memory

There is an unexpected victim of the rise in RAM memory prices: the very modern connected cars

by

Which what’s happening with the RAM memories is making one thing clear: the best time to buy memory modules is yesterday. The price increase is so extraordinary which is already affecting other classic components of our PCs such as SSD units or graphics cards. However, the crisis that these components are generating goes further. Much further. Data centers devour memory. The AI ​​fever, we already know very well, has generated a voracious hunger not only for cutting-edge AI chips, but also for RAM and HBM memories that accompany these chips. As indicated in The Wall Street Journaldata centers (both conventional and those dedicated to AI) will consume more than 70% of the high-end memory chips that manufacturers produce in 2026. And if they could take more, they would take them. This is not (only) about PCs or mobiles. It is evident that the first affected by this problem are conventional desktop and laptop computers, as well as our mobile devices. Hundreds of millions of them are sold every year and they all have a certain amount of RAM that is now more expensive than ever. The shock wave is already causing other components such as SSD drives or graphics cards affected, but in reality memory chips are everywhere. And above all, in one. From TV to car. The frenetic rise in memory prices is certainly going to affect other segments that we had not thought about soon. Of course it will do so on other consumer electronic devices, and this certainly includes Smart TVs, which They have their own processor, memory and storage to offer us its functions. But the problem may be even more critical for cars, which for years were already computers with wheels and which are now even better and more powerful computers (and with more memory) with wheels. Memories of all kinds. Although car electronic systems have traditionally used RAM, the latest in most cases was not needed. But that was in the cars of a few years ago, because the arrival especially of the electric car and the fever for screens in our vehicles has made these needs different. Now our cars need various types of memory, but in some cases those modules are as good (or better) than the ones we have in our cell phones and computers. The ECUs. A modern car makes use of so-called ECUs (Electronic Control Units) for issues such as controlling the transmission, the airbag system or the engine itself. It is normal for them to have between 50 and 150 of these control units or microcontrollers, and almost all of them contain RAM for temporary data and a ROM for firmware and software. Infotainment systems. The most obvious component that surely comes to mind as that “car computer” is the infotainment system, which usually consists of a touch screen, navigation functions, support for CarPlay and Android Auto systems, and voice assistants. Although in many cars these systems use 1 GB or 2 GB of DRAM memory, there are more modern cars that They reach 4 GB and even 8 GB of LPDDR4 memory. And if we talk about some manufacturers like BYD or NIO, there are models in which They use 16 GB of LPDDR5 memory. The Ford SYNC 5 system, for example, is based on a Qualcomm SoC with 16 GB of RAM. Driving assistance requires memory. In addition to these components, there are others that also require the use of RAM. Advanced driving assistance systems (ADAS) allow you to activate functions such as adaptive cruise control, lane keep assist, automatic emergency braking or parking assistant. And to achieve this they use RAM with high bandwidth, which allows working with real-time images and processing of sensor signals. Samsung knows this well and in fact manufactures modules specifically oriented to this market. Tesla’s well-known autopilot hardware, Hardware 4 (currently used) makes use of 16 GB of RAMFor example. Micron already warned. In December 2023 Micron already indicated that “a car needs more memory than a (space) rocket.” The firm, an absolute protagonist in the field of RAM memory module manufacturing, indicated how in 2023 the average vehicle used 90 GB between RAM and NAND, but in 2026 that figure was estimated to be 278 GB and would reach 2 TB in high-end vehicles. That was good news for it and other manufacturers, and even then it pointed to how “generative AI is transforming automotive.” What they probably didn’t realize is that this revolution was going to need many data centers, and those data centers were going to need a lot of memory. And this is where we are. In Xataka | “Not a phone, it’s a car”: Volkswagen believes that screens in cars are going too far

a -50°C sanctuary to save the memory of glaciers

by

The climate crisis What we are experiencing is not only threatening to redesign world maps with sea level risebut it is also erasing traces of the planet’s history. After confirming that 2025 was the third warmest year in historythe scientific community has completed a critical mission: inaugurate the Ice Memory Sanctuary in Antarcticaan underground library designed to preserve ice from mountain glaciers before they melt permanently. A real bunker. Today we have on the planet a seed bank to prepare ourselves in case there is a global catastrophe, and also data servers. And now we also have a large bench for ice, which logically requires extreme thermal stability. This sanctuary, which can be considered an authentic glacier cemetery, has been promoted by the Ice Memory Foundation and led by institutions such as the French CNRS and the Italian CNR. The location chosen could not be other than the Antarctic plateau itself, specifically the Concordia station. What is stored. Inside there is not simply “ice, but we find what scientists called “ice witnesses”. For science there is a fairly clear difference, since these glaciers are authentic hard drives that contain the thermal chemical history of our planet. And unfortunately it was being lost due to rising temperatures. With these ice cylinders it is possible to analyze the air that existed thousands of years ago or even at analyze the hydrogen and oxygen isotopes inside calculate the exact temperature it was in the past. Something that allows us to reconstruct global temperature graphs with a precision that tree rings or marine sediments do not always achieve. A disaster record. Bonus, this ice also acts as a filter that traps anything floating in the air. That is why we have already seen, for example, cvolcanic sand or dust from the Sahara which allows studying historical eruptions or the cycle of wind movement. Although technology logically has limitations, and in the future it is quite likely that these technological means will increase considerably. That is why the real objective is to leave this ice for the scientists of the future who will surely have many tools to continue extracting information from these blocks of ice that we cannot do today. Engineering behind the cold. Logically, ice cannot be at unstable temperatures, so the location at the Franco-Italian Concordia station is not a conventional building. It is a cave excavated directly under the snow, taking advantage of the extreme conditions of the white continent. Something that allows you to maintain a stable temperature at -50ºCwhich is also essential for storing the genetic material that may be inside. But unlike freezers in European laboratories, this sanctuary does not depend on the electrical grid or motors. If there is a blackout or energy crisis, the ice remains intact. That is why its design is perfect to last for centuries. There are already tenants. This sanctuary already has several members in its exhibition. Two ice cores have already been found that come from the Alps, specifically, one Col du Dôme block drilled in 2016 and from Gran Combin (Switzerland) extracted in 2025. Logically, the problem is in logistical transportation from Europe (or any location) to Antarctica. The samples traveled for 50 days on the research icebreaker Italian Laura Bassi from Trieste to Antarctica, completing the last leg by plane to the Concordia base. Something that logically is not easy at all. What’s next now. The Ice Memory Foundation plans to continue rescuing samples from at-risk glaciers in the Andes, Himalayas and Pamirs. The Concordia sanctuary is ready to receive the legacy of a world that, year after year, breaks temperature records and this is what has caused this project to move so rapidly today in order not to lose more glaciers that are melting. Images | Cassie Matias In Xataka | Eight months ago a robot disappeared under the ice of Antarctica. Today we have recovered it and it brings disturbing data

Everyone blames the manufacturers for the lack of memory. Micron says real bottleneck lies elsewhere

by

For months, memory shortage It has established itself in the technological debate as one of those phenomena that do not seem to need too many explanations. If RAM is missing and prices risethe immediate conclusion is that someone is privileging AI and leaving the consumer aside. That idea has resonated strongly, especially after visible decisions that have affected the domestic channel and have reinforced the feeling of abandonment. But when you get down to how memory is manufactured and kept stable today, the diagnosis becomes less obvious: the bottleneck doesn’t seem as obvious as it seems. A controversial decision. In this climate of widespread suspicion, Micron has become a preferred target, shared with other large manufacturers, but for a very specific and recent decision: the announcement of the end of Crucial consumer products. The company recently announced that will stop selling RAM memory and storage under that historic brand, with shipments expected through February 2026. For many users, that move was interpreted as a direct consumer recall just when memory is short. Micron justified that decision by noting that AI-driven growth in data centers has skyrocketed demand and that Crucial’s exit seeks to improve supply and support to its strategic customers in higher-growth segments. The market has changed size. From Micron’s perspective, the problem is not a renunciation of consumption, but an abrupt change in the scale of the market. Christopher Moore, vice president of marketing for the client and mobile business, He said in an interview with Wccftech that the company continues to have a relevant presence in PCs and mobile devices, while serving data centers. What has altered the balance is the growth of the data center business, driven by AI, which has gone from representing around 30% of the market to approaching, according to its figures, 50% or even 60%. That leap, he defends, has left the entire industry without sufficient margin. Variety also creates scarcity. For Micron, the bottleneck is not so much the lack of factories as how the existing ones are used. Moore explains that producing memory is not about making a single type of chip seamlessly, but rather about switching between multiple densities and configurations depending on what customers ask for. Each change, for example going from 12 GB to 16 GB modules or from 16 GB to 24 GB, forces lines to be readjusted and reduces the total output volume. In a context of skyrocketing demand, this variety, which was previously acceptable, becomes a direct brake on production. Micron’s new Idaho factory under construction Faced with the temptation to think that new factories will solve the problem, the manufacturer asks for patience. Moore explains that expanding memory capacity is not an immediate process, because it requires not only building facilities, but equipping them, validating them and certifying each product with customers. The company laid the first stone three years ago in its ID1 plant in IdahoUnited States, whose entry into operation is scheduled for mid-2027. Even so, it warns that there will be no significant impact on supply until the entire qualification process is complete, which it places in 2028. Crucial is gone, the channel is not. Moore assures that, although Crucial has disappeared from the consumer showcase, the company continues to provide memory to major PC and mobile device brands through channels less visible to the end user. This OEM channel, in which Micron supplies memory directly to integrators and manufacturers, concentrates a very relevant part of the market and ends up being incorporated into commercial designs and equipment. From their point of view, the consumer continues to receive Micron memory, even if it no longer does so under a recognizable label. With this panorama, the lack of memory ceases to be a problem of isolated decisions and is revealed as the result of several overlapping tensions. AI-driven demand for data centers that has changed the scale of the market, operational limits on production and long lead times to expand capacity explain why supply will remain tight for years. Micron places the relief horizon no earlier than 2028 and, until then, the consumer will live with fewer options and pressured prices. The bottleneck, the company insists, is not only in who buys the memory, but in how it is manufactured. Images | Micron In Xataka | The situation with RAM prices is so desperate that there are already those who build their own memory at home

The situation with RAM prices is so desperate that there are already those who build their own memory at home

by

The crisis we are experiencing with RAM memories and its exorbitant price is shaking the technology industry in multiple ways, precisely because they are components that found in the vast majority of devices that we use in our daily lives. Faced with the crazy prices, there are users who have not given up and have resorted to extreme solutions: building their own RAM memories. Untenable. Memory prices have skyrocketed to unsustainable levels. DDR5 modules that previously cost between 100 and 150 euros now easily exceed 350 euros in many markets. You can blame the AI. And the demand for DRAM for artificial intelligence applications has absorbed a large part of global production. OpenAI alone has accounted for 40% of all productionleaving home users paying the consequences. The worst thing is that it doesn’t look like this is going to be solved soon. And it is that according to analysis firm IDC, the shortage could last until 2027. cheap adapters. In recent weeks we have seen how some users have chosen to build their own RAM memories to face the price crisis. One of the approaches is to use SODIMM to UDIMM adapters. In a video The Hardware Canucks YouTube channel shows how they have tested this solution on Ryzen 7000, 9000, Intel LGA 1700 and LGA 1851 systems without too many problems. The approach is simple: buy DDR5 SODIMM modules (the ones for laptops) that are still relatively cheap and connect them to the system using adapters that cost between 10 and 15 euros. It must be said that this method also has its limitations, and that is that the data transfer speed that these adapters achieve is quite a lottery. In the tests carried out by Hardware Canucks they say that some do not exceed 4,800 MT/s stably, while others reach 5,600 MT/s or even 5,800 MT/s, although it depends a lot on the adapter model and the platform. In terms of performance, the good thing is that the difference is practically imperceptible. According to the content creator’s tests, with an RTX 5090 and a Ryzen 9800X3D, the difference is between 5 and 7% in the worst case compared to conventional DDR5-6000 memory. A more radical solution. Another approach is the one that the Russian modder VIK-on has opted for. And just as they count From Videocardz, the enthusiast has built a functional 32GB stick by combining chips from two 16GB SODIMM modules from SK Hynix, a Chinese PCB, and a heatsink from AliExpress. The total cost: 17,015 rubles, about $218. As explained in the media, in Russia an equivalent module costs at least three times more. Images: VIK-on After physically assembling the parts, a process that requires BGA reballing stations and considerable soldering experience, the modder then integrated ADATA firmware to enable an XMP profile that allowed the memory to run at a speed of 6,400 MT/s. In this way, VIK-on has achieved a functional 32 GB stick that any motherboard could recognize and that according to the modder works stably in games. Between the lines. That making a RAM from home is economically viable says a lot about the state of the market. Furthermore, not everyone experiences the situation in the same way, since in some markets such as Russia prices are especially prohibitive. Of course, soldering memory chips is not trivial, as it requires specialized equipment, technical experience and taking risks of damaging expensive components. The adapter method is much more bearable, but it is most likely that these homemade solutions will continue to be a niche. Most users will prefer to pay the extra price rather than risk soldering components or dealing with third-party adapters. Although if the forecasts end up being true and the crisis extends for several years, the emergence of a secondary market for professionally refurbished modules taking advantage of surpluses is no small feat. It would probably be the last solution we would resort to, but in other markets there might be enough demand. Cover image | Andrey Matveev In Xataka | The computers of the future have found an unexpected ally to store information: fungi

In the midst of the RAM memory crisis, Samsung takes a leap with its HBM4 memory. It does not imply good news for the pocket

by

We are in full RAM price crisis. The industry is a cake that three large producers share and the data centers and the artificial intelligence They want to eat the whole cake. Samsung is one of the companies that manufactures memory for consumption and data centersand will soon begin mass production of its latest broadband memory chips: the HBM4. Don’t throw the bells in the air too soon. HBM4. This technology represents a crucial advance in stacked memories. Its density allows double the bandwidth, key to transmitting more data per second, but they are also up to 40% more energy efficient than HBM3. In short: they consume less energy and have fewer bottlenecks, which translates into an improvement in data processing. Industry sources point out that Samsung will use the 10-nanometer D1c manufacturing process for the matrix of these HBM4 memorieswith an internal structure of 4 nm. It’s a more advanced process than the 12-nanometer D1b from its main rival, SK Hynix. In addition, it will achieve a data transfer speed of 11.7 Gbps compared to 9-10 Gbps of the current standard. Hello Nvidia. South Korean media they point that these new Samsung HBM4 modules they would have passed Nvidia certification testing and will be in february when the company starts mass manufacturing them. Where will they end up? Some to Nvidia’s new AI acceleration system, called Vera Rubinothers at the heart of Google’s seventh-generation TPUs. After these reports, the company’s shares they went up 5.3% in the Seoul market. The enemy at home. In statements To South Korean media, Samsung representatives have commented that they feel quite confident with a new product that will clear up doubts about the company’s ability to supply the demanding needs of data centers. The fifth-generation HBM3E memories were a bottleneck for the company, so major players in the AI ​​industry looked next door: SK Hynix. Also South Korean, she is the second leg of memory chip manufacturing. The third is the American Micron Technology, a considerable distance from the two South Koreans. A year ago we already told that SK Hynix had achieved enormous efficiency in the DRAM stacking process to create these HBM memories, which allowed it to be 8.8 times more efficient than Samsung or Micron and, therefore, produce more modules for an industry that never stops asking. Meanwhile, the two South Koreans were in a race for the development of the new generation HBM4, and Samsung seems to have struck the first blow. Of course, it is estimated that Hynix will also begin mass production of these new memories on the same dates. And the consumer… what? Well nothing. If you were expecting good news related to the price of RAM, it must be said that no improvements are expected. These HBM4 modules will go to Nvidia, but we recently commented that OpenAI had reached an agreement with Samsung and SK Hynix to supply with 900,000 wafers per month. It is the volume equivalent to 39% of the estimated global capacity… and only for one company. Translation? Bottleneck in the market, a manufacturing speed that may not meet that demand and more bad news for the user. We have seen that Micron has abandoned its Crucial brand for consumers in favor of RAM for data centers, and that Samsung and SK Hynix are focused on HBM4 memories en masse, although they are not used in consumer devices, implies that this is where they will focus on this lucrative AI market. In short: Samsung may be dominating the new generation of memories, but 2026 seems difficult for anyone who wants to build a PC, expand RAM of yours, buy a new mobile or even wait for good news from the Steam Machine. Image | TSMC, Google In Xataka | RAM has become so, so expensive that there are manufacturers selling computers in an unprecedented way: “pre-assembled”

The RAM memory market is broken and there are those who point to a new player: Asus

by

The RAM memory market it’s gibberish. The voracity of the data centers has caused energy companies to rethink your renewable goals and? RAM memory increases meteorically in price. This short term is so attractive that Micron, one of the three RAM giants, recently announced that killed its branch of Crucial consumption. And, king dead, king in place: leaks suggest that Asus would be considering its arrival in the RAM market for 2016. It’s not going to be easy at all. In short. The middle Sakhtafzarmag is the one that has sprung the hare: Asus would enter the DRAM market over the next few months. The medium now filter previously information about new processors from AMD and Intel so, although we are talking about a rumor, it is not a medium that comes out of nowhere. At a time when reports point to a RAM shortage until the end of 2027it is not uncommon for other players in the PC market to become interested. It makes sense. And Asus is one of the greats. Your income surpass 18,000 million dollars annually and is present as one of the largest PC hardware manufacturers. Apart from its motherboards and GPU, Asus sells complete desktop and laptop computers (for gaming and office automation) and consoles (there’s the recent Asus ROG Xbox Ally). The RAM segment is one that I had not entered, but the way of doing it has other precedents: Corsair. As I say, Asus starting to sell RAM memory makes sense if we take into account that Crucial, one of the most powerful brands in the consumer segment, has ceased to exist. Crucial was a Micron brand focused on the user: if you wanted RAM, you could buy one from Crucial and mount it on your PC, but with this rise of AI, Micron has seen that the mine is in the data centers. Your explanation is that it is a movement to “improve supply to strategic customers.” The reality is that it is a chore for all PC users. Corsair style. If you have built a PC, it is easy that you have opted for Corsair RAM memories. This brand has monitors, boxes or power supplies, but also memories. However, it is not a memory manufacturer: is an assembler. What Corsair does is design its own PCB, stability systems and heatsinks, and then to that PCB solder the RAM modules from manufacturers such as Samsumg, SK Hynix or Micron. three paths. Entering a new segment is not easy, but Asus has three paths: Be a assembler. Buy memories from large manufacturers and integrate them into your own PCBs. This is what it does, for example, with its graphics cards (Nvidia chip, but its own PCB and dissipation system). ‘Pass’ from big manufacturers that are having difficulties supplying data centers and opting for other emerging ones. For example, the Chinese company CXMT, which has recently achieved validate DDR5 memory modules (and which is on the US blacklist). It would be a win-win for both: Asus validates this Chinese company in the international market and CXMT gets a high-profile international partner. The third is the most risky: Become a memory manufacturer. Asus has the financial resources, but not the experience to do it. It would be the best to create a more controlled product, but in the end it means facing a greater risk. wasp nest. As we said from the beginning, the arrival of Asus in the RAM memory segment is a rumor that arrives just when RAM supply chain is broken. It is something that affects us as consumers because we see exorbitant prices, but ‘Big Tech’ also has to pay more for RAM, there is a lot of speculation about the price of machines like Steam Machine that will be launched right in this price hurricane (some RAM modules are more expensive than any console) and even memory manufacturers they may face difficulties in their products, such as Samsung. The arrival of another assembler does not change the balance of power that the big three – Samsung, Hynix and Micron – have since Asus would buy from them, but if it associates with Chinese companies, things change, and a giant like Asus will be lat the gateway of a CXMT or Fujiuan Jinhua would add pressure to the current oligopolistic system. Decongestion? Difficult. Now, just because a new player enters this playing field does not mean that prices will drop immediately. Everything will depend on how they enter, but if they assemble memories from the three most established manufacturers, there will still be no decongestion in the market because they will be more likely to distribute the same finite product. If they enter through a Chinese manufacturer, the situation could be alleviated as long as the stock is not broken. In any case, if they are really going to make some move for 2026, it wouldn’t take long for us to have official news – and CES is just around the corner. We have contacted Asus, we will update when we hear back. Images | Hector Reyes In Xataka | AMD’s problem is not that it doesn’t make good GPUs for AI. It’s not even close to NVIDIA

The RAM memory crisis seemed to have its months numbered. Micron has a completely different perspective

by

They say that there is never a bad time to do something you really want and that, many times, the only thing that stops us is finding the right excuse not to start. That idea can work in many areas, but today it doesn’t fit very well if what you have in mind is build your own PC. At least not without assuming that the current context clearly works against it. We are witnessing firsthand how the so-called “memory crisis” is pushing upand notably, the prices of NAND memories, key in SSD units, and of the DRAM used in computers and laptops. We have more and more reasons to be patient, because an actor as relevant as Micron already warns that the challenges for the sector will persist for quite some time. The memory crisis is still far from resolved The company has put a date on the table and it is not what many expected. In its communication of resultsSanjay Mehrotra, CEO of Micron, spoke of “tight conditions” in DRAM and NAND and said that they are expected to “persist through and beyond 2026“In other words, if today we are already seeing how memory is becoming more expensive, Micron is warning that it does not point to a clear return to normality in 2026. That nuance matters, because it changes the horizon of anyone who is thinking about building or upgrading a PC in the short term. There are numbers that help to understand why this phenomenon does not remain a simple market swing. The firm once again reported record revenues of $13.64 billion in the last quarter, compared to $8.71 billion in the same period of the previous year, driven by the push for AI. That does not mean that there is excess product in all segments, because capacity and industrial priorities do not adjust to the pace of demand. Micron says it expects to increase its DRAM and NAND shipments by 20% next year, but acknowledges that boost isn’t enough to catch up. To understand why the domestic market is more exposed, it is worth looking at the photo of the factory. Micron is pushing its business toward HBMa memory designed for high-performance systems in data centers, and that has an opportunity cost. It is a technology that uses three times more silicon wafers than conventional DRAM, which means that, with the same capacity base, fewer units can be produced for the rest of the segments. It is not that consumption memory disappears, it is that it has less priority. Micron is pushing its business toward HBM. The first consequence is already being noticed by those of us who look at prices to build or expand a PC. Memory is what is becoming more expensive and the effect is seen, for example, in DDR5 kits. From there, the tension begins to filter through the rest of the chain, not only due to price, but also due to availability. The decision with Crucial also fits into this shift in priorities. Micron will stop selling consumer products under that brandwhich means one less player in this market, and greater pressure for those who are still in the race in the domestic sector. If Micron makes one thing clear with its roadmap, it is that standardization is not around the corner. The company is accelerating investment and capacity, but with a calendar that moves in yearsnot in weeks, and that forces us to look at 2026 with another face. For those who are thinking about buying or building a PC, the prudent reading is simple: it is advisable to assume that memory will continue to be a sensitive component, both in terms of price and availability, for a good period of time. Images | Micron | Samsung In Xataka | RAM is so expensive that smartphone manufacturers already have a plan: return to phones with 4 GB of RAM

Pebble wants us to carry an “external memory” for the brain on our finger

by

There are everyday moments when a fleeting idea crosses our mind and we know that if we don’t save it instantly, it will probably disappear without a trace. It can happen while we are cycling, cooking or simply walking with our hands busy, when taking out the cell phone is inconvenient or outright impossible. That feeling of losing something that seemed important has led some companies to explore an unexpected solution: turning the index finger into a place to capture quick thoughts before they escape. The fear of forgetting what is important. For Pebble, the challenge is not just in coming up with an idea out of the moment, but in how often it happens. Its founder states that it happens to him between five and ten times a day, and that the most frustrating thing is not the idea itself, but the subsequent certainty of having forgotten something without being able to recover it. That recurring sensation is what, according to the company, justifies finding a more direct mechanism to record brief thoughts before their context is lost. A notepad ring. The device proposed by Pebble, the Index 01takes the form of a compact ring, built in stainless steel and equipped with a physical button and a microphone. By pressing it, the user can capture a short voice note immediately. It is available in various colors and sizes, and has water resistance to withstand continued use. Its main function is to offer a quick entry point to save information without depending on the phone at the exact moment it arises. From finger to app: Each recording begins with a press of the button, which activates the ring microphone and saves the audio to its internal memory, without any additional processing. When the phone is nearby, the recording is transferred via Bluetooth and that’s where all the work happens: the Pebble app converts voice to text using a recognition model that works locally, and then an LLM that also runs on the phone itself determines whether to create a note, set a reminder or add an event to the calendar. It never plugs in, but it runs out: Pebble opts for a silver oxide battery similar to what hearing aids use, allowing the ring to run for years without needing to be recharged. According to the company, an average use of between ten and twenty daily recordings of a few seconds is equivalent to about twelve or fifteen hours of accumulated audio, enough to achieve that long autonomy. When the stack nears the end, the app notifies the user, who can purchase another ring and send the previous one for recycling. The approach means that the battery cannot be replaced or recharged, something Pebble openly acknowledges. When the end-of-life notice arrives, the user must purchase a new ring. As we say, the company offers the possibility of sending the old device for recycling, but does not mention discounts, replacement programs or return compensation, so the replacement apparently works as a separate purchase. Pebble insists that the ring is designed to process information locally and limit its scope to what is strictly necessary. The connection between the device and the mobile is encrypted, and both the speech-to-text conversion and the classification using a language model occur on the phone itself and, by default, do not require sending the data to external servers, although the company offers an optional cloud backup system for recordings that is still in development and plans to offer encryption. The ring does not listen continuously or record health data, and it does not integrate a speaker or vibration. Its operation is limited to the moment in which the user keeps the button pressed. When memory lets itself be hacked. Beyond recording notes, Pebble allows you to configure the ring to perform additional actions with single or double presses, from controlling music to taking a photo or activating home automation routines. The app supports sending reminders to services like Notion and offers support for over 99 languages. The company also describes an action system based on MCP, small extensions that run on the mobile itself and that, according to its roadmap, should expand what the device can do without depending on a central server. From watch to ring: Pebble is going through a relaunch phase in which it seeks to expand its catalog beyond smartwatches. After recovering your brand and sending your new Pebble 2 Duoprepares the arrival of Pebble 2 Time with a significant level of prior demand. In that scenario Index 01 appears. The founder himself summarizes its bet stating that the ring has ceased to be a technological device and has become “an extension of the brain”, a phrase that reflects the ambition with which the company presents this project. Price and availability of Index 01. The company puts the starting price at $75 during pre-sale, with a rise to $99 when the first units begin shipping globally starting in March 2026. The device is in the design validation phase and is produced in the same plant that works with Pebble Time 2, where the current prototypes are assembled. Shipments will depart from Asia under a DDP system, so taxes and duties will be handled prior to delivery. Images | Pebble In Xataka | We have tested the new Google glasses with Gemini: AI and today’s technology drive the dream that Glass promised

make Apple’s memory configurations look cheap

by

For years, criticizing Apple’s pricing policy has been a more than justified constant: we have seen how the company charged $100 per a leap in capacity that it cost them ten, or asked astronomical figures such as 690 euros for a 2 TB SSD which in the market cost 100. However, the current memory crisis has turned the tables: the shortage is so serious and the inflation so aggressive that Apple’s rates are beginning to seem even “economical” compared to the competition. Context. As we have been telling in Xataka, the memory industry is facing a critical scenario. The AI ​​fever has caused chipmakers such as Samsung and SK Hynix to prioritize the production of HBM memory for servers and AI GPUs, leaving aside the consumer market. The result is a supply deficit that has skyrocketed prices by up to 300% for some components, forcing manufacturers like Xiaomi to warn that our next mobile will be more expensive. Others like Micron have not brought good news either: in fact, the manufacturer closes its Crucial consumer division to focus on the lucrative data sector. Dell and Lenovo raise prices, Apple freezes. The first chip to fall has been Dell. According to industry reportsthe company plans a price increase of 15-20% this month, while Lenovo has begun to warn its customers of imminent increases for early 2026 precisely due to the DRAM shortage. But the most striking thing is not the general increase, but the cost of the expansions. Dell is currently charging a surcharge of $550 for going from 16 to 32 GB of RAM in some of its XPS laptops, a figure that easily exceeds the $400 that Apple asks for the same jump in its MacBook Air. Bounce effect and an exception. The current situation has led to paradoxes never seen before: the traditionally expensive Apple maintains its prices stable (for the moment and thanks to already high margins), while the PC world suffers from market volatility. The Framework Modular Laptop Manufacturer He took the opportunity to point out the play: denounces that Dell’s prices are “abusive” and highlights that they charge 85% less for the same memory upgrade. However, even they warn that their cheap inventory will run out sooner rather than later. A future of pressured margins. Although Lenovo is well positioned to weather the storm thanks to its scale, financial analysts They warn that the rise in memory prices threatens their margins and end-user demand for the next 12-18 months. With Samsung and SK Hynix refusing to increase production To avoid a new bubble, it seems that the industry has entered a phase where paying premiums for RAM will be the new normal. This makes Apple’s historical “dunks” seem, ironically, like a refuge for its stability. Cover image | Composition with images of Applesfera and Andrey Matveev for Unsplash In Xataka | We have been assembling computers in the same way for many years. The RAM memories of the future promise to change that and more speed

The memory of young people is deteriorating at a record pace. Science thinks it knows why

by

The memory problems among youth are beginning to be worrying. This is what a new study scientist published in the magazine Neurology and that tries to answer why this happens and above all the reasons that exist for our youth to begin to be in decline in regards to to your memory. The surprise. What can logically be expected is that with the passage of time and accompanying aging, memory problems begin to appear that anticipate dementia. But in the United States, after analyzing millions of people, they have seen that the population most affected by this ‘mental fog’ is precisely the youth. And the result in this case is very important: self-reported cognitive problems among young adults aged 18 to 39 have almost doubled in the last decade. But it is something that we are not understanding. The study. To reach this conclusion, a total of 4.5 million people who responded to the national survey of the Behavioral Risk Factor Surveillance System (BRFSS) from the CDC and collected between 2013 and 2023. In this way, there was a truly large sample of people to analyze, although limited only to the United States. The results in this case were quite clear: the prevalence of adults reporting a cognitive disability increased from 5.3% in 2013 to 7.4% in 2024. But what was truly interesting came when separating the results by demographics: In young people aged 18 to 39, the rate skyrocketed from 5.1% in 2013 to 9.7% in 2023. This group is, in fact, the driver of the overall increase in the entire population. In those over 70 years of age we saw a decrease in prevalence from 7.3% to 6.6%, when logic tells us that it should increase. Other factors. In order to know the reason for this increase, other factors behind the respondents had to be traced as well. In this case it aimed at the income level: Have low income with less than 35,000 dollars a year left us with a prevalence that increased from 8.8% to 12.6% With high incomes (>$75,000) the rate was much lower, although it also dropped from 1.8% to 3.9%. But the same thing happens with the educational level, where young people who did not even have high school went from 11.1% to 14.3% while those with university degrees increased from 2.1% to 3.6%. And even in order to obtain much more information, they wanted to analyze the prevalence according to the race of young people, where it could also be seen, for example, that Asian adults are the ones who reported the least cognitive problems. Specifically, the data is the following: American Indians/Alaska Natives: continue to have the highest prevalence, rising from 7.5% to 11.2%. Hispanic adults: saw a significant increase from 6.8% to 9.9%. Black adults: The rate rose from 7.3% to 8.2%. White adults: increased from 4.5% to 6.3%. Asian adults: Consistently maintained the lowest rates, going from 3.9% to 4.8%. What is happening? With all the data in hand, it is logical to think about what is happening so that young people increasingly have more cognitive problems. And for researchers there is not only one valid answer, but there are several that are being proposed. The first of them is that there is greater awareness about this problem, and that is why there are more people who raise their hands when presenting it and have no doubts when it comes to seeking help. But there are also other factors such as economic stressors or work problems that seem to be contributing to these trends. All this without forgetting that the greater presence of digital tools may have meant that our memory is not as trained. But all the social and economic factors we face today can also mark an important milestone when it comes to the real burden on our minds. This ‘overload’ can condition the appearance of these highly relevant cognitive symptoms. Images | Eliott Reyna Milad Fakurian In Xataka | Finding a job had always been a good way to escape poverty: in Spain it is no longer true

Log In

Forgot password?

Forgot password?

Enter your account data and we will send you a link to reset your password.

Your password reset link appears to be invalid or expired.

Log in

Privacy Policy

Add to Collection

No Collections

Here you'll find all collections you've created before.