contracts with Samsung, SK Hynix, Micron or major OSATs are important because they help the company get the things they need. Having these public contracts, with Samsung, SK Hynix, Micron or major OSATs is a thing to do.

When I look at job postings and permits to build factories I see things about putting together modules, packaging and testing lines in areas like Taiwan, Vietnam and China. This is a way to figure out what people are really planning to do with module assembly, packaging and test lines, in these regions.

Financial reports describing increased capex directed at manufacturing or inventory purchase.

ASUS is doing something with module branding. They are coming out with RAM modules that have the name on them.. They might work with other companies to make memory products that have both their names on them. This is a sign that ASUS wants to control more of the module business. This means they want to be in charge of making the RAM modules from start to finish, which is called module-level verticalization. ASUS module branding is what this is, about.

11) Bottom line — a realistic verdict

The news that ASUS might start making its RAM in 2026 is a mix of real problems and rumors that are not true. The fact is that it is really hard to get DRAM now because a lot of companies need it for artificial intelligence. This is a deal for ASUS.. When ASUS says it does not plan to make its own RAM that makes sense. Building factories to make RAM is a very complicated and expensive process that takes a long time. ASUS knows this. That is why it says it is not going to do it. The company is being honest about its plans, for RAM. ASUS is not going to build its RAM factories anytime soon. The most likely path for ASUS — and the one that would actually help in the short term — is module-level actions, supplier contracts, partnerships, or targeted investments in packaging/assembly/test capacity rather than building its own wafer fabs in 2026.

12) Quick recommendations for stakeholders

• For ASUS I think they should try to make term deals with other companies work with more partners to put together computer parts think about buying some companies that make parts for them and look into ways to design their products so they do not need as much memory when they are, at their busiest this way ASUS can manage their memory needs better.

• For people who buy things: the price of computers with a lot of RAM memory is going to go up in 2026. So if you need a computer with RAM memory for your job or for playing games you should think about buying it now instead of waiting.. You can wait until 2027 or 2028 when the price of RAM memory might go back down, to normal.

• For people who invest in companies or analyze them it is an idea to pay attention to when big companies that make memory like Samsung, SK Hynix and Micron say they are going to spend money on new projects. You should also watch for announcements from companies about the deals they make to get the things they need. The Samsung, SK Hynix and Micron companies will have an impact on when it becomes easier to get the things that are, in short supply.