chips

The true bottleneck of AI is not chips but electricity. And there China has a great asset

by

Data centers are Electricity devourers. A single training cluster can spend more light than 100,000 houses. And while we become obsessed with the semiconductor war, the main problem is still there, in silence: electricity. Why is it important. Training the most powerful models require superoring workers 24 hours for whole weeks. It doesn’t matter if you have the best chips in the world if you don’t have electricity to feed them. It is like having a Ferrari with hardly any gasoline. In figures: The context. The United States has an still incontestable domain in AI and controls 75% of world computing capacity. But their companies are being delays with years to connect new data centers to the electricity network. Google, Microsoft and Amazon They have to deal with years of delays to connect new data centers to the electricity network. China, on the other hand, can lift and connect plants in a few months. Unidso’s bottleneck is in the slowness to connect new capacity to the network. Between the lines. American chip restrictions have forced China to be more efficient. He Deepseek earthquake At the beginning of the year it was the best test. China is riding its “National Integrated Computer Network”, connecting public and private data centers. Your “East Data, West Computing” plan “ Build eight big hubs in provinces with cheap renewable energy. It is something similar to what we saw at the time with the cryptocurrency mining. The key: China puts the AI ​​where there is electricity. Yes, but. Coal is still 58% of its mix energetic. Although renewables is expanding at a good pace, its current advantage is based partly on fossil fuels. That questions the sustainability of its strategy. And now what. The AI ​​career will be decided both in power plants and in laboratories. China is building the infrastructure to feed gigantic models. The United States has better chips, But worse electrical infrastructure. AI needs two things: Silicon. Electric muscle China is betting very strong for the second and that could be its winning asset. Outstanding image | ダモ リ in Unspash In Xataka | China’s three master moves to “independent” technologically from the West: raw materials, chips, AI

Chips manufacturers seek talent urgently. In Europe only 100,000 more engineers are needed

by

During the next five years the global semiconductor industry will need to incorporate nothing less than One million qualified workers. This prognosis is no elucubration; It comes from SEMIan international organization that watches over the interests of the electronics industries and integrated circuits. According to their forecasts Europe will face a deficit of 100,000 engineers, and Asia will need 200,000 qualified technicians. These a priori figures may seem exaggerated, but they are not at all if we consider that for 2024 the chips industry grew by 19.1% compared to 2023 thanks to the demand for GPUs for artificial intelligence (AI) and consumer electronic products, as well as to the expansion of 5G communications throughout the planet and the development of the car market. In 2024 the global semiconductor industry invoiced 627.6 billion dollars. There are not enough professionals to support the growth of this industry TSMC, The biggest chips manufacturer on the planethe goes hunting again year after year to be able to meet his needs. During 2023 recruited 6,000 engineers For its Taiwan facilities, and presumably this trend also remained for 2024. And between 2025 and 2028 it will start several semiconductor manufacturing plants in the US, Germany, Taiwan and Japan. TSMC is one of the most successful companies in this sector, but with all probability other chips designers and manufacturers will also need to strengthen their templates. The average salary of an engineer without previous experience, backed by a master’s degree and newly arrived at TSMC exceeds $ 65,500 annually In this situation the salaries offered by these companies are very high. The average salary of an engineer without previous experience, endorsed by a master’s degree and newcomer to TSMC exceeds $ 65,500 annually (Approximately 56,000 euros), But this is just the starting point. It is assumed that as their salary acquires experience. The problem facing semiconductor companies, According to semiis that as many people with technical profile are not being formed in universities as they will need in the short and medium term. In addition, many of the most experienced engineers are retiring or will do so before 2030. As a button shows: in the US, a third of employees of integrated circuit companies have 55 years or more. And in Germany a third of the technicians who have developed their work career in the chips industry will retire throughout the next decade. However, there is another challenge that also compromises the future of these companies: the next batch of engineers will have to have advanced skills in AI and Automatic learning. The companies that are dedicated to the semiconductors are aware of the problem that already looms on them, which has caused some to have launched initiatives that initiatives that initiatives that initiatives that They go beyond offering good salaries. Some of these measures are to invest in the progression of their professionals and offer them flexibility to prevent them from leaving; in seeking candidates with non -traditional profiles in which their skills prevail and not their training, or in promoting the incorporation of women into this industry. At the moment Women represent only 17% of technical positions in the semiconductor industry. Image | TSMC More information | SEMI In Xataka | We already know what the chips that will arrive until 2039 will be. The machine that will manufacture them is close

An existential threat looms over several US chips companies. And China is just one of its problems

by

Several US companies involved directly in the integrated circuit industry are suffering. Some of them, in fact, They are flirting with bankruptcy. In this context it is inevitable not to think of Intel and The difficult stage that crosses Currently, but from the company Led by Lip-Bu Tan We have spoken a lot in recent months. The companies in which we are going to investigate in this article are smaller, but, even so, they play a very important role in the US chips industry. The worst is happening is, without a doubt, Wolfspeed. And this company of Durham (North Carolina) It has officially declared in bankruptcy. Its specialty is the development and manufacture of silicon carbide (sic) and gallium nitride (GAN) semiconductors for high -power electronics applications, which positions them as extraordinarily attractive integrated circuits for the electric car industries, renewable energy, data centers, fast -charged infrastructure and aerospace and military industries. Robert Feurle, the general director of Wolfspeed, has declared That “after evaluating the options available to strengthen our balance and adjust our capital structure we have decided to take this strategic step (declare in bankruptcy and look for new financing routes) because we believe that it will place Wolfspeed in the best possible position for the future.” An interesting note before moving forward: Spain has launched The Godic project with the purpose of contributing to the independence of Europe of silicon carbide semiconductors from China and the US. China is not the only cause of the problems of American companies Wolfspeed is not at all the only company in the US chips industry that is in difficulties from a financial point of view. The American subsidiary of the Israeli company Tower Semiconductor, Globalfoundries, on semiconductor, Skywater Technology or Ayar Labs are some of the companies that are also facing economic difficulties. The competition from China is subjecting them to a pressure that does not cease to increase, but the competitiveness of Chinese companies is not the only factor that is putting in trouble some US chips companies. The competitiveness of Chinese companies is not the only factor that is putting in trouble some US chips companies Tariffs linked to Imports from China which has deployed US administration seeks to protect the business of US companies within your market. However, most of the companies that I have mentioned in the previous paragraph also sells their semiconductors outside the US. And the competitiveness of Chinese companies in other markets is very high due mainly to two crucial factors: their prices are moderate thanks to contained production costs and the quality of their products is comparable to that of their western and Asian competitors. In any case, the economic problems faced by some designers and manufacturers of American chips do not have their origin solely in China’s competition. Wolfspeed is at a very delicate moment because, beyond the pressure exerted on it, its Chinese competitors, The global demand of the high -power semiconductors it produces has fallen. In addition, during the last five years He has borrowed To be able to build new integrated circuit factories, and currently you cannot assume this economic burden. The panorama of other US companies that are dedicated to chips is not very different to the context with which Wolfspeed is dealingso it is possible that in the short or medium term other companies also enter in bankruptcy. Image | Xataka More information | GLOBALDATA In Xataka | The US is willing to do anything for advanced chips not to reach China. And Malaysia is an obstacle

Huawei states that he is staying behind the chips. It is just what we want us to create

by

Huawei doesn’t like to get attention. He doesn’t want headlines. It is not generating too many and since 2019 The US veto and google occur The company has adopted a low profile and a shy, silent narrative. But as long as he does, Huawei does not stop growing and consolidating his role as one of the most important technology companies in the world. The silent leader. Ren Zhengfei (Anshun, 1944) founded Huawei in 1987 and is not only the company’s CEO: he is considered a spiritual leader. He doesn’t see it like that and affirms which is rather “a puppet leader (…). It only develops a symbolic role, like an idol of cement in a temple. Without it the temple would seem empty, but in reality the idol does nothing … that I am in Huawei or in Huawei has no real impact.” Who wants to be forgotten. Despite its absolute relevance, whether symbolic or not, Zhengfei has no intention of being in the foreground. His hope is in fact end up being forgotten. “I am just an old man. What is the point of reminding me? People should think more about the future and the world. My greatest desire is to have a coffee in a cafeteria without anyone noticing.” Zhengfei is known for his reserved and discreet character, and has transferred that way of being to his company. “Our chips are a generation behind the US”. The phrase is from the aforementioned Ren Zhengfei, which In an interview In the Chinese state media People’s Daily talked about his latest advances in semiconductors. There the head of this giant explained that the sanctions that prevent Huawei from obtaining or manufacturing state -of -the -art chips cause that “our chip continues a generation behind the US.” Better not believe it much. Huawei is Property of your employees And he has ties with the Popular Chinese Liberation Army, the country’s armed forces, of which Ren Zhengfei was part for several years. The commercial war has caused the company to adopt a very discreet profile, and REN’s comment on the state of its chips is possibly a strategy not to attract attention. Because although there may still be differences with the advances of the US, the Chinese giant does not stop expanding. For all without being noticed. I explained it William Huo, an expert who explained in China how Huawei is currently 25,000 million dollars a year on R&D. One of the results of these investments have been New Ascend chips for ia that has not been designed to overcome Nvidia chips, but to climb big. And there is a blunt example. Source: Semiianalysis Cloudmatrix 384. Huawei chips may not be so powerful, but their clusters are. It is demonstrated by the cloudmatrix 384, direct competitor of the NVIDIA GB200 NVL72. This monster has 384 GPUS Ascend 910C connected through an optical network that allows an interconnection at speeds much larger than those offered by copper cables of the Nvidia cluster. But it is also a complete cloudmatrix 384 system, as indicated in Semiianalysisoffer up to 300 flops (in BF16 accuracy) of computing, almost double what the NVL72 GB200 of NVIDIA offers. Not only that: it has the memory capacity 3.6 times and 2.1x times the memory bandwidth. The Huawei machine is an absolute beast, no matter how much its chips “are a generation behind” – which is debatable. The confusion. A few weeks ago there was talk of how the new laptop Huawei Matebook fold Huawei would be based on the promising Kirin X90 chip. That soc, manufactured by SMICit seemed to be manufactured with 5 Nm photolithographic node, but that was difficult. Especially since SMIC – La Foundry China that Try to follow the steps A TSMC or Samsung – has no access extreme ultraviolet photolithography equipment asml. How was I going to achieve it then? China advances in semiconductors. SMIC (semiconductor manufacturing international corp) is the largest Chinese semiconductor manufacturer, and has a quota of about 5% In the world market. The US veto has caused this manufacturer to have access to EUV equipment, but Some rumors They pointed to what He had achieved A 5 NM node with deep ultraviolet technology (UVP). It is not so, and new data reveal that the X90 does not use a node of 5 nm (N+3) but the already known 7 Nm (N+2) As it happened with the Kirin 9020. Be careful, because in SMIC they are working on their own lithography of 5 nm, and they seem be close to achieving it. And beyond, of course, The 3 nm. A wolf with lamb skin. Huawei does not usually generate headlines, but it seems just what he is looking for. And meanwhile, as Huo states, he manages to buy time in which the industry does not look so much in it. And meanwhile, scale without rest. But as this analyst says, not to pursue quarterly benefits, but to endure. Image | Rubaitul Azad In Xataka | USA says that Huawei will not be able to manufacture more than 200,000 AI chips in 2025. We know what the reason is

Apple has confessed that you are using AI to design its next chips. What comes to us to users is anthological

by

American companies Cadence Design Systems and Synopys are the most important in the software industry specialized in Integrated circuit design automation, known as Eda for its English denomination (Electronic Design Automation). And both have taken the same path by integrating into their solutions models of artificial intelligence (AI) designed to optimize semiconductor design processes. Nvidia, AMD, Intel, Broadcom or Medialatek are some of the chip designers who use the EDA software of these two companies, and some of them are already starting to flirt with the design through AI. In fact, Johny Srouji, Senior Vice President of Apple hardware technologies, has confirmed That the apple company is interested in taking advantage of the generative artificial intelligence to design the integrated circuits that it incorporates into its devices. “EDA software companies have a critical role when supporting the complexity of our chips designs,” Soruji holds. “The generative the techniques have enormous potential that can help us carry out more design work in less time, so they can give us a great impulse in the field of productivity (…) Passing the Mac to Apple Silicon was a great bet for us. We did not have a plan B or we were going to divide the product line. We bet everything.” This is what we can expect from the chips designed with AI From Johny Srouji’s statements a very clear idea follows: his commitment to EDA software endowed with AI It is definitive. This is the reason why he recalled the moment in which Apple decided to leave Intel’s chips and bet on his own designs. Somehow this engineer is hinting that the use of the generative AI in the design of integrated circuits is a milestone for chips designers. And he is right. The main consequence derived from the use of EDA software is the acceleration of the integrated circuit design process The main consequence derived from the use of EDA software is the acceleration of the integrated circuit design process. Apple, Nvidia, Google, AMD or Intel, among other companies, already have on the table the possibility of invest much less time in design of their own semiconductors. However, this is not all. The AI ​​can also be responsible for carrying out a part of the work carried out by the engineers of these companies. The EDA software reinforced with AI is able to optimize the performance per watt of the chips and their gross productivity, two objectives that until now were in the hands of the experts in the design of the microarchitecture of the integrated circuits. For this reason Apple and the other semiconductor designers will be able to reduce the time between two generations of consecutive chips, and, more importantly, they will manage to adapt more quickly to the changing needs of the market. Somehow we have already revealed what implications the popularization of EDA software has from the user’s point of view. And it is that in all likelihood we can expect that the improvements implemented in two consecutive chips generations have a perceptibly greater than the current range. In practice this resource should allow integrated circuit designers to improve a lot both its gross performance and its performance per wattalthough it remains to be seen to what extent the limitations imposed by current silicon technology restrict the capacities of the next generations of semiconductors. Anyway, there is no doubt that a new and unexplored world opens before us. Image | Apple More information | Reuters In Xataka | Apple believes that its rivals are not doing well in ia. It is the perfect excuse to delay Siri one more year

It’s no longer about doing better chips, but connecting half a million

by

The new battle of the AI ​​is no longer earned only with integrated circuits, but orchestrating complete systems that connect half a million chips working at the same time. Why is it important. For decades, the technology industry has been governed by the Moore law: The number of transistors in a chip doubled every two years. Today, calculation power advances come from the ability to connect endless amounts of chips between them, an effort that does not follow any predictable “law”. The turn. AMD has presented Your mi400 chip Not as an isolated piece, but as a complete system called helios Rack of the pre -installed data center. It is the implicit recognition that the era of individual chips is over. The context. Nvidia dominates the AI ​​because Jensen Huang had the vision of using GPUS for these calculations, which gave him the advantage of those who hit first. But AMD is shortearing fast distances (Bet on smooth his usually goes wrong), and the inherent change of the industry has also changed competition. Between the lines. Companies no longer hire chips engineers, but also more concrete profiles: Specialists in Design of racks. Systems architecture specialists. Specialists in data centers optimization. AMD is now using More than 1,000 specialized engineers in design of racks. Architecture chiefs who previously focused on concrete domains should now be somewhat more general to create more integrated systems. The threat. The biggest competitive pit (‘Moat‘) Nvidia is no longer hardware, but software. His suite CUDA It is the standard in the AI ​​industry to train models. AMD competes with ROCM. And although changing system can be too expensive, especially for small companies or independent developers, large companies can justify spending if the greatest efficiency gives them the right return. Yes, but. Companies are packing so much hardware in the racks that High soils and electrical networks They begin to be too tension. The question here is when they can sustain this race without finding physical limits. Deepen. The CEO of COPEAidan Gomez, said In the event that was “very easy” to migrate one of your company’s models to the AMD platform. Openai, Oracle and other great technological ones are looking for alternatives to Nvidia since he boom of chatgpt began to make your chips scarce. Supplier diversification is also strategic for projects like Stargatethe initiative of half a billion Openai dollars next to Oracle. Outstanding image | İsmail Enes Ayhan in Unspash In Xataka | Taiwan has decided to prevent by all means that TSMC technology arrives in China. Your future is at stake

The US is willing to do anything for advanced chips not to reach China. And Malaysia is an obstacle

by

Malaysia is in the spotlight of the US. The administration led by Donald Trump has suspected for many months that Chinese companies and research centers dedicated to artificial intelligence (AI) acquire the most advanced NVIDIA GPUs through Singapore and Malaysian intermediary companies. In fact, this possibility It is no longer just a hypothesis. And it is that the Singapore government has confirmed that it has identified those responsible for diverting to China, and presumably towards the parent company of Deepseekservers that contain the high performance GPUs produced by NVIDIA. This information was revealed at the end of last February by the Channel News Asia television channel, and shortly after the Minister of Internal Affairs and Justice of Singapore, K. Shanmugam, He confirmed it. Interestingly, it did not realize what the GPUs that incorporate these machines are, but made public a very important fact: the name of the companies that manufactured the servers. And they are Two US clients from Nvidia Very important: Dell Technologies and Super Micro Computer. For Malaysia, the eeuu is not an option Malaysia is a global power in packaging and verification of integrated avant -garde circuits. In August 2023 we had the opportunity to check it first hand because Intel invited us to Visit the chips factories that has in this country of Southeast Asia. These plants reside in Penang and Kulim, two exotic and beautiful cities located north of Kuala Lumpur, and are specialized in the processing of silicon wafers that come from Intel factories in Israel and Ireland. Malaysia is a very attractive country not only for Intel and other manufacturers of American semiconductors; It is also for Chinese and Taiwan chips companies. Its appeal lies in four fundamental pillars: has a stable supply chainnotable technological development, makes available to these very qualified personal companies and their production costs are moderate. For integrated circuit manufacturers, a country with these characteristics is a safe bet. Many Chinese manufacturers are diverting the packaging of their chips to Malaysia to protect themselves from the tension between the US and China Intel has been present in Malaysia since 1972, and during the more than five decades that have since elapsed its infrastructure in this Asian country has not stopped growing. His model has worked so well that other large companies in the semiconductor industry have adopted it. One of them is the Chinese giant Huatian Technology, and many other Chinese manufacturers of integrated circuits are diverting the packaging of their chips to Malaysia to protect themselves from the uncertainty that has triggered the tension between the US and China. The chips industry is one of the engines of the Malaysian economy, and, as we have just seen, the US is an essential actor for this country. For this reason, Malaysia’s government is interested in doing everything in his hand to Do not get into the US administration. The presumable participation of some of its companies in the export of avant -garde chips to China places it in a delicate position that, for more inri, has just complicated even more. And it is that a company from China is suspicious of training AI models in Malaysia infrastructure and using Nvidia hardware. According to Reuters The Ministry of Investments, Commerce and Industry of Malaysia is investigating whether this or any other company housed in its territory has violated some local law. Interestingly, this institution is analyzing a report that argues that Four Chinese citizens They traveled from Beijing to Kuala Lumpur, each transporting several hard drives with dozens of calculus leaves and images, with the purpose of training an AI model into 300 servers with Nvidia chips. Malaysia is not subject to sanctions or restrictions by the US government, and it is evident that this Asian country is interested in its relationship with the Trump administration be as cordial as possible. What is not clear is that he gets unscathed from his involuntary mediation between China and the US. More information | Reuters In Xataka | We can forget an AI without hallucinations for now. The general director of Nvidia explains why

Intel will fit one of the toughest blows in its history. One directed directly to its chips factories

by

“These decisions are difficult, but they are essential to face the challenges we are already facing the company’s current financial situation. It is very painful.” These words They are part of the statement That Naga Chandrasekaran, the manufacturing vice president of Intel, sent the employees of this company last Saturday. In his concrete message that the Board of Directors is preparing A template reduction that will range between 15 and 20%as well as most of the cuts will arrive in July. The challenges a Those who currently face Intel They exceed the other challenges that he has faced during his more than half a century of history. The leadership that has sustained for decades in the manufacturing industry of integrated circuits is in the hands of The Taiwanese company TSMC since the mid -2000s. In addition, the stagnation during the last years of the PC market and the slowness with which Intel has participated in the industry of the industry of the artificial intelligence (AI) have placed it in a very compromised position. In July 2024, the company that at that time led Pat Gelsinger gave a tremendous batacazo in the stock market. Their actions fell 30% in a few days and stabilized in the value they had in 2011. In addition, Intel lost $ 1.6 billion During the second quarter of 2024 and its year -on -year income fell by 1%. These circumstances triggered a crisis that still persists today. This cut will have an inevitable impact on Intel factories Last August 2 Intel announced that was about to start a structural adjustment plan that pursued to reduce the costs of the company and increase its agility when adapting to the challenges that the market currently imposes. Its objective was to fire 15% of the workforce (more than 15,000 employees) and reduce costs by approximately 10,000 million dollars. Pat Gelsinger had declared shortly before Intel was inefficient because he had an excessively complex structure. It is still early to identify how this measure will affect the competitiveness and performance of Intel factories Not even Gelsinger himself “survived” this regeneration process. On December 2 This executive came out in a somewhat precipitated way of the company, whose course was uncertain until last March 18 Lip-bu took the reins of Intel. This veteran physicist and nuclear engineer He hastened to confirm that two of the pillars of its strategy would pursue reinforce the company’s position in the AI ​​market and reposition Intel as a leading company in the integrated circuit manufacturing industry. Shortly after his arrival It was leaked that lip-bu so He planned to launch a new cut of the Intel template in a clear attempt to reduce their operating expenses, among which personnel costs or marketing expenses are counted. The figure that the company shuffled on this occasion amounted to 20% of its workforce, which in practice implied to dispense with approximately 20,000 workers. These people join the more than 15,000 employees of which Intel has dispensed with during the last months of 2024. Now we know something else. From the statement that Naga Chandrasekaran has sent just four days ago to Intel employees it follows that during the next weeks they will be forced to leave their jobs Between 8,000 and 10,900 factor workers That has this company disseminated throughout the planet. However, presumably the most affected plant will be the largest of all: that of Oregon (USA). It is still early to identify how this measure will affect the competitiveness and performance of Intel factories, but its size invites us to anticipate that it will have a deep impact on the production infrastructure of this company. Image | Intel More information | Oregon Live In Xataka | Intel has confirmed that the 20A node will be skipped to reduce expenses. The 18A node will enter production in 2025

bet on your own chips

by

China wants to lead the electric car industry, and plan to achieve it by dominating the chip supply chain that give them life. The country, in search of its technological self -sufficiency in an industry that represents about 10% of its national GDPhas urged its manufacturers to win in the semiconductor career for electric cars. A large -scale plan. Sources of Nikkei Asia They claim that the main groups of Chinese car, including Saic Motor, Byd, Li Auto and Geely are preparing to launch models equipped with 100% national manufacturing chips. From 2026, at least two Chinese companies are expected to produce semiconductors for automotive. It is something that has already begun to be seen, with movements such as Xpeng or child betting on your own manufacturing chips for its most recent electric vehicles. On June 13 we saw the XPEng G7, the first car of the Chinese company to replace Nvidia chips for own chips that promised a three -time performance. 2027 as a key year. With the start of mass production of vehicle semiconductors, China would be setting the most ambitious objective in 2027: 100% national manufacturing in each of the chips equipped in its vehicles. It is a project under the supervision of the Ministry of Industry and Information Technology of China (MIIT)who would be summoning manufacturers regularly to know their national chip adoption rates. What is being done. Recently, Xpeng announced its Turing AI chip, with a computing capacity that, supposedly, triples to that of the Nvidia Orin. It is one of the most ambitious responses to Nvidia, and expects divisions such as Volkswagen in China to end up adopting this type of solutions. Why Volkswagen? Because Both have agreements for Manufacture electric cars. Other Chinese manufacturers, such as child, They announced their Shenji Nx9031your chip to boost autonomous driving functions. The problem? It was manufactured in the lithographic process of five nanometers, and that can only mean one thing. External agreements. China is in full Sprint stage with its semiconductors. Is about to have the ability to make 5NM chips and SMIC expects to reach 3NM lithographic processes in 2026 But, at least for the moment, the chips of their rivals are more advanced. This is why companies like Nio have opted for companies external to China (They do not detail who manufactures the Shenjibut its 5 nm process details that it is not national), and Xiaomi has allied with TSMC To manufacture your Xring 01the company’s first smartphones processor that competes directly with the A18 PRO of Apple and The best of Qualcomm. The company He has already announced that he will take his chips to carsand the big doubt is whether he will do it with TSMC as an ally. China has the ability to master this industry. There are three pillars that largely determine the success of an electric car that intends to sell anywhere in the world. Batteries (autonomy and load) Price Software (operating system, cabin functionalities, etc.) In the 2023 China Electric Vehicle Forum, the Institute for Market Economics Studies revealed that the external dependence of automobile chips was close to 95%. The key point is that cars chips are very varied: microcontrollers, memories, sensors … Most of them can occur with the current lithography teams that China has. In fact, there is a key here. Among the large chips suppliers for cars are already companies such as Huawei, Semiconductor and STM Microelectronics ByD. With the focus on the domain of the semiconductors, manufacturers outside China betting on Chinese software, and the country winning the battery careerChina has everything to end up dominating this industry sooner or later. A conquest outside and inside. The war that China is playing is not just internal. The country has the potential to dominate in markets such as the European. Spain is the best example, with brands like Mg heading the sales rankingsand giants like Chery disembarking in our country with Jaecoo and omoda. China no longer has the West as a mirror. His new heads in automotive fight with their own identity, a clear domain in areas where the European car has been left behind, and A deeper consumer understanding Regarding that of the brands they manufacture in Europe. Image | Geely In Xataka | Chinese cars no longer compete against the West. Compete against the future itself

USA says that Huawei will not be able to manufacture more than 200,000 AI chips in 2025. We know what the reason is

by

Jeffrey Kessler, Undersecretary of Commerce for Industry and Security of the US Department of Commerce, has done a few hours ago This statement in Congress: “Our evaluation indicates that the production capacity of Chips Ascend de Huawei by 2025 It will be 200,000 units or lessand we anticipate that the majority or all of that production will be delivered to companies within China. “ Besides, Kessler pointed out that “China is investing A lot of money to increase your chips production of artificial intelligence (AI), as well as the capabilities of the GPUs it produces. For this reason it is essential not to have a false sense of security and understand that China is reaching the US rapidly. “These statements express unambiguously how worried the US administration is in the development of China undergoing China in the field of AI. China goes only between three and six months behind the US in AI This is What David Sacks arguesan IA and cryptocurrency expert who exercises as an advisor to the Government of Donald Trump in this area. In addition, it acts as a link between the government and the American technology industry. The White House has clarified the statements of Sacks clarifying that this guru referred to AI models, so that Chinese chips are between one and two years behind their US equivalents. Earlier this week Ren Zhengfei, the founder and general director of Huawei, declared that the GPU Ascend of this Chinese company are still A generation behind of the chips for the US. His reflection is aligned with what the US government defends, but it is important that we do not overlook that Huawei invests more than 25,000 million dollars annually in the development of its hardware for AI, so presumably it will not take long to match the benefits of the GPUs produced by NVIDIA or AMD. Blea performance of integration technologies used by SMIC to produce Huawei GPUs has a very wide margin of improvement In any case, the prediction of Jeffrey Kessler about the production capacity of avant -garde chips for Huawei is sustained on an irrefutable fact: the performance by wafer of the integration technologies used by the Chinese semiconductor manufacturer SMIC (Semiconductor manufacturing international corp) To produce the Huawei GPUs it has a very wide margin. SMIC already has the ability to manufacture 6 Nm integrated circuits, and soon it can also produce 5 Nm semiconductors, but is limited by the performance of the deep ultraviolet lithography equipment (UVP) you have in your possession. It is meritorious that SMIC and Huawei engineers have managed to refine their integrated circuit manufacturing processes what is necessary to be able to produce 5, 6 and 7 nm chips With ASML UVP teams, but a priori it is very unlikely that with these machines they will be able to go beyond the 3 Nm. And it is because the technique of Multiple patterningwhich is what they are using, imposes important limitations. A note: This strategy in broad strokes consists in transferring the pattern to the wafer in several passes with the purpose of increasing the resolution of the lithographic process. His problem is that he usually has an upward impact on the cost of chips and the decline in production capacity. For Huawei it is a big problem not to have the necessary technology to produce avant -garde semiconductors comparable to those who manufacture Intel, TSMC or Samsung, so it is working on the development of its own team of extreme ultraviolet photolithography (UVE). More information | Reuters In Xataka | Nvidia has to deal with the absolute distrust of several US legislators. His plan in China is in danger In Xataka | The US wants to end the chips for the Chinese that are sold abroad. And China knows how to defend oneself

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.