chips

The US believes to have the definitive solution to prevent its avant -garde chips for China

by

Despite the efforts you are making The US government to avoid itthe avant -garde GPU for artificial intelligence (AI) that produce nvidia, AMD or brains They continue to arrive in China. Controlling secondary markets and parallel import routes is extremely difficult, especially when they run through countries such as India, Malaysia or Singapore, in which The US action It is very limited. This is the reason why, According to Bloombergthe US Department of Commerce, which is responsible for regulating exports, is considering starting an unpublished strategy. This measure will consist, if it finally comes to fruition, to introduce the technology necessary to allow them to be tracked at all times. A priori would be a form of geolocation that would identify where each integrated circuit resides during its journey once it leaves the semiconductor factory. This strategy raises technical challenges Michael Kratsios, the director of the Office of Scientific and Technological Policy of the White House and one of those responsible for the US government’s action plan for the development of AI, has confirmed that are considering the possibility of introducing tracking solutions both implemented in the hardware and software itself: “We are discussing what software or physical changes could be introduced into the chips themselves to improve the tracking of their location.” The constant tracking by GPS is not practical in this scenario of use, according to Kratsios himself The US administration is determined to protect its current leadership in the development of hardware for AI as the adoption of this technology accelerates throughout the planet. However, implement the solution proposed by Kratsios and other White House advisors It implies important technical challenges. Constant tracking by GPS is not practical in this scenario of use, According to Kratsios himselfso a possible solution is to introduce in each chip the logic necessary to confirm its location periodically. In this way, external networks can be reduced. The problem is that this design increases the complexity of the logic of each chip and can presumably impact its performance, which would penalize its productivity in the calculation centers. In addition, the cost of each GPU will increase and hackers They could find new ways to violate the safety of these chips by resorting to possible weaknesses of tracking logic. All probability to Nvidia, AMD, Intel and other designers and chips manufacturers for this measure will not like them. They will hardly introduce a logic that will make them more vulnerable and make them more vulnerable. His problem is that so far in this area the US government has adopted a little conciliatory position. And it does not seem that your attitude will change. Image | Nvidia More information | Bloomberg In Xataka | The US will not be able to contain the technological development of China. Experts from the chips industry forecast it

Industrial espionage is threatening the largest chips manufacturer on the planet: TSMC

by

It is no secret: espionage is very present in The semiconductor industry. It is in other sectors and we can be sure that a strategic industry such as integrated circuits It is not at all immune to him. On this occasion “the victim” is the Taiwanese company TSMC, The largest integrated circuit manufacturer of the planet. And, curiously, this plot of industrial espionage does not arrive at any time. And this company is about to start the large -scale manufacture of chips using its most advanced integration technology: The 2 Nm. What we know at the moment is that, According to Reutersthe Taiwanese authorities have arrested three TSMC employees because they have allegedly stole commercial secrets of this company. As we can expect, behind this detention is TSMC itself, as He has revealed The Taiwan Superior Prosecutor’s Office in a statement. According to Nikkei Asiathose responsible for this company have realized that two employees and a former employee have been made with critical information about the photolithography of 2 Nm of TSMC. This information is very valuable. In fact, it could be used by a competitor to optimize its own semiconductor manufacturing processes. The research has not yet determined whether this stolen information has reached another company, but United Daily News ensures that researchers have registered the offices of the Japanese company Tokyo Electron. At the moment this last signature has not confirmed or denied anything. Other industrial espionage frames that have given much to talk Our next protagonist is Asml. This Dutch company is the only one that is currently capable of manufacturing teams of extreme ultraviolet photolithography (UVE) that are necessary to produce avant -garde chips. At the end of 2023 Peter Wennink, the then general director of ASML, confirmed that a Chinese origin employee had abandoned the company in 2022 after being recruited by Huawei with the purpose of revealing this Chinese company secrets of the Dutch firm that a priori could compromise their business. According to the Dutch medium, Russia has something important in its favor: it has spies inside Asml This event was so serious that those responsible for ASML decided to include it in their 2022 annual report and Wennink was forced to Give explanations In a meeting with investors. According to this executive, the information stolen by the former Chinese employee It was partial. In fact, he described it as “one piece in a puzzle whose box is not in your power.” This Wennink statement rightly reflects The titanic complexity that have the most advanced lithography equipment produced by ASML. At the end of last year this company starred in another plot of espionage. Again as “victim.” The Government led by Vladimir Putin intends to have a prototype of UVE lithography equipment capable of manufacturing 130 Nm chips ready in 2026. And in 2028 another similar one trained to produce integrated 7 nm circuits. A priori it is difficult to believe it, but, according to the Dutch medium USRussia has something important in your favor: it has spies within ASML. Like China. In fact, one of them, a former employee of Russian origin, He has been arrested and is accused of stealing critical commercial secrets. The last plot of industrial espionage that I propose that we review is starring the South Korean company SK Hynix and the Chinese company Huawei. In mid -April 2024 a former employee of Chinese sk Hynix origin It was stopped at a South Korean airport when he was about to enter the country. The accusation that was cernia about it was very serious: it was suspected of having stolen confidential information about semiconductor manufacturing processes Used by SK Hynix to give it to Huawei. This employee decided to print 3,000 pages of technical documents, and that movement gave it away because it aroused suspicion immediately within the company. He faces 18 years in prison. Image | TSMC More information | Reuters | Nikkei Asia In Xataka | South Korea fears US reprisals. To avoid their old lithography equipment, they take dust on a warehouse

Taiwan urgently needs talent for its chips industry. Surprisingly he is looking for it in summer camps

by

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. Anyway, for Taiwan, its semiconductor industry is strategic for three fundamental reasons: it represents between 13% and 15% of the gross domestic product from the country; It is the engine of its exports with a close value to 40% of the total; And finally, the production of avant -garde chips gives the country a huge relevance from a geostrategic point of view. For this reason for this Asian country it is crucial that TSMC, UMC, Foxconn, MediaTek and its other large technology companies have the workforce they need. One of the strategies that are using some Taiwanese companies or with important businesses in Taiwan to recruit young talent is summer camps and university courses. Its purpose is to capture young people who have the right skills, although, curiously, some of these camps are held abroad for a reason for weight: The birth rate in Taiwan It goes down while, as we have seen, Taiwanese companies that are dedicated to semiconductors need to constantly increase their templates. In Europe, 100,000 more engineers are needed The Taiwanese government supports this strategy. It is fully aware that its main semiconductor companies, especially TSMC and UMC, need a constant supply of well -trained engineers to support their growth. Its current demand cannot be filled only with Taiwanese students, which has led to the NTU (Taiwan National University) to implement A global degree program in semiconductors that precisely seeks to attract foreign students. In Germany a third of the technicians who have developed their work career in the chips industry will retire throughout the next decade In any case, the Taiwan integrated circuit industry is not at all the only one that needs to recruit new talent. 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. 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 You 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. Image | TSMC More information | Reuters In Xataka | We already know what the chips that will arrive until 2039 will be. The machine that will manufacture them is close

The strongest Chinese companies in Chips and IA have created a historical alliance

by

The future of Nvidia in China is increasingly uncertain. The suspicions of the administration of the Chibespace of China about The safety of your GPU H20 for artificial intelligence (AI) force this company to convince the Chinese government that their chips are reliable. However, this is not the only challenge that Nvidia faces in the country led by Xi Jinping. And it is that a good part of the most relevant Chinese companies in the semiconductor industries and AI have allied to encourage the adoption of chips for the Chinese. During the last fiscal year, which expired on January 26, 2025, China represented approximately 13% of total income of the company led by Jensen Huang with a figure of about 17,000 million dollars. In practice, this Asian country is the third best client of this company only behind the US and Taiwan. In fact, according to TrainingView At the beginning of 2025 Nvidia had a fee in the Chinese chip market for no less than 95%. However, during the last weeks it has dropped to 50%. This abrupt decrease is largely due to the export restrictions of chips for the The US government has imposedalthough it is also caused by the development of competition within China. And, to curl the curl, Stepfun, which belongs to Tencent Holdings; Infinigence ai; Siliconflow, from Huawei; Metax; Biren Technology; Focus me; Iluvatar Corex; Cambricon Technologies and Moore Threads, which are some of the strongest Chinese companies in the development of chips for AI, They have constituted an alliance who seeks to stop Nvidia. For Nvidia it is increasingly difficult to sustain its position in China The Chinese government is urging Chinese companies that are dedicated to the development of large models of the use in their servers integrated circuits of Chinese origin. The alliance for innovation in the model-chips ecosystem, which is what is called the organization that the Chinese companies that I have mentioned a few lines have constituted, represents an important step in this direction. Nvidia has in its favor the mass implementation of CUDA (Compute Unified Device Architecture) In the AI projects that are underway, but the panorama is already beginning to change. This technology brings together the compiler and development profits used by programmers to develop their software for NVIDIA GPUs, and replace it with another option in The projects that are already underway It is a problem. Huawei, who aspires to an important portion From this market in China, it has Cann (Compute Architecture for Neural Networks), which is its alternative to CUDA. However, this is not the only asset in the country governed by Xi Jinping. Nvidia has in its favor the massive implantation of CUDA, but the panorama is already beginning to change Moore Threads It is one of the Chinese companies that are dedicated to the production of hardware for which companies aligned with the interests of the US and its allies cannot sell software or advanced equipment. Although it is very young (it was founded in 2020) it has something very important in its favor: its founder is Zhang Jianzhong, former general manager of the Nvidia subsidiary in China, so it is evident that he knows well what he has in hand. Moore Threads has developed several GPU for AI applications that, on paper, rival some of the advanced solutions that have placed in the Nvidia, AMD or Huawei market. The cards MTT S4000 and MTT S3000 They are its most interesting proposals right now, although, curiously, in its porpholio the MTT S80 card also appears, a proposal for games and content creation that, according to Moore Threads itself, has a calculation capacity of 14.4 Tflops in single -precision floating coma operations. It doesn’t impress, but it’s not bad at all. However, this company has something else: a software package with which the domain of CUDA seeks to break. Muse calls itis compatible with the range of MTT cards that I have mentioned a few lines above and incorporates a compiler, execution libraries, specialized libraries and code purification tools. However, this is not all. On paper its most attractive capacity for China is that it allows to reuse the code written in CUDA, transferring it so that it can be executed on the cards for Moore Threads. It is difficult to predict what reception the hardware and software of this company will have in your country of origin, but there is no doubt that it is worth following the track. Image | Moore Threads More information | SCMP In Xataka | AI is the best thing that is happening to nuclear fusion. It is already accelerating the construction of Iter

The Samsung chips manufacturing subsidiary walks on the tightrope. Tesla will save her from collapse

by

Samsung has reached an agreement with Tesla. An important agreement. According to Bloombergthese two companies have signed a commitment that will run until 2033 and for which the subsidiary specialized in the manufacture of integrated circuits of this South Korean company will produce for Tesla chips for a value of 16.5 billion dollars in its Texas plant (USA). Not bad at all, especially if we are in mind that Samsung urgently needs that its semiconductor division Increase your presence in the market. Jay Y. Lee, the president of this company, sent in the middle of last March An internal statement in which it synthesizes with great precision what this company faces: “Our technological advantage has been compromised in all our businesses. It is difficult to see that efforts are being made to boost great innovations or assume new challenges. There are only attempts to maintain the status quo instead of generating disruptive changes. “ The Samsung subsidiary specialized in the manufacture of semiconductors is largely the company’s engine, and to recover health it is essential that its competitiveness increases. However, to carry it out it is necessary that its integration technology of 2 nm pelee from you to you both with the equivalent lithography of Intel, and, above all, with that of TSMC. However, Samsung’s starting point is favorable. And it is because He has been working in which it is undoubtedly The most important photolithography in its history. The 2 nm are crucial Samsung has led for more than three decades the industry of integrated dram circuits, but the rise of the artificial intelligence (AI) has triggered something that just two or three years ago would have seemed unthinkable: now it is also the South Korean SK Hynix the manufacturer of integrated memory circuits that LEADS THE HBM Chips Market so much (High Bandwidth Memory) that work side by side with the GPUs for the The one of the DRAM memories. Chips manufacturers need the candlestick performance of their avant -garde nodes to be at least 70% In current circumstances it is evident that Samsung needs to trace as soon as possible. And it seems that it is in it. At the beginning of 2025 several South Korean media anticipated that Large scale manufacturing of 2 nm chips It had already begun in the company’s South Korean plants. However, this does not mean that Samsung already has everything tied. Chips manufacturers need The performance by wafer of his avant -garde nodes is at least 70%and, according to the South Korean newspaper Munhwa Ilbothis company currently moves in the range of 40 to 50%. Even so, the Japanese chips designer for the preferred networks (PFN) and a South Korean company specialized in the design of neuronal processing units (NPU) They were already interested in early 2025 in which Samsung manufactured his designs in his new 2 nm node. There is no doubt that at the delicate moment this company is going through, having several agreements tied before its competitors initiate large -scale production with equivalent photolithographs is very important. However, this is not all. And is that just five months after that news the South Korean medium Chosunbizwhich in the matter of semiconductors does not usually give stitch without a thread, said Samsung I had already started the evidence In its 2 Nm node For Nvidia and Qualcomm. This test process does not guarantee that Samsung is finally going to manufacture integrated 2 Nm circuits for these two US companies, but their interest in the technology of this South Korean company is an oxygen ball. There is no doubt about that. Image | Samsung More information | Bloomberg In Xataka | This is the chips war: a former SK Hynix employee is suspected to deliver stolen technology to Huawei

The US hardened their restrictions for Nvidia chips not to reach China. So they are sweeping the black market

by

A few days ago Nvidia got what It seemed impossible: That the US government allowed you to sell your H20 GPU very soon to Chinese clients. It was an important turn in the Trump administration policy, which since April had raised his restrictions. Despite the hardness of the sanctions, the plan has had lagoons, according to an investigation of the Financial Times. What happened. The Financial Times has accessed sales contracts, to presentations of companies already involved in the industry and has drawn a conclusion: three months after Trump harden the export control, chips worth $ 1 billion were sent to China. In context, it is a huge figure, if one takes into account that Nvidia entered 17,000 million last year. The process began in May according to the Financial Times, when Chinese distributors began marketing GPU on which restrictions on data centers that work with Chinese Laboratories of AI weigh. The investigation points to something that the US already suspected: Many chips enter from Southeast Asia They are not any chips. The investigation reveals that Chinese AI companies are managing to acquire the GPU B200 of Nvidia, a beast that It already has a successorbut that is a candy to train models. And it promised multiply by four The performance of the desired H100 in MLPERF 4.1. The investigation also mentions the sale of other chips on which export controls weigh, such as the aforementioned GPU H100 and its successor, the H200. The price of a Rack Of eight B200 ready to use is approximately $ 489,000, and has dropped in price since they arrived in China in May. The difference with the sale that is authorized? An extra 50%. Publication in social networks announcing the sale of ASUS H200 racks. Image: Financial Times Nvidia’s position. The company has maintained a defense position to remain in China these months, and it is that this market supposed 13% of its global income. Jensen Huang has been praising the Chinese models of AI At a complicated moment for the company, in need of strengthening its complicity among Chinese companies while just selling chips. In full Dilemma for the US to sell or not sellHuang has placed the power of Huawei’s solutions at the height of the H200. A path full of difficulties. As Nvidia told Financial Times, buying chips clandestine is not something that allows expected yields. Mounting a data center with the GPUs is not just a process of installing parts, but about giving them service and support, something that the company does not supply to chips sold outside authorized channels. According to an operator of a data center, the export control does not prevent NVIDIA chips from reaching China, and what it does is create inefficiencies and “huge profits for intermediaries that assume risks.” Even so, a distributor recognized that “there is no shortage.” The effect of relaxation with H20. With the announcement of the permission to the marketing of the H20 to China, the sales of the B200 and other chips marketed in the black market have fallen, according to several distributors. The reasons are not clear, but buying Nvidia guarantees the aforementioned support as well as a more competitive price. Despite this, there are Chinese distributors announcing stock of the B300, which is not yet manufactured in mass. Image | Nvidia and Flickr In Xataka | The US machinery to win the “war” of AI to China is already underway. And it goes faster than expected

Openai has invested billions in chips. His only problem is that he has forgotten something important: cables

by

While Openai designs increasingly complex models and promises cities size data centers, a much more terrestrial reality threatens to stop the advance: the electrical infrastructure is not ready for what is coming. At the beginning of the year. Sam Altman (OpenAI CEO) promised, together with Oracle and SoftBank, an investment of 500,000 million dollars in AI infrastructure Under the name Stargate. The announcement, Made at the White House with President Trumphe talked about building 10 gigawatts of data centers on American soil by 2029. However, the start has been slower than expected. According to The Wall Street Journalnot a single formal contract with SoftBank has not been signed, and the first center, in Ohio, is still in the evaluation phase. Meanwhile, Openai has advanced on his own, expanding his alliance with Oracle to develop 4.5 more gigawatts, adding more than 5 GW under construction, As the company itself has reported. A big problem behind. This does not only affect Openai, rather the entire AI sector since no investment or quantity of GPUS can solve the true bottleneck of the sector alone: the electricity grid. As Le Monde explainedtraining models such as GPT-4 consumes dozens of gigawatts -hora, but the real challenge is in the “inference”, that is, in daily use. Each consultation to a model like Chatgpt implies complex calculations that consume energy every second. From the International Energy Agency (IEA), It was warned That the global electricity consumption of data centers could double before 2030, exceeding 945 TWH, more than all Japan today. And there the paradox appears: we can generate more energy, but we have no how to move it. In other words, high voltage lines, transformers, substations, adequate land, permits, technicians are missing. It’s like wanting to fill a city with bottled water, but without having pipes. The project continues. In a last statementOpenai has affirmed that its expansion with Oracle is already creating tens of thousands of jobs in Texas and that the objective of the 10 gigawatts is on its way to being surpassed thanks to new alliances. The Stargate I site, in Abilene, has already begun to operate with NVIDIA GB200 chips. However, As the Wall Street Journal has detailed, Disagreements with SoftBank persist on where to build, how to finance and how to connect data centers to electrical networks that are already saturated. Sam Altman recognized the challenge in an internal memorandum cited by axios: “The thirst for computer science is beginning to tension the supply chain and demands some real creativity.” Other paths: atomic energy. Faced with these limitations, technological giants are looking for solutions outside the traditional electrical system. The answer, for many, is in a surprising return: nuclear energy. Goal has signed a 20 -year contract with Constellation Energy to supply part of its data centers from a nuclear plant in Illinois. Google and Amazon They have also opted for small modular reactors (SMR). Microsoft, meanwhile, will reopen a closed nuclear power plant since 2019exclusively to support your AI infrastructure. Nvidia has not been left behind. In 2024, it invested 650 million dollars in Terrapower, the company founded by Bill Gates that is building the first Natriat reactor, a fourth generation machine that, According to their developersIt will generate electricity by half of the cost of a conventional reactor. The project, which takes place in Wyoming, has a Spanish participation: the public company sees the reactor cover. Without cables, there is no ia. Meanwhile, the number of users continues to grow. Chatgpt reached 800 million active users in April, According to Altman cited in axios. Each of them generates requests, questions, images, instructions, and all that consumes energy. According to Le Mondetasks as simple as writing an email can be more than 7% of a complete mobile phone charge. And generating an image consumes even more. Elon Musk says his company XAI already operates with 230,000 GPUS, and expects 550,000 more. The AI does not rest, but the cables that feed do not supply. Artificial intelligence promises to change the world, but before it will have to face something more basic than any algorithm: the physical laws of electrical systems. There is no needless. There are no servers without energy. And there is no energy without cables. Image | Pexels Xataka | The AI is opening the doors of a radical revolution on the Internet: that we can all create apps without knowing

The TSMC factory in Arizona is going well, although its chips are more expensive than those of Taiwan

by

Lisa her, the general director of AMD, He has just confirmed what we suspect since the beginning of this year: semiconductors that It is already manufacturing The new TSMC plant in Phoenix (Arizona) are Between 5 and 20% more expensive that the comparable integrated circuits produced in Taiwan. AMD will receive its first chips manufactured in this plant at the end of 2025, and given the increase in costs with all likelihood its market price It will be taller than that of semiconductors from Taiwan. Even so, Smooth his He maintains that the existence of this factory in the US is good news for all, for companies and users, because it contributes to the diversification of the supply of the chips and the strengthening of the distribution chain. According to histhanks to this plant and those that will arrive in the future will be more difficult for it to occur again A semiconductor crisis As serious as the one that triggered the Covid-19 pandemic between 2020 and 2023. Despite all the performance of the TSMC factory in Arizona it is high According to financial journalist Walter BloombergTSMC will increase the price of integrated circuits produced by 30% in the US to compensate for the costs triggered by tariffs on the production equipment of imported chips from Europe and Japan. Of course, the Government led by Donald Trump has not yet revealed how the import tariffs of photolithography machines used by TSMC, Intel or Samsung plants in the US will affect the importation tariffs. N4 and N4P lithographs are part of the TSMC Finfet Integration Technologies Family In mid -April 2024 CC Wei, the executive who at that time held the reins of TSMC, advance that the increase in costs derived from the manufacture of integrated avant -garde circuits in the plants that the company has outside of Taiwan It would be assumed by both TSMC and its customers: “If my client wants to manufacture in a specific area (out of Taiwan) then TSMC and the client himself will have to share the increase in costs (…) We are already discussing it with our customers.” Whatever it is beyond the costs linked to the production of chips in the US, The TSMC factory in Arizona is fine. This plant is manufacturing semiconductors in the N4 node (5 Nm). Lithographs N4 and N4P are part of the TSMC Finfet integration technologies family, although on paper the N4P process is a bit more refined. In any case, the plant we are talking about will not be the only one that this company will have in Arizona. The second factory will be operational in 2028 and will produce integrated circuits in N3 (3 Nm) and N2 (2 Nm) nodes. And finally, the third factory will not be listed at all until the end of this decade and It will produce chips in the node N2 (2 nm). TSMC can manufacture integrated 2 nm circuits In its US plants, although the Taiwanese administration will cauture the use of this technology in the country led by Donald Trump. “Private companies must make their own commercial decisions covered in their own technological progress (…) TSMC is building factories in the US with the purpose of serving their US clients because 60% of the world’s chip designer companies are based precisely in the US.” These words of Jw KuoMinister of Economic Affairs of Taiwan, are a declaration of intentions. It will be interesting to verify how the relationship of the US and Taiwan governments in the future prosper. Image | TSMC More information | SCMP In Xataka | The US confesses its worst nightmare: if China invades taiwan and controls TSMC the US economy will go to pique

Openai and Oracle prepare a brutal data center with 2 million chips for ia

by

The plan ‘Stargate’ It is the great bet of the government led by Donald Trump to keep the US at the head of development in artificial intelligence (AI). When this project was made public on January 22, it hit the economic endowment that would make it possible: nothing less than 500,000 million dollars. This money will leave the coffers of the Japanese investment group SoftBank; of those of OpenAI, the creators of Chatgpt; of those of Oracle, and, finally, it will also be provided by the investment firm Emiraratí MGX. These companies will support the construction during the next four years of an advanced network of data centers that will house the high performance computing infrastructure necessary to sustain US leadership in the AI field. The spearhead of these facilities is already being built in Texas (USA), in a town called Abilene. And it is colossal. In fact, this first data center of the ‘Stargate’ project will bring together, According to OpenAimore than two million chips for ia. 4.5 GW Dan for a lot When the US government announced to Bombo y S pay this plan left a great question open: how did he plan to solve the supply of electricity required by the new facilities? Large data centers for AI consume a lot of electricity, which has caused Some technology have opted for investing in nuclear centrals to guarantee the supply of electricity that these facilities require. Oracle began to install the first ‘racks’ with NVIDIA B200 platform servers in June At the moment this question is not completely resolved. And it is not because the ‘Stargate’ infrastructure should be completely ready before President Trump’s current mandate expires, and a new nuclear power plant can hardly come into operation in four years. Even so, Openai and Oracle have formalized A few hours ago they have reached an agreement to build the necessary infrastructure to Deliver additional 4.5 GW to your data centers. Interestingly, SoftBank does not participate in the financing of this expansion, although, as I mentioned a few lines above, it does in the ‘Stargate’ project. According to OpenAia part of the gigantic Abilene data center is already in operation. In fact, Oracle began installing the first racks With platform servers B200 of NVIDIA In June. It is surprising that in just six months a part of this installation is already active, but it is important that we do not overlook that the Abilene data center seeks to demonstrate that the ‘Stargate’ plan is viable within the deadline that those responsible have promised. Anyway, in this equation there is another unknown that also has a lot to say: China. “We hope that China significantly increase its investments in AI and semiconductors in response to the US domain in AI,” CBM consultancy analysts foresee. It makes sense. These two great powers world supremacy are being disputedso it is understandable that each significant step that give one of the two Receive a more or less overwhelming answer from the other. We can be sure that 2025 will be an even more agitated year than 2024 in the geopolitical and technological fields, so we will be attentive to the steps that US and China will surely give. Image | Christina Morillo More information | OpenAI In Xataka | Huawei attacks Nvidia positions in China: he wants to have dominant hardware in inference processes in AI

Nvidia cannot sell her most powerful chips to China for sanctions. So you have found a plan B: Risc-V

by

Nvidia has announced that Its CUDA platform will be compatible with RISC-V processors. He has done so during the Risc-V Summit in China and the chosen place is not accidental: this announcement clearly points to the Chinese market. For the first time, the technology that allows applications to communicate with the NVIDIA GPUs will be extended beyond ARM and X86, towards an open source architecture. Why is it important. CUDA It is the software that operates the Nvidia’s ecosystem. Without a CUDA, the GPU would lose much of their parallel calculation apacities. That Nvidia opens this technology to RISC-V It means that processors based on this open architecture can now serve as the main CPU in NVIDIA GPU systems. The background. The announcement, in addition to making in China, comes while China is accelerating its efforts to reduce its dependence of western processors. Nvidia can’t sell your most powerful models GB200 and GB300 to China for US sanctions, so in this way finds a way to maintain relevant Cuda in the Chinese market. Between the lines. There is a lot of geopolitical strategy in this decision: NVIDIA has been integrating RISC-V nuclei for years into its own GPU for low-level control tasks. Now it makes the jump to support RISC-V as the main processor. And that responds to a reality: if China is going to develop its own processors using open architectures, Nvidia wants to be there from the beginning. In detail. The configuration shown by NVIDIA shows a heterogeneous system: The GPU handles parallel loads. The RISC-V processor executes the CUDA controllers and the application logic. And a DPU manages network tasks. This architecture allows you to orchestrate GPU computations completely within the CUDA environment, something impossible so far with RISC-V. Deepen. Historically, Nvidia has behaved Cuda to each important architecture: X86, ARM, PowerPC and even Sparc de Sun. The company understands that it must be present from the first day on any platform that can take off in the business sector. With a value already exceeding 4 billion dollarsNvidia can afford to bet on all promising architectures. And now the movement positions RISC-V as a viable alternative for future designs of AI processors and high performance computing. If the stars align, other manufacturers could follow the example of Nvidia. And that would accelerate the adoption of RISC-V in data centers beyond China. Outstanding image | Wikimedia Commons In Xataka | On his way to the authentic quantum supremacy, China has set an objective: a “real” quantum computer before 2030

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.