step

China has taken a new step in its high-speed race. The CR450 has just reached a new milestone in its tests

by

China has spent years perfecting machinery that not only symbolizes speed, but also industrial precision. Its last exponent, the CR450has shown the scope of that search: in its most recent tests, two trains reached a combined speed of 896 km/h at the intersectiona new record in the Chinese system. It is not an isolated gesture, but a step within the innovation program launched in 2021 to raise the bar for high speed with more reliability and performance. The new registration was confirmed on October 21. During tests on the high-speed line connecting Shanghai, Chongqing and Chengdu, two CR450 trains crossed each other, reaching a relative speed of 896 km/h. In the same test campaign, one of the prototypes once again reached 453 km/h per unit, equaling the record set in 2023. The tests, they explain, are part of the “evaluation operation” that is currently being carried out on the Wuhan–Yichang section, a prior step to a more demanding phase scheduled for 2026. Speed ​​is on the table, but the operation is not yet At first glance, it might seem that two trains traveling at 453 km/h should add up to a crossing speed of 906 km/h. In practice, testing conditions prevent this. As China Railway Group explainedthe exact moment when both units are on parallel tracks it only lasts a few secondsand getting them to maintain the same speed at that point is extremely complex. For safety reasons, technicians increase speed gradually, ensuring stability and synchronization before attempting new records. The CR450 is not an isolated project, but one more piece of the railway plan that China launched in 2021 to raise commercial speed to 400 km/h. The challenge is not minor: maintaining that pace without increasing consumption or noise. Before entering service, the prototype must complete 600,000 kilometers of tests under real conditions, an essential requirement for its certification. This year, trials have extended from the Chongqing to Qianjiang sections to the Wuhan–Yichang line, where technical teams continue to fine-tune the train’s behavior in prolonged use scenarios. How Sina collectsmuch of the CR450’s advancement can be understood by looking inside its engineering. The train incorporates permanent magnet motors with a total power of 11,000 kW. The weight has also been reduced about 50 tons thanks to the use of carbon fiber and magnesium alloys, and the aerodynamic profile has been optimized with a longer nose, 15 meters. They claim that at 400 km/h, the noise level inside the car barely reaches 68 decibels. Although the CR450 has already demonstrated its technical capabilities, its commercial deployment remains without a clear destination. Today there is no operational line in China prepared to travel at 400 km/h. The first that contemplates this possibility is the Chengdu–Chongqing Central Line, approved in 2021 with a base design of 350 km/h and adaptable sections for future tests at higher speeds. According to China Economic Newsthe plan is that next year the train will undergo a more demanding testing phase there, the closest so far to a real service scenario. The development of the CR450 is divided between two of the main railway manufacturers in the country. The CR450AF version has been built by CRRC Qingdao Sifang, while the CR450BF is built by CRRC Changchun. Both They share an eight-car configuration —four engines and four trailers—. Official information indicates that they incorporate advanced communication and braking systems, as well as high stability bogies designed to maintain balance even in extreme speed tests. The immediate future of the CR450 passes through the aforementioned line, where over the next year it will undergo tests closest to real operation. There is still no confirmed date for its entry into service, and those responsible for the project emphasize that the priority continues to be technical validation. For now, we have to wait to see if all the promises of the program materialize and if the new train manages to transfer its laboratory achievements to the operational field. Images | China Railway Group In Xataka | The shortest launch in history: a million-dollar luxury yacht sank just 200 meters from the dock

China is taking a giant step in its quest for technological self-sufficiency: its own EDA software

by

A company called Qiyunfang just done a unique presentation at the Bay Area Semiconductor Expo held in Shenzhen, China. In it he has presented two EDA platforms. And with them it has opened the door to something in which the Asian giant totally depended on the US: designing your own chips. What is EDA. Electronic Design Automation (EDA) software is the fundamental tool and the gateway to be able to design chips and printed circuit boards (PCBs). Historically, this segment has always been dominated by American companies: Synopsys, Cadence and Mentor Graphics / Siemens EDA were the absolute references. They “translate” the ideas of the engineers who design the chips, and convert them into functional plans that manufacturers such as TSMC or SMIC can manufacture. Who is Qiyunfang. This company, founded in 2023, is not just any one: it is SiCarrier subsidiarywhich in turn collaborates with Huawei Technologies. As if that were not enough, SiCarrier is a semiconductor manufacturer that has financial support from the Shenzhen government. The US had China tied. In May of this year, China began to block the export of its rare earths, and the United States responded with a blockade that was equally harmful to China: the aforementioned American companies They could no longer sell their services and their EDA software to its Chinese partners. It was one of the most effective ways to “strangle” the Chinese semiconductor industry: if you can’t design the chip, it doesn’t matter if you have factories to produce it. SMIC, for example, it’s been 20 years using Synopsys EDA design suites. With the veto he was left unable to use them. China once again opts for self-sufficiency. The solutions presented by Qiyunfang theoretically allow for domestic solutions for EDA software for both schematic designs (conceptual design) and PCB (physical design of the board). Not only that: these platforms run on a completely Chinese software stack—operating system, database, middleware. Qiyunfang is not a product, it is a break with dependency in two of the crucial stages of chip design. The key is technological independence. If these platforms comply, China will have a solution immune to sanctions, also integrated into the Chinese national technological ecosystem. The trade and technology war that the country maintains with the United States has encouraged both parties to seek precisely to avoid dependence among themselves and also with other countries. It’s the “I’ll stew it, I’ll eat it” taken to the maximum extreme. The other challenge: advanced chips. Even having its own EDA software, China still has a colossal challenge with advanced chips that use 3 and 5 nm photolithography and that take advantage of UVE technologies. They continue working on these types of systems, but until they have them, Qiyunfang’s software platforms are a fantastic option for developing more “mature” but equally important chips such as those for the automotive sector or industrial applications. China continues to move key chips. This news confirms the trajectory that Xi Jingping established with its famous “Made in China 2025” plan. It seeks to conquer the key technologies of the future: AI, robotics, automotive and of course the manufacturing of semiconductors without external dependencies are little by little a reality in the Asian giant, and this new milestone of this Chinese company seems to demonstrate it. In Xataka | Before the tariffs, China bought most of its beef from the US. After the tariffs another country has won

PLD Space, one step away from becoming the company that has developed an orbital rocket the fastest

by

Whether in the Elche factory, on a test bench at Teruel airport or on the launch pad under construction in French Guiana, PLD Space is abuzz. The company advances one milestone per week and he tells us why: the Miura 5 rocket is practically ready at the design level. “I would tell you that it is 99%,” says Raúl Torres, CEO of the company, in an interview with Xataka. Candidate to become the Europe’s first private orbital rocketthe Miura 5 is about to finish the Critical Design Review (CDR) and take shape for the first time. “Now we are finishing the QM1 qualification models and starting the QM2, which means that shortly, and I’ll leave it there, we are going to have a first teachable Miura 5,” he reveals for the first time. This first fully integrated model will not fly, but will allow PLD to close engineering fronts and carry out key tests before the end of the year. If everything goes according to plan, the rocket chosen to take off will begin assembly in January. “The idea would be that in May we would be in Guyana to start doing the combined tests with the French space agency CNES,” confirms Torres, adjusting the schedule that originally pointed to a launch at the end of 2025. It is not an unexpected adjustment, but it was pending official confirmation since Chris Larmour, founder of Orbex, PLD’s British competitor, 1,000 euros were bet with Raúl Torres that the Miura 5 would not fly in 2025. Raúl accepted the bet. Will he pay Larmour now? “We have invited him to come sign the rocket at the end of the year, we are waiting for him to answer us,” says Torres. “I would like Orbex to also invite me at the end of the year to sign their rocket. Mine is going to sign it, so I only have to pay half of the bet.” Works in Guayana, lighting in Teruel If the Miura 5 flies in early 2026, PLD Space will be one of the fastest companies to have developed an orbital launcher, which is even more impressive considering the Spanish company’s financing compared to several of its competitors. But PLD Space is not starting from scratch. The successful launch of the Miura 1 suborbital rocket in October 2023 was the graduation of a team that now faces a higher challenge. “Miura 1 has been like primary school, ESO and high school, and now we are at university,” explains Torres. “That is why we have developed Miura 5 so quickly, because we have gone one step ahead with many developments.” Technologies such as the stage power system, cryogenic protections or the welding techniques of the Miura 5 are a direct inheritance from its little brother. However, “university” brings new and more complicated subjects. The most obvious technological leap is in the Miura 5 engines. The five TEPREL-C of the first stage and the vacuum-optimized TEPREL-C of the second They are beasts of another categoryespecially due to the introduction of turbopumps. PLD has developed most of the critical components in-house, such as liquid oxygen and kerosene valves. Combustion chambers are manufactured by electroplating copper and nickel, turbopump housings are 3D printed, and high-precision rotating components are machined. The objective is to achieve a production rate that allows one engine to be manufactured every two weeks in the Elche warehouse. PLD Space passed a fundamental milestone on October 6 with the first static ignition of a fully integrated TEPREL-C Vac in its facilities at Teruel airport. With 75 kN of thrust, it is one of the most powerful vacuum engines ever powered by a private company in Europe. But the real muscle of the rocket will be in the five TEPREL-C engines responsible for takeoff. Each one has 190 kN of thrust, almost double than its competitors. When will we see the first roar of a Miura 5 with the TEPREL-C fully integrated? “In one quarter you should expect the long and qualification tests of both the first and second stages, and also the restart test of both engines,” Torres told Xataka. To validate each component, PLD Space has also deployed new infrastructure at the Teruel airport. The T3 bench has been the protagonist of the static and compression tests of the rocket structures. Valves and gas generators are tested on bench T6. Bench T7 will be used for qualification of first stage Teprel-C engines and second stage long duration ignitions. The T9 bench will be used to test the separation between the first and second stages. Meanwhile, thousands of kilometers across the Atlantic, PLD Space construction in French Guiana has begun. PLD has become the first New Space company to begin construction of its own launch base at the Guyana Space Center. “It is very likely that Miura 5 will be before Kourou’s works,” says Torres. The first structures of the launch pad They are being built in Spain. The rocket should arrive in South America in May. Advances in reuse since flight 1 Inspired by SpaceX, PLD does not conceive of a modern launcher without reuse. And their plan for the Miura 5 is to start collecting landing data from the first flight. If it achieves stage separation on its debut launch, the rocket will perform a maneuver boostback like that of the Falcon 9. “In flight one mission, in the test flight that we will do next year, we are going to try to re-enter the stage,” confirms Torres. After separation, the rocket will turn around and turn on its central engine for a few seconds to brake. “The booster will be ready to re-enter. We don’t want to miss the slightest opportunity to collect data.” And he talks about data because he does not expect to recover the rocket. “Evidently, it’s not going to happen the first time.” The first flight won’t even have a parachute. The main objective is to survive reentry from a hypersonic speed at Mach … Read more

OpenAI has taken its first step towards Latin America. Behind it there is an investment of 25,000 million in Argentina and many questions

by

For almost any country in Latin America and the world, a company like OpenAI announcing a multi-million dollar investment sounds like a golden dream. It is not only the most influential company on the planet in artificial intelligence, but also one of the pacesetters in the industry. Its arrival promises jobs, economic movement and global visibility. But, as with any large-scale project, it also has doubts: energy consumption, water use or the sustainability of a data center of hundreds of megawatts are not minor issues. Argentina, at least on paper, has been chosen to attempt that leap. The announcement of the Argentine Government It is based, at least for now, on a single document: a letter of intent signed between OpenAI and the local company Sur Energy. The text, published on October 10, 2025, mentions an investment of “up to $25 billion” for a data center of “up to 500 megawatts,” under the Incentive Regime for Large Investments (RIGI). The location of the project is not specified nor are deadlines or construction phases detailed, which keeps it in a preliminary stage. The Argentine president met at the Casa Rosada with representatives of OpenAI last week Silences that weigh. There are details that attract attention. A multimillion-dollar announcement, linked to the expansion of OpenAI in the region, and yet neither its CEO nor the company itself have communicated it through their official channels. That they have not done so does not invalidate the project, but it does mark a distance with the institutional enthusiasm on the Argentine side. In this type of operations, communication is usually part of the message. Here, for the moment, it is conspicuous by its absence, at least on the side of the American startup on its website and social networks. The plan: AI factories at scale. Stargate is not an isolated project, but the name that OpenAI uses for its global infrastructure program. Its objective is to build a network of data centers capable of supporting cutting-edge artificial intelligence models, the technology that gives life to tools such as conversational chatbots or image generators. In the case of OpenAI, those models are the ones hidden behind products like ChatGPTbased on systems such as GPT-4 either GPT-5. The plan began to take shape months ago, when the company announced an ambitious infrastructure project in the United Stateslater expanding it to other countries. Interior of Stargate 1, the first large-scale data center developed under OpenAI’s own program Power, density, permissions. Data centers for artificial intelligence operate in another league. They concentrate massive training on GPUs with industrial-level consumption and an energy density much higher than that of a conventional data center. Each room requires advanced cooling systems capable of constantly keeping the temperature under control. And, although permits and licenses are required as in any facility of this type, its scale and technical requirements make building one of these projects a much more complex and lengthy process. {“videoId”:”x8jpy2b”,”autoplay”:false,”title”:”What’s BEHIND AIs like CHATGPT, DALL-E or MIDJOURNEY? | ARTIFICIAL INTELLIGENCE”, “tag”:”Webedia-prod”, “duration”:”1173″} RIGI and financing: promise vs contract. As we say, the project is covered by the Incentive Regime for Large Investments, a tool created by the Argentine Government to attract foreign capital through tax, customs and exchange advantages. In practice, the RIGI facilitates the conditions so that a large-scale project can be financed, but it does not guarantee that the investment will materialize. Patagonia sounds loud, it’s not official. On paper, there is no defined location. Neither the Argentine Government nor OpenAI have mentioned Patagonia in their statements. Even so, the name of the local company that appears in the letter of intent, Sur Energy, fuels the idea that the project could be developed in the south of the country. The president of Argentina, Javier Milei, with the CEO of OpenAI, Sam Altman, in May 2024 Climate and design: allies or burden. If the southern hypothesis gains strength, it is also for a technical reason: the climate plays in its favor. Colder areas allow you to operate with less cooling energy and take advantage of outside air, something that reduces costs and emissions. In parallel, the availability of water continues to be a decisive factor. The new artificial intelligence campuses, aware of this risk, are adopting cooling systems that minimize the use of water resources. We will have to wait to know the option chosen by OpenAI. When the network or the water say no. The location of a data center does not depend only on the weather or tax incentives. Factors such as the electrical grid or the availability of water can mark the success or failure of a project. Mexico, for example, is one of the largest technology hubs in the region, but even there a Microsoft data center ran into the limitations of the national network. and had to resort to gas generators. In Chile, Google saw its plan blocked due to excessive water use. They are reminders that it is not enough to have space: you need infrastructure. In Xataka In the nineties, no one saw how the Internet would starve factories. Thirty years later, AI is doing the same thing From exclusivity to autonomy. For years, OpenAI’s infrastructure depended almost entirely on Microsoft. In 2019, the Redmond company invested 1 billion dollars and became your exclusive cloud partner. Over time, that alliance grew to exceed 10 billion, consolidating Azure as the platform where the company’s models were trained and executed. However, OpenAI has been seeking greater operational autonomy. The Stargate program responds precisely to that idea: having its own computing resources and diversifying its technological dependence. From paper to concrete. For now, it all depends on the next steps. For the initiative to move forward, a definitive contract between OpenAI and Sur Energy, the presentation of environmental studies and electrical interconnection licenses will be necessary. The financing scheme and long-term energy agreements will also have to be defined. Only with these pieces in place can we speak of a real work. Until then, … Read more

Alibaba has one of the best open source AI models. Your next step: use it in robotics

by

Alibaba has taken another step in its commitment to artificial intelligence by creating an internal team dedicated to roboticswhich will operate from qwenits AI modeling division. The Chinese giant, owner of one of the best open source AI models, now wants the Qwen team to know how to apply its knowledge in robotics, a sector that is beginning to awaken interest, not only in industry, but also with the arrival of projects in the domestic sphere. Who leads the project. Justin Lin, technology manager at Qwen and expert in multimodal models (capable of processing text, sound and images), was the one who has confirmed the creation of this “small team for robotics and embodied AI” through their social networks. Lin has worked on the development of the Qwen models, which are currently among the most popular in open source globally. The vision behind the movement. According to Linmultimodal AI models are evolving into “fundamental agents” capable of performing complex long-term reasoning tasks thanks to reinforcement learning. “They should definitely make the leap from the virtual world to the physical world,” he said. explained the manager, making clear the intention to apply these technologies in tangible devices. Alibaba’s big bet. This announcement is part of Alibaba’s broader strategy in the sector. Last month, the company led a financing round of 140 million dollars at X Square Robot, a Chinese robotics startup. In addition, its CEO Eddie Wu esteem that global investment in AI will reach $4 trillion in the next five years, a figure that reflects the sector’s expectations. Global competition. Alibaba is not alone in this race. Nvidia and SoftBank are also making significant moves in smart robotics. SoftBank just announced the acquisition of ABB’s industrial robots business for $5.4 billion, while Nvidia CEO Jensen Huang has qualified the combination of AI and robotics as a “multi-billion dollar” long-term growth opportunity. China is also the world’s leading power in the robotics sector. And only in 2024, Chinese factories installed nearly 300,000 industrial robotsa figure higher than the rest of the world combined. The Qwen factor. The choice to place this team within Qwen makes all the sense in the world. Seven models of the Qwen series are currently listed in the top 10 Hugging Facewith the multimodal model Qwen3-Omni occupying first place. This strength in AI provides the company with a solid foundation to develop advanced robotic applications based on the journey they already have with Qwen. Cover image | zhang hui and Possessed Photography In Xataka | AI companies have just encountered an unexpected challenge: insurers have started to turn their backs on them

29 years ago we clone to the Dolly sheep. Thanks to this, today we are one step closer to solving the fertility crisis

by

Imagine a future in which the infertility caused by the lack of ovules or sperm is no longer an insurmountable obstacle. A future where two men can have a biological child together, or where A woman who has lost her ability to produce ovules For age or for a cancer treatment may have offspring with its own genetics. This future, which until now belonged to science fiction, is a little closer thanks to a revolutionary advance that has been published in Nature. Biology This advance, which seems like a science fiction, has been made by the team of researchers at Oregon Health & Science University, led by Dr. Shoukhrat Mitalipov. In this case they have managed to develop an experimental technique that forces a skin cell (somatic) to reduce your number of chromosomes in half. It is, in essence, the most crucial and complicated step in the creation of a gamete (an ovule or a sperm). A process they have called mitomeiosis. To be able to understand it, you have to know that all the cells of our body have in total 46 chromosomes in its nucleus. But there is an exception: sperm and gametes that They have 23 chromosomes. A very important number so that when an ovule and sperm merge, they have a total of 46 chromosomes. That is why it is revolutionary that they have managed to get a skin cell to have 23 chromosomes to be an ideal candidate to give offspring. The trick. The natural process to create these haploid cells (with 23 chromosomes) is called meiosis. A very complex type of cell division that has been investigating for a long time. This made it replicate in a laboratory, which is known as in vitro gametogenesis (IVG) was one of the greatest challenges of biology. Something that now reminds us of what we already saw with the Dolly sheep in the cloning process. OHSU’s team addressed the problem in an ingenious way. Using a technique similar to cloning, called nuclear somatic cell transfer (SCNT). A technique that is complex, but can be summarized in three different steps: The first thing is to take a donated human ovule and extract the genetic material. In this way, the ovule maintained all its cytoplasm with the organelles, which ultimately is like the machinery that the cell has to produce energy and carry out many processes such as meiosis. Once done, a skin cell is taken (a diploid cell with 46 chromosomes in a 2N state) and is extracted the nucleus inside. Now it only remains to introduce the core of the skin cell into the ovule that has been emptied. The result. In this case it was amazing, since the ovule cytoplasm could ‘deceive’ the skin’s core, forcing it prematurely into a state similar to the metaphase of the Meiosis. This caused its 46 chromosomes to be organized in a spindle ready to divide, despite having skipped the DNA duplication phase in the cell cycle that is before the division of the genetic material. The problem. However, here they met a wall. In nature, the entrance of the ‘active’ sperm to complete its division, being mediated by a large number of zinc. But in this case, when they tried to fertilize the SCNT ovules with sperm, the vast majority (almost 77%) remained ‘arrested’ without reacting. The natural signal was not enough for this artificial construction. The solution in this case went to develop an artificial ‘starter’ key. After sperm fertilization, they applied an assisted activation protocol an electrical pulse by electroporation to simulate the calcium entry caused by sperm to its entrance, followed by a treatment with a chemical inhibitor called Roscovitin. And it is something that ended up working. Forced activation made the modified ovules leave their arrest and complete the division. The 46 chromosomes of the cell were separated, leaving an average of 23 chromosomes within the fertilized ovum (now a zigoto) and expelling a small polar body from the rest, having achieved the long -awaited ploidy reduction that was the objective of this experiment. Progress. The embryos resulting from this experiment containing chromosomes of the skin and sperm cell, beginning to divide and even some reached the blastocyst phase (an early development of about 5-6 days), with a success rate of 8.8%. This shows that genomes can integrate and work together. It’s just a test. The authors who are still a long way forward, since for now it is a “proof of concept” by not being a perfect replica of natural meiosis. In this case, segregation is random unlike meiosis, where it is ensured that each daughter cell receives a copy of each of the 23 types of original chromosomes. In this project the separation of homologous chromosomes (the paternal and maternal) was completely random. This generates aneuploid embryos that are incompatible with life. In addition, it also lacks ‘cross -rise’ or crossovera vital mechanism in meiosis where paternal and maternal chromosomes exchange fragments creating genetic diversity. This is something that is not present in this process and that takes away a lot of variability. The future. Despite the limitations of this study, work is a fundamental milestone. It aligns with other laboratories such as the Japanese Katsuhiko Hayashi that in 2023 managed to create functional ovules From male mice skin cells, with which healthy offspring were born. In the long term, the implications of these studies give hope to those women who suffer from infertility due to lack of functional gametes and who want to have offspring with their own genetics. The same happens in same -sex couples that also open the door for a couple of men (using a skin cell to create an ovule) or women (creating sperm) can have biologically related son between both members of the couple. Although we cannot also forget that right now there is a fertility crisis that causes that in Spain, for example, there is reduced birth rate. This is also conditioned that it is … Read more

The step of a Levantine town hall

by

Surely once traveling through some point in Spain you’ve seen a Plate hanging on the facade of a historic buildingCity Council or train station indicating the altitude of that place with respect to the sea ​​level In the city of Alicante. Because Compare the altitude of a city like Pamplona or Soria with respect to the Levantine city is a question that has a scientific answer and that we are going to inevit today in Xataka. The first thing to keep in mind is that sea level is not the same or constant by the different tides, Nor is it much less the same throughout the earth. That is why each country takes as Reference level a specific point of its territory. It is what is known as zero level. And any altitude that you want to calculate in that country will be made compared to that data. Nor can we forget that The technology we have now compared to what existed two centuries ago it is not even for the same. Then there were no altimeters, which we now see anywhere: the car, the mobile, gas stations … that made calculating the altitudes of any place was a very laborious work that required an infinite patience. The task of choosing that reference point also occurred in Spain. Specifically, he fell to the National Geographic Instituteinaugurated in 1856, which was entirely dedicated at that time to elaborate precise topographic maps of the country. In a publication of the institution in the 10th Hispanic-Portuguese Assembly of Geodesia and GeophysicsThey explain that Alicante’s choice was not a coincidence. This article mentions that for the measurement to be as stable as possible, a low point had to be chosen so that negative heights were not produced, hence the sea level was chosen. So a site was sought on the Spanish coast in which The variations between low and pleamar were the minors possible. That is, I had to have Little maritime oscillation. That already completely ruled out some cities of the Cantabrian, because in places like Santander or San Sebastián the sea level can vary up to four meters during the day. Four years of manual measurements With that in mind, the scientists thought of the coast and the waters of the Mediterranean. Precisely Alicante, is inside a baywhere the daily variation of sea level is just a few centimeters. It was also found that this city was the best in terms of weather criteria, atmospheric pressure or seasonal variability. But the most curious thing about the IGN’s decision was not the choice of the city, but the specific place where they decided to establish the zero level. To know exactly where to locate it, They made daily measurements from 1870 to 1874 by handhelping only a strip on the queen stairs located in front of the port of Alicante. After almost four years of work, it was decided to establish that right point on the first step of the interior stairs of the City of Alicante. Specifically it was determined that this exact step It was 3.41 meters above sea. There a plaque was installed that says: “NP 1”. That is, precision level 1. since then, all the altimetry in Spain was measured according to the difference with this point, adding 3.41 meters. “From there, the altitudes were transferred to the rest of Spain through leveling lines and branches that would form the first high precision leveling network (Rednap),” Explain the institute In the report. Obviously, over time and the continuous expansion of the port network in Spain, the Mareographers have changed site, other new ones being built to achieve a continuity of the data series. In fact, today there are 10 operational stations throughout the country: Alicante 1, Alicante 2, Cartagena, Almería, La Coruña, Alborán (Isla de Alborán), Santa Cruz de Tenerife (Island of Tenerife), Puerto de la Cruz (Isla de Tenerife), Los Cristianos (Isla de Tenerife), Puerto del Rosario (Isla de Fuerteventura). Image | Dean Milenkovic In Xataka | Two provinces, four municipalities, three regions: the most complex town in Spain is also that of Feijóo In Xataka | The Júcar Hydrographic Conference has been sacrificing thousands and thousands of fish from its reservoirs for months. And the worst thing is that it makes sense

Apple has taken a step back with the iPhone 17 Pro. It was necessary to make a leap forward

by

Apple has just presented the iPhone 17 PROy, with him, we already have all the letters on the table in terms of high performance processors. The new iPhone sets the soc a19 pro, the Pixel 10 Pro has the G5 tensor and we already knew the Snapdragon 8 Elite of mobiles like the Samsung Galaxy S25 Ultra. Those of the apple carry generations fighting to mark the rhythm in terms of power Thanks to some chips that have less nuclei than those of the competition, but great performance both in CPU and GPU. They are also usually sparse when talking about concrete specifications, such as the frequency of these nuclei, but this year they have been overcome. They have given very few data from the A19 Pro and, nevertheless, They announce how big they are “The most powerful pro models ever created”. The funny thing is that this power of the iPhone 17 Pro can be more marked by the design of the mobile than only by the benefits of the processor itself. Because you don’t have to talk about how premium that is titanium and glass: you have to talk about aluminum. And it makes all the meaning in addition to a great enemy: the covers. THE DELIBLED MYSTERY OF THE A19 PRO Apple is a company that takes great care of communication and does not say a word higher than another without having seriously meditated. Before seeing how they are talking about the power of the new iPhone, let’s see what they wanted to teach the processor. To read something about A19 PRO, you have to go down to the middle of the terminal’s presentation page, where we can see that the soc “It offers the greatest power that an iPhone has had”. On the page characteristicsthey simply detail that it has six CPU nuclei (with two performance and four efficiency), six GPU cores with neuronal accelerators (this is the great novelty of this soc, by the way) and 16 nuclei in its ‘Neural Engine’. Same CPU nuclei and same configuration as the A18 PRO (And that the A17 PRO) and same GPU nuclei, but now each of those nuclei with a support for the locations of the AI ​​at the local level. We have not had the classic “yields x% more than the previous generation, something they detailed before. The only percentage of yield we have had is that of sustained yield 40% better. And this is what must be explained because it is not so much thanks to the A19 Pro … as the design of the iPhone 17 Pro. Steam chamber and aluminum body, performance keys On its website, Apple mentions that “thermal design innovation makes the GPU and CPU achieve up to a 40 % more constant performance“, And all” thanks to their advanced cooling technology and the unibody structure of aluminum “, which has” allowed them to create the most powerful iPhone of all time. “ In the first seconds of the presentation of this model, we saw that Apple focused on showing the new chassis of the iPhone 17 Pro: A unibody body which implies that the sides, the interior and much of what we will have in sight and between hands it has been created from one piece, and not joining several to create the body. It is something that Apple itself has already done in the past and that companies like Tesla, with her gigapressThey also use to build a chassis with the least possible number of pieces. But what attracts attention is the material: aluminum. Apple has a couple of generations selling the idea that titanium and crystal of the iPhone 15 Pro and iPhone 16 Pro It was the culmination of what we could understand as a mobile manufactured with premium materials and … do they return to aluminum after all these years? As paragraphs said: it is the decision that makes sense if they want solve a historical problem. Aluminum is a material that is used, next to copper, when creating the dissipators that we have in consoles, computers and other devices that generate heat. He is a great driver, so the heat of his chips is easily transferred to aluminum and, from there, to the environment, helping to refrigerate the circuitry. Titanium and, above all, crystal … are not great heat drivers. It depends on many factors such as alloy and thickness, but it is estimated that thermal conductivity, in watts per meter and Kelvin, of the three materials is as follows: As I say, it is a measure that depends, but w/mk above or down, it is evident that aluminum is much better dissipating the heat than the other two materials. But if the iPhone 17 pro were only aluminum, it would be uncomfortable in the hand under heavy tasks (or recording video, that there is no need to render something with the iPhone so that TEMPERATE TEMPERATE). That’s where Apple’s other great rectification comes into play: a dissipation system at the height. Last year We already saw that Apple incorporated a structure of graphite to drive the heat of the processor and distribute it better for the rear of the phone to prevent it from concentrating on a single point. It was something necessary, but we also commented that it might not be enough. Thermal strangulation tests, as evidenced: When the iPhone is demanded, it is heated and the SOC, to protect itself, enters what we know as Throttling: its frequency – and performance – low to heat less. When the iPhone returns to normal temperature, the frequency of the processor increases automatically. This generation, finally, have been given and accepted reality, incorporating a rumored steam camerawhich is one of the most advanced systems to dissipate heat. This mechanism It uses a metallic duct full of a low pressure fluid that, when changing state between liquid and vapor, transfers heat to reduce the device temperature. That is why Apple, both in the presentation … Read more

China is no longer made up of moving away from Nvidia. His next step is the heart of the AI ​​with a system that breaks molds

by

In 2017, the Paper “Attention is all you need”Google changed the technical basis of language generation: the Transformers They allowed to process long sequences in parallel and climb models to sizes that were previously unfeasible. That climbing route has driven architectures such as GPT and Bert and has converted self -how The central piece of generative AI Contemporary. But this new approach was accompanied by growing costs in memory and energy when the context lengthens, a limitation that has motivated research to develop alternatives. Spikingbrain-1.0 aims to break molds. Of the “attention is all you need” to the brain: the new commitment to break limits in the A team from the Chinese Academy of Sciences Automation Institute He has just presented Spikingbrain-1.0. We are talking about a family of spiky models aimed at reducing data and computation necessary for tasks with very long contexts. The experts propose two approaches: Spikingbrain-7B, of linear architecture focused on efficiency, and spikingbrain-76b, which combines linear attention with Mixture of Experts (MOE) mechanisms of greater capacity. The authors detail that much of the development and the tests were carried out in clusters of GPU Metax C550, with libraries and operators specifically designed for that platform. This makes the project not only a promising advance at the software level, but also a demonstration of own hardware capabilities. An especially relevant aspect if China’s effort is taken into account for reducing his dependence on Nvidia, A strategy that we already saw reflected with Depseek 3.1. Spikingbrain-1.0 is directly inspired by how our brain works. Instead of having neurons that are always “burning” by calculating numbers, uses spiky neurons: units that accumulate signals until they exceed a threshold and trigger a peak (spike). Between peak and peak they do nothing, which saves operations and, in theory, energy. The key is that not only does it matter how many peaks there are, but when they occur: the exact moment and the order of these peaks carry information, as in the brain. In order for this design to work with the current ecosystem, the team developed methods that convert traditional self -acting blocks into linear versions, easier to integrate into its spiky system, and created a kind of “virtual time” that simulates temporal processes without stopping the yield in GPU. In addition, the Spikingbrain-76B version includes Mixture of Experts (MOE), a system that “awakens” only to certain submodos when we are needed, which we have also seen in GPT-4O and GPT-5. The authors suggest applications where the context length is decisive: analysis of large legal files, complete medical records, DNA sequencing and massive experimental data sets in high energy physics, among others. That lace appears reasoned in the document: if the architecture maintains efficiency in contexts of millions of tokens, would reduce costs and open possibilities in domains today limited by access to very expensive computer infrastructure. But validation in real environments is pending outside the laboratory. The team The code of 7,000 million parameters has released in Github next to a detailed technical report. It also offers a web interface similar to chatgpt to interact with the modelwhich according to the authors are deployed entirely in national hardware. Access, however, is limited to Chinesewhich complicates its use outside that ecosystem. The proposal is ambitious, but its true scope will depend on the community to reproduce the results and make comparisons in homogeneous environments that evaluate precision, latencies and energy consumption in real conditions. Images | Xataka with Gemini 2.5 | ABODI VESAKARAN In Xataka | Openai believes having discovered why the IAS hallucinates: they don’t know how to say “I don’t know”

Windows 10 is one step away from running out of support. Almost half of the world’s computers continue to trust him, even more

by

In just over 40 days, Windows 10 will stop receiving patches and security updates. Microsoft will thus mark the end of support for an operating system that still Almost half of the global market dominates (In Spain, 54%). Windows 11 It has been in the market for several years, but the change has been slower than expected. The countdown is still underway And millions of users apparently have not yet decided how to face the jump before October arrives, although those willing to pay may keep it with extra support. What real options now has the standard user The end of support does not mean that your computer stops working, but it is more exposed. Microsoft offers three main outputs: Make the leap to Windows 11. If your PC has the necessary hardware, the update is free and guarantees updates and new functions. The equipment that does not meet requirements will have to be renewed to continue protected. Pay for extended updates (ESU). With this plan they continue to receive security patches paying an annual fee. It is designed to win margin, not to stay indefinitely in Windows 10. Keep Windows 10 or explore alternatives. Install Linux or using the system without support in controlled environments is possible, but requires assuming risks and planning in the medium term. Compare the situation of Windows 11 with that of its predecessors helps to understand if this slow change is an exception or simply history that is repeated. According to statistaat three years and ten months of life Windows 11 is around 49% of the global market. Windows 10, at that same age, had reached about 57%, and Windows 7 moved in even higher figures, around 61%. At the opposite end, Windows 8.1 barely reached 10%, a reflection of his warm welcome. These data draw a clear pattern: Windows 11 progresses more slowly than its successful predecessors, but leads widely against Windows 10. Everything indicates that the hardware barrier, which leaves millions of old teams out, explains a good part of this rhythm. It is not that users resist by custom, but that the jump to the new generation implies more effort than ever, and that shows in global and European figures. The slow advance of Windows 11 is better understood when looking at the requirements imposed: TPM 2.0 and a closed list of processors leave out millions of computers that were runnings 10 without problems before. For many users, update implies renewing the equipment, and that slows the transition. There are unofficial methods to install Windows 11 on non -compatible PC, But doing so entails risks. The result is a slower adoption and a Windows 10 that is still very alive at the doors of its end. Microsoft Operating Systems Market share at a global level Europe and Spain advance to another rhythm. While the global average shows that Windows 11 is already present in almost half of the teams, in the old continent Windows 10 maintains more than 53% quota and Windows 11 is around 43.9%. Spain follows that same pattern (54% in Windows 10 and 42.61% in Windows 11), with practically traced figures. ANDThe result is a photograph other than the global: Here the version jump progresses more slowly and Windows 10 is still the protagonist at the doors of its support end. Microsoft operating systems market share in Spain Update resistance is not unpublished either, although the current context has nuances. Windows 7 reached its support end in January 2020 with a still significant presence, around 27% of the global market, but already surpassed long by Windows 10. Today the situation is different: Windows 11 leads, but Windows 10 retains almost half of the teams, just over a month of running out of patches and official support. Images | Microsoft In Xataka | If you have an old PC, there is an effective alternative against the requirements of Windows 11 and the Bloatware: Flyoobe works

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.