powerful

Aliexpress’s summer promo starts with very powerful discounts on mobile phones, tablets and more devices

by

There are many users who, when they seek to renew one of their devices, usually expect very specific dates, such as Black Friday. These promos usually bring chollos and offers of all kinds, although the negative part is that they are only held at specific times of the year. The good news is that the different stores also launch promos the rest of the year And now he has touched Aliexpress. From yesterday until June 25 at 23:59 we have available your Summer promoan event loaded with very good offers in technology. In addition to these, we also have at our disposal A series of discount coupons To save even more in our carts: code discount Minimum purchase ES02 2 euros 15 euros ES04 4 euros 29 euros ES10 10 euros 69 euros ES20 20 euros 139 euros SSES30 30 euros 199 euros ES40 40 euros 279 euros SSES45 45 euros 309 euros ES50 50 euros 349 euros ES60 60 euros 439 euros All these discount coupons cannot be used combined nor are they valid for virtual products or other promotions. In addition to the above, we have others available that, in this case, can only be used with products that Miravia sells through Aliexpress: Code discount Minimum purchase Master153 3 euros 15 euros Master5910 10 euros 59 euros MES400 80 euros 599 euros An event loaded with chollos in technology As we say, there are few offers that we have available right now in this summer promo. To facilitate the work, we leave you down a selection of the most prominent that are still available. Little x7 proa smartphone with a great value for money, by 219.99 euros. Bose Quietcomfort SC headphonesof headband and with good autonomy, by 124.29 euros With the ES20 code. OnePlus Watch 3with 1.5 -inch screen and an elegant design, by 183.49 euros With the SSES30 code. Galaxy Tab A9+a very economical tablet to study or work, for 119.99 euros With the ES20 code. Nintendo Switch Pro commandcompatible with Nintendo Switch 2, by 49.66 euros with the ES04 code. Little x7 pro We start with him Little x7 proa telephone that stands out for a very high money ratio (and more, with this offer). 6.67 inches AMOLED screen with 3,200 nits peak shineinsured power with the dimensity 8400 ultra of MediaTak and a considerable battery of 6,000 mAh. In addition, it has a good double camera photographic system. It is available by 219.99 euros. * Some price may have changed from the last review Bose Quietcomfort SC If we are looking for headband headphones, these Bose quietcomfort scs perfect us. They stand out for being light, for offering a sound at brand height And for having a good noise cancellation. In addition, we can also configure them in their ‘Aware’ mode to be aware of what happens around us. Available by 124.29 euros With the ES20 code. * Some price may have changed from the last review OnePlus Watch 3 He One Plus Watch 3 It is a smartwatch with a classic cut design, but has enough characteristics to get our attention. Its 1.5 -inch screen is protected by sapphire crystal and has a peak shine of 2,200 nits, so we will be able to see it out of problems without problems. We cannot forget that it also has a good number of sensors to be able to control our health and our training to the millimeter. It comes out by 183.49 euros With the SSES30 code. * Some price may have changed from the last review Galaxy Tab A9+ You don’t have to spend a fortune to have a tablet at home and this Galaxy Tab A9+ It is the test. It is a manageable and comfortable device to use, both for work and leisure. It has a 11 -inch screen (with 90 Hz) and a 7,040 mAh battery to offer us good autonomy. Besides, It is compatible with Dolby Atmos soundideal to watch series or movies. We can acquire it for 119.99 euros With the ES20 code. Galaxy Tab A9 + (4 + 64 GB) * Some price may have changed from the last review Nintendo Switch Pro command We close with the Pro Nintendo Switch Uno command, which It is compatible with Nintendo Switch 2. It is a wireless and very comfortable command to use, ideal if we seek to have a traditional experience in the Nintendo console. The best thing is that the summer promo leaves it at a price that is not easy to find: it goes out for 49.66 euros with the ES04 code. Pro Nintendo Switch command * Some price may have changed from the last review Some of the links of this article are affiliated and can report a benefit to Xataka. In case of non -availability, offers may vary. Images | Aliexpress, little, Bose, OnePlus, Samsung, Nintendo In Xataka | The best smartwatch: their analysis and videos are here In Xataka | Best wireless headphones. Which to buy and 19 models from 20 euros to 450 euros

The first quantum computer with error correction already has date. It is close and will be 20,000 times more powerful than the current

by

It seems that in the end Israeli mathematician Gil Kalai was wrong. This professor at Yale University has been very critical for several years with Quantum computing. In fact, he argues that the increase in the number of states of quantum systems and their complexity will cause They end up behaving like classic computersso the superiority of the former will end up evaporating. It defends, in short, that totally functional quantum computers They will never arrive. However, we currently have very solid reasons to anticipate otherwise. And there are several companies that ensure that they will have a quantum machine equipped with the ability to amend your own mistakes Before I expire this decade. These computers will be very important because they will presumably have the ability to solve a very wide range of problems with which not even the most powerful classic supercomdators available can deal. Just a week ago we told you that Xanadu, a young Canadian company founded in 2016, Plan to have ready before 2030 A quantum computer of one million phototonic ulna with error correction. However, it is not the only company that intends to make this milestone come true. IBM plans to make available to its customers in 2029 ‘Starling’, its first large -scale quantum computer equipped with the ability to amend their own mistakes. It will execute 20,000 times more operations than current quantum computers IBM is going to build The quantum computer ‘Starling’ in a new data center that will be housed in Paughkeepsie, New York (USA). This machine will bring together 200 logical cubits that, in theory, will allow you to execute 100 million quantum operations. Logical cubits represent a way to overcome the difficulty involved in the use of hardware or physical cubits, which are extremely noise sensitive, and, therefore, prone to make mistakes. Each logical ul Each logical cubit is constructed abstractly on several physical or hardware cubits, so that a single logical cubit encodes a single cubit of quantum information, but with redundancy. It is precisely this redundancy that allows detect and correct errors that are present in physical cubits. Until very recently the number of hardware cubits that was necessary to implement a single logical cubit immune to errors was impracticable, but IBM says it has found the solution to this problem. Their engineers have published two very interesting technical articles in which they develop the strategy they will use to bring to fruition their large -scale quantum computer and error correction. In the first These articles explain what procedure will continue to process instructions and execute operations in an efficient way. AND In the second They describe how they will decode the information that physical cubits contain in an efficient way, and also what strategy they will use to identify and correct real -time errors using conventional computational resources. As I mentioned a few lines above, IBM says that ‘Starling’ will be ready in 2029, but before its arrival this company expects to reach other important milestones. At the end of 2025 ‘Quantum Loon’ will prove the technology used to connect cubits to a larger distance Within a single chip. In 2026 ‘Kookaburra’ it will be the first IBM modular processor designed to store and process coded information. In 2027 ‘Cockatoo’ will arrive to demonstrate that it is possible to intertwine two ‘Kookaburra’ modules, and, consequently, that it is viable to connect several quantum chips such as nodes of a larger platform. And finally, after ‘Starling’ will arrive ‘Blue Jay’, which will bring together 2,000 logical cubits that, in theory, will allow you to execute 1 billion quantum operations. Image | IBM More information | IBM In Xataka | Bitcoin encryption and other cryptocurrencies will fall. And those responsible will be quantum computers

The most powerful telescope in Europe has seen something rare while observing two brown dwarfs: a perpendicular planet

by

Today, in things that we did not know that there were in other star systems: a planet that orbits perpendicular to its two stars. It is as strange as it sounds. A different planet. We have seen everything in the exotic exoplanet gallery that the observable universe offers us, but a new surprise has forced astronomers to readjust their expectations. Using the Vary Large Telescope del Southern European Observatory (ESO) In Chile, astronomers have found the first exoplanet known in a polar orbit around a binary system of stars: two young brown dwarfs. Orbiting at 90 degrees. Baptized as 2m1510 (AB) B, this world has two Tatooine style soles, something extremely common when it comes to more massive stars, but not so much in the case of brown dwarfs. However, what is common is that orbits its two host stars in a way that until now had only theorized: turning perpendicular to them, in a polar orbit of 90 degrees. We sensed that they could exist. Astronomers had already detected planetary training discs in polar orbite and theory suggested that They could form stable planetsBut finding them was another story. 2M1510 (AB) B is the first credible test that this configuration exists. And the most curious thing is that the team was not actively looking for this type of planet, they found it while refining the orbital and physical parameters of the two brown dwarfs, seeing that the path of the two stars was being pushed and pull in unusual ways. It is not a usual system. The planet is not the only rare in this neighborhood, taking into account the host stars. Brown dwarfs are larger than giant giant planets, but too small to maintain nuclear fusion which characterizes the stars “really”. But the rarity does not end there. These brown dwarfs form an eclipsest binary, which means that from Earth we can see how they hide each other. 2M1510 (AB) is an incredibly infrequent system: the second pair of eclipsessing brown dwarfs known to date. With more than 5,800 exoplanets confirmed to date, only about 16 orbit around two stars. That one of them does precisely around a system as atypical as a binary of eclipsessing brown dwarfs, and with a polar orbit, it is a real cosmic jewel. Image | THAT In Xataka | These real images were unthinkable before the Webb Telescope: they are planets orbiting other stars to 130 light years

There is a career to achieve the cheapest and most powerful model. And against all prognosis, China is winning it

by

On May 20 Google launched Gemini 2.5 Pro and Gemini 2.5 Flash in preliminary version. These new AI models were better than ever, and to demonstrate the company included in its announcement several graphs and comparative tables. They looked at how both surpassed their rivals both in the field of reasoning and in the traditional performance (Benchmarks of Mathematics or Programming), but there was also another fact that Google presumed: Google: The cost of Gemini 2.5 Flash. Source: Google. That table published by Google made it clear that Gemini 2.5 Flash It was clearly the winner of that comparison in the important price/benefits ratio. What Google did not say is that this success of this model was the exception to the rule, because in that race for having cheap and powerful models, China seems to take the lead. He does it at least if we attend to the cost of using these models. In Xataka we have analyzed that cost based not on the price of subscriptions for end users, but on the cost of access to the API, which is the one that allows developers to integrate these models into their own chatbots and their services. The API prices of each model clearly differentiate two uses from artificial intelligence. On the one hand, how much does it cost to write something for the model to then process it (the so -called input tokens). On the other, how much does the text generated by the model once processed the answer (the so -called output tokens). The entrance tokens They are usually five times cheaper than exitbecause processing the request and generating text is much more expensive than receiving it, analyzing and “understanding it.” We wanted to compare the cost of the main models of the AI ​​developed in China and those of the US, and although as always are not all that are, if they are all. The resulting table is as follows: These prices are public and very easy to find in the case of US AI models (OpenAI, Anthropic, Google) but not so much in the case of Chinese models (Deepseek, Qwen (Alibaba), Doubao (Bytedance), GLM-4 (Zhipu), Ernie (Baidu)). Be that as it may, the table, ordered from the cheapest to the most expensive, demonstrates that today Chinese models are especially cheap. Only Gemini 2.5 Flash Preview manages to compete – and does it exceptionally. In the rest of the cases, the AI ​​models in China win the battle for cost. It must be noted that, Like all comparisons, this is unfair. And it is because that table does not take into account the benefits of each model. Openai O3 and Anthropic Claude Opus 4, the last and most powerful models of those companies, are especially precise in their answers, but each consultation consumes much more resources (computation, energy) and that makes it logical that they are much more expensive than their competitors. But these models are also designed for very special cases and for specialized, detailed and deep consultations. In the vast majority of cases it is not necessary to use these models, and that is where they are competing for example Deepseek R1 or Gemini 2.5 Flash Preview: in the price/benefits ratio. Models with variable prices That price battle has made us see in recent times two techniques that some companies are applying to the prices of use of their APIs. The first one is to differentiate normal entries and outputs of inputs (and even exits) cache. Deepseek API prices. Attentive to the lower left: according to the time you use them, they can leave cheaper. Source: Deepseek. The explanation is simple: a “normal” entry is a request or question that the model has never processed and therefore has to process completely. If the entrance has been caught (Cache hit) is because that request has been processed in the past, so the system can obtain the response of its cache, which significantly reduces the computational costs. Deepseek, Google, Anthropic and Openai offer this type of option, as can be seen in the table. The second technique is to use variable prices according to (at the moment) the time slot in which we use these models. This is what Deepseek has done, which has “day” and “night” prices according to UTC schedule. If you use the Depseek API from 18:30 to 2:30 (Peninsular schedule in Spain), it will be half price. Good news: AI is every time (much) cheaper While China and the US fight who has the most powerful model or who has the cheapest model, what is constantly happening is that the AI ​​price is falling remarkably. It is an observation that several experts such as Ethan Mollick, a professor at the University of Pennsylvania who recently analyzed how That price/benefits ratio does nothing but improve. The models are getting better and cheaper. Other experts such as Raveesh Bhalla – explained by Netflix and LinkedIn – also reflected this evolution at the beginning of the year. Then he showed how the cost of an O1 level model had dropped 27 times in the last three months. Moreover, at this rhythm the GPT4 level models – which a year ago were absolute referents – will be reduced 1,000 times in just 18 months. We are living it in price reduction. Dane Bahey, from Openai, said at a conference in September last year how the cost per million tokens had fallen from 36 dollars At just 0.25 dollars In the last 18 months. And that price drop is still clear and fantastic for users. Thus, we are facing a career that at the moment has a lot of stretch: China’s models carry the lead if we attend only to their cost, but careful, because we must also take into account the benefits. It is true that these Chinese models have already shown in the benchmarks that have been showing that they compete from you to you with the best US models, and now it remains to … Read more

It aims to go beyond what ai offers today at WhatsApp with these powerful functions

by

It all started with a promise. Goal announced more than a year and a half ago that would integrate generative artificial intelligence into WhatsApp. And he did. Goal AI was deployed first in the United States, then in other markets, and finally, in March 2025, In the European Union. Today it is possible to ask you things, ask you to generate images or summarize content. Of course, some functions, such as the creation of images, are not yet available in the old continent. Telegram, on the other hand, has opted for a more direct route. Instead of developing its own technology, it has signed an agreement with XAIthe artificial intelligence company founded by Elon Musk. The result is Grokan assistant with ambitious aspirations that already works on the social network X and now prepare a greater deployment in Telegram. Its objective is not less: integrate AI in user’s day -to -day life. Telegram supervitaminated with AI In a video shared by Pavel DurovFounder and CEO of Telegram, shows how Grok will be present constantly. It can be top of the conversations and will appear in the search engine, prepared to answer doubts or generate content directly from there. Something similar to what we find in WhatsApp. From there, the functions are raising level. Grok can generate images from the application itself, as a goal is already in other regions, but will also go further with assisted writing tools. That is, you can reformulate messages directly in the text field. For example, if we write “Great!” and the system would suggest something like “excellent teamwork. Keep it up!” It may seem lower, but introduces a customization layer that takes another step in the naturalness of the conversation. The list continues. The wizard can also generate animated stickers, summarize groups of messages, links or PDF documents. And those summaries will not only be displayed on the screen, but they can also be read aloudfacilitating content consultation while we do other tasks. In groups, Grok also aims to play a role as a moderator, although for now there are no details about its operation. What has been shown is a possible verification system in the public channels of Telegram. In the examples shared by the company, the AI ​​resorts to Fuentes such as Bloomberg to analyze and validate the published information, although it remains to be seen how, and if, this function will really be implemented. While WhatsApp keeps moving step by step, Telegram bets on a more ambitious integration. And he does it with an AI that not only responds, but begins to get involved actively in the conversation. It is expected that Grok’s deployment in Telegram will begin in Summer of 2025although for now it is not clear if all the functions will arrive at the same time, if they will be activated gradually or what will be the road map for availability in the EU. Images | Telegram/XAI In Xataka | Openai has just demonstrated that the AI ​​no longer completely obeys: O3 sabote its own shutdown to continue working

Elon Musk turned an abandoned US factory into the most powerful supercomputer in the world. Nobody thought of the neighbors

by

It would be said that Elon Musk has created A perfect circle Around what until recently it was an abandoned factory on the outskirts of Memphis. There has installed Colossusthe most monstrous supercomputer on the planet to boost AI to new limits and that it defines the future of cars (of Tesla). In order for the combo, it is complete to the richest type in the world, everything is in turn fed by the megabaterías of Tesla. A Win Win For Musk’s framework, although with a problem: the neighbors. A colossus … toxic. The story brought it This week the CNN. In the summer of 2024 Musk transformed an old abandoned factory in southwest Memphis into what he himself proclaimed as “the most powerful supercomputer on the planet.” The project, promoted by its artificial intelligence company XAIpromised to turn the city into a New Technological Mecca (The so -called “Delta Digital”) with Quality jobs and dozens of millions in taxes. However, for Botxtown residents, a mostly black and impoverished community that has lived with decades with Industrial pollutionXai’s arrival has meant a déjà vu environmental: A new source of pollution settled without clear permits, with an apparent contempt for public health. A computer that consumes. To feed “Colossus”, XAI installed 35 gas turbines capable of generating Up to 420 megawattsreleasing toxic gases Nitrogen oxidesultrafine and formaldehyde particles. The problem? Who did Without licenses of air required, welcomed to a legal exemption for temporary machinery, which according to experts does not correspond to it. The area already houses 17 polluting facilities, and various studies indicate that the risk of cancer in the area quadrupple acceptable levels For the EPA. Memphis also has the higher rates of children’s hospitalizations by asthma in all Tennessee. Realities. While the mayor of Memphis, Paul Young, celebrated the transformative potential of the project and anticipated more technological investment, local leaders such as state representative Justin Pearson They have denounced have been excluded from the process. The lack of transparency adds to an obvious regulatory collapse: an installation with the power of an electric plant operating without permits in the middle of a residential neighborhood. To this we must add the most recent thermal snapshots (image below) that indicate that at least 33 of the turbines They were operational in April. Following the controversy, XAI finally requested permission for 15 of them and withdrew 12, but, as CNN counteddamage to trust is made. Promises Project defenders say that “leading standards in emissions” will be achieved, but residents see the employer repeat: Employment promises well paid that do not specify (because the reality is that the data centers use very few), while the environmental load falls on those who have less resources to defend themselves. Plus: The story of Botxtown It is not new. Already in 2021, its inhabitants achieved stop a pipeline that would cross their lands, and in 2023 they closed A sterilization plant that emitted ethylene oxide. For them, therefore, XAI is simply the last chapter of a long struggle for the right … to breathe. Innovation or regression. It is the last of the legs to be treated with the controversy. XAI installation reflects a broader national dilemma than We have counted before on the rise of artificial intelligence and its real cost. Amid the enthusiasm for turning the United States into the “Global Capital of AI” (according to New EPA Guidelines Under the government of Donald Trump), the expansion of data centers Devoradors of energy advances without a serious evaluation of its environmental implicationsespecially in vulnerable communities. The unconditional support of the Executive to Musk, one of Trump’s closest advisors, coincided with the Weakening of environmental policiesthe elimination of ecological justice programs and a rhetoric that prioritizes economic efficiency over human health. The first “stone.” The contradiction seems clear: IA is promoted as the future, but it is fed with fossil technologies of the past, generating private benefits while the risks and damage are socialized. “If innovation chains you to fossil fuels, that’s not progress,” Keshaun Pearson remembereddirector of Memphis Community Against Pollution. Thus, residents fear that what happens in Memphis is just a general essay of what could soon be replicated in similar neighborhoods throughout the country. A tireless struggle. He counted An NBC report That in Botxtown, indignation coexists with fatigue. Many, such as Sarah Gladney (respiratory and resident patient to a few km from the installation), feel they live in a perpetual battle. The possibility of a second meginstalization of XAI, already projected in the city, only increases the sensation of siege. “It seems that we are always at war,” He underlined. A paradox, since while local officials speak of economic transformation, neighbors simply speak of survival. In the background, the collision between the promises of peak technology and the old reality of systemic pollution raises an uncomfortable question: Who pays the price of this digital revolution? In the southwest Memphis, the answer seems sadly clear. Image | Southwings for the Southern Environmental Law Center In Xataka | Musk has created the perfect circle: Tesla’s megabers feed the AI ​​that will define their next cars In Xataka | Some researchers have disassembled the batteries of Tesla and Byd. You know which one yields better and is much cheaper

The ice age had a solar storm so powerful that its effects can still be detected in trees

by

How powerful can it become a solar storm? More than one will have ever asked this question for mere curiosity. And also out of necessity. Answering this question is important in order to prepare in the face of potential storms of great magnitude, but it is complicated: we have a limited sample and the most intensity events are those that happen less frequently. But now we have a new track. The mother of all storms. A recent study has analyzed The remnants of the greatest solar storm of which we have record to date. The storm, about 500 times more powerful than the largest solar storm since the beginning of the space era, would have happened about 14,300 years ago. Although We already had enough previous clues about this eventThe new study facilitates the work of preparing for such events, and can also help us improve our radiocarbon dating techniques. 14,300 years ago. The discovery of this solar storm has been possible thanks to the detection of a radiocarbon peak, the well-known carbon-14, happened towards the year 12,350 AEC, towards the end of the last glacial period. This implies that the storm is not only the most powerful of which we have record, it is also the only solar storm known outside the Holocene, the contemporary geological era (if we exclude the existence debated from the anthropocene). Different storms. The storm analyzed was a solar particle storm. There are different events That we can catalog as solar storms, each with its characteristics, such as radio blackouts, solar radiation storms or geomagnetic storms. Radiation storms, such as the one studied, are produced when large amounts of charged particles from the sun reach the magnetic field of the earth. This field tends to divert the particles towards the poles, making the effect on high latitudes greater. Carbon 14. The team responsible for the new study He turned to a new model Chemist-Climatic, Socol: 14C-EX, for analysis. This is a model designed to rebuild solar storms of particles in climatic conditions of the past glaciations. Thanks to this model, they explain, it was possible to verify that this solar storm was 18% stronger than the event of the year 775 EC, the largest solar storm known until the discovery of this event. “Compared to the largest event in the Modern Era of Satellites (the 2005 particle storm) the old 12,350 AC event was about 500 times more intense, according to our estimates,” explained in a press release Kseniia Golubenko, co -author of the study The details of the study were published In an article In the magazine Earth and Planetary Science Letters. More than establishing a record. The study allows us to establish a new framework to the “most pessimistic scenario,” says Golubenko. Knowing what we can face gives us essential tools when preventing this type of event. The study can also help scientists who study something very different: archaeologists. The analysis has been based on carbon-14, a very important isotope when it comes to very diverse organic matter remains, from fabrics to ships. Understanding the peaks of this isotope caused by the impact of loaded particles from the Sun can help us to date back objects created in past times. In Xataka | NASA has calculated how much time we would have to prepare before a devastating solar storm and has got to work to get that time Image | NASA/GSFC/CIL

In case they were happening few things, the sun has just caused a radius blackout with its most powerful eruption of 2025

by

The sun has decided to remind us that we are in their hands. A new active region of sun spots, called AR4087, has released this morning a solar eruption of class X2.7, the most powerful so far this year. The sun was not asleep. Eruptions, glows or solar flares are classified as Five categories According to its intensity: A, B, C, M and X. Each letter represents events ten times more intense than the previous letter. An eruption X2.7 is located at the lower end of the most powerful category, but remains an important and unusual eruption. Observatories that constantly monitor the solar crown They detected it on May 14 at 08:25 UTC (10:25 in Spain). The X2.7 class flares has been the strongest of 2025, demonstrating that the sun is still awake After last year’s activitywhen it reached the solar maximum. Radio lock. Although it was not directed directly to the Earth, the eruption launched an intense X -ray pulse and extreme ultraviolet radiation that reached our planet at the speed of light. When impacting with the upper atmosphere of the Earth, this radiation has rapidly ionized, interfering with the high frequency radio signals (HF) on the visible face of our planet. As a result, a R3 level radius blackout was detected (cataloged as “strong”) in Europe, Asia and the Middle East. The show begins. The sun stain AR4087 remains very active and Soon he will look directly to the Earth As the sun revolves on its own axis. In the last hours he has produced class rashes M5.3, X2.7 and M7.74. The big question now is whether the most powerful eruption was accompanied by a coronal mass ejection, the famous solar plasma and magnetic field languages ​​that, when impacting with the Earth’s magnetic field, can cause geomagnetic storms and intense auroras to different latitudes, Like the historic Red Auroras of 2024. At the moment, the AR4087 region is at the visible edge of the sun, so the earth seems to be out of reach. Image | Spaceweather In Xataka | Webb and Hubble telescopes watched Jupiter’s auroras at the same time. The problem is that they did not see the same

Huawei plans to advise Nvidia in China. It has a new GPU for theory that in theory is extremely powerful

by

Huawei is putting all the meat on the grill to absorb so much share in the Chinese GPU market for artificial intelligence (AI) as I can. And it is that the entry into force of the last US sanctions package is compromising with all probability Nvidia leadership in China. The US Department of Commerce It has imposed restrictions to the export to the country led by Xi Jinping of The H20 GPUand this in practice means that this chip presumably will not reach the Chinese clients of Nvidia. This last company has announced that this ban will cause a hole in its accounts of 5,500 million dollars Due to the commitments linked to the H20 GPU that had already acquired the reserves of this chip that it will not finally satisfy. Some of the Chinese companies that have bought large amounts from the H20 chip to NVIDIA and who presumably planned to continue doing them are Tencent, Alibaba or Bytedance, but at the current situation they will have to resort to an alternative. And Huawei has put it on a tray. The GPU Ascend 910D aspires to snatch the leadership in performance from Nvidia Huawei reacted immediately to US sanctions. And is that just a few hours after the entry into force of the new regulation of the Department of Commerce He presented his chip for the ascend 920a solution that is clearly destined to occupy in the Chinese market the gaps that the NVIDIA H20 GPU is going to leave. This proposal will enter large -scale production during the second half of 2025 using 6 NM integration technology that have presumably developed elbow with Huawei elbow and SMIC. Until now Huawei wanted to get his hardware to dominate the inference processes in AI However, this is not the only asset that Huawei has to increase its market share both in China and beyond its country of origin. And is that, According to Reutersthis company is preparing to start the testing and validation phase of a new GPU for AI: The Ascend 910D chip. Unlike the GPU Ascend 920 that, as we have seen, presumably aspires to compete with the NVIDIA H20 chip, the GPU Ascend 910D seeks to overcome the performance of the chip NVIDIA H100. If this movement is confirmed, already priori this information is reliable, it will be evident that Huawei will have chosen to fight in all hardware market segments for the nvidia. Until now this Chinese company wanted to get its hardware dominate the inference processes in AIand not the training of the models, as Georgios Zacharopoulos, a senior researcher of AI who works on the acceleration of inference in the Huawei laboratory in Zurich (Switzerland) points out in this statement. “The training is important, but it only happens a few times. Huawei focuses mainly on inference, which will ultimately give us access to more customers,” says Zacharopoulos. Inference is broadly the computational process carried out by language models with the purpose of Generate the answers which correspond to the requests they receive. In any case, the information we have reflects that the GPU Ascend 910D will allow Huawei to compete with the chips for the most advanced NVIDIA both in inference and in training. Image | Huawei More information | Reuters In Xataka | In a low voice, China has begun to remove some tariffs from US products. Your concern: the chips

No one is safe from tariffs. Not even one of the most powerful and necessary companies for the US: ASML

by

ASML occupies a privileged position. This company of the Netherlands is the only company on the planet capable of producing equipment of extreme ultraviolet photolithography (UVE) that are necessary to manufacture avant -garde chips. In fact, these are the machines that use TSMC, Intel, Samsung, SK Hynix and Micron Technology to create Your most advanced integrated circuits. It is currently not exaggerated to affirm that a very important part of the semiconductor industry is supported About Asml’s shoulders. And, despite its indisputable relevance, it is not immune to tariffs. Or, at least, to the uncertainty that has triggered the government’s tariff strategy led by Donald Trump. In fact, According to Reutersthe investors of this Dutch company are very concerned about the impact that US taxes will have on the company’s economic performance. The next tariffs on chips have plunged Asml in uncertainty The US administration has excluded manufacturers of lithography equipment from the base tariff of 10% announced just a few days ago. However, ASML, Tokyo Electron, Canon and other companies that produce these machines have reason to be expectant. And worried. A little over 24 hours ago Donald Trump confirmed that during the next few days will announce the tariff rate with which it will finally tax the imported integrated circuits. He also anticipated that some companies in the semiconductor industry will have some flexibility, although he did not specify either what will be those organizations or to what extent they can continue to import integrated circuits avoiding tariffs. The problem is that with these statements has plunged the entire semiconductor industry in uncertainty. This is the problem facing ASML and the other machine manufacturers to produce integrated circuits. If the US government finally decides to impose tariffs on the importation of lithography equipment manufactured abroad, the semiconductor sector will suffer If the US government finally decides to impose tariffs on the importation of lithography equipment manufactured abroad, the semiconductor sector will suffer. ASML is in a more comfortable position than other chips processing equipment manufacturers because, as I mentioned in the first lines of this article, it is the only company that produces UVE lithography machines. And not having competition in this segment helps. In addition, Intel and TSMC are equipping, or, at least, they plan to do it in the short or medium term, Its new plants of manufacture of semiconductors in the US. And it is impossible to start them Without ASML machines. However, the panorama is gloomy even if the US administration decided to exclude tariffs definitively to the manufacturers of lithography equipment, which is something that can happen. If this last scenario was given, but the US government decides to impose tariffs on imported chips, which right now seems most likely, the business of some of the best ASML customers, such as TSMC, Samsung or SK Hynix, presumably will resent to a greater or lesser extent. And, consequently, it is reasonable to assume that ASML’s economic performance will degrade. This is what worries your investors. TSMC and Samsung They already have factories in the USand SK Hynix is ​​building a plant in West Lafayette (Indiana), but in the short term they will hardly be able to meet the needs of the US market Only with your US plants. We will see what happens finally, but whatever happens is evident that uncertainty is already hurting even an industry as robust as that of integrated circuits. Image | ASML More information | Reuters In Xataka | Nvidia will continue to sell its H20 GPU in China. It has cost a dinner of 1 million dollars per diner

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.