NVIDIA

Nvidia has beat its income record again. Paradoxically, Wall Street does not seem to import much

by

The Fever continues to spur Nvidia in the stock market. The company led by Huang has presented again record results for your fourth fiscal quarter. And it has been making the rupture of records into a while. Why is it important. This comes shortly after concerns about Deepseek’s impact and its efficiency in the market that dominates Nvidia. About the idea that less hardware and less power was necessary to continue advancing in AI. Which caused the huge hole in Nvidia bag. The most prominent point of the results was the Awesome Blackwell startthe new chips architecture for Nvidia. Colette Kress, financial director of the company, explained that Blackwell generated 11,000 million dollars in income, exceeding all expectations and becoming “the fastest product launch in the history of our company.” In figures. The data centers business, where technology is concentrated for AI, remains the main growth engine with 35.6 billion revenue in the quarter, 16% more than in the previous quarter and 93% more than in the same period of the previous year For the full fiscal year, this segment reached 115.2 billion, more than double the previous year. The total revenue of the quarter reached 39.3 billion, 12% more than in the previous quarter and 78% higher intelligible. Between the lines. Nvidia is changing her strategy to deal with competition. Jensen Huang emphasized The call with investors that are designing specific platforms for “reasoning AI”, a type of AI that requires up to 100 times more computing capacity than the “traditional” generative AI, which reinforces their most powerful need for hardware even against more efficient models. Reasoning models such as O3 of OpenAi, Deepseek R1 and Grok 3 They require much more intensive processing. Blackwell offers 25 times more yield for these models compared to the previous generation. Infrastructure demand for AI is promoting initial orders of Clusters of 100,000 Gpus or more. And now what? Nvidia has already announced that he will present Blackwell’s successor, called Blackwell Ultra, in the second half of this year, maintaining his annual rhythm. Jensen Huang has confirmed that this transition will be simpler than Hopper’s Blackwellsince it will use the same system architecture. The contrast. Despite the impressive results, Nvidia’s shares barely reacted in the market after the presentationa behavior very different from that seen in previous quarters. The reason: investors perhaps see with caution the reduction in gross margins (from 75% to 73%) and the lowest percentage growth compared to previous quarters, although it remains extraordinary for a company that has established itself as the second most valuable in the world. The big question. Will Nvidia be able to maintain your technological and financial domain against increasingly intense competition and that also threat from China? It seems inevitable that, although Nvidia’s position will remain strong, a certain decrease in demand in favor of new competitors comes. And above all, as the huge investments made by companies in AI do not have a clear and lucrative return, something that seems quite feasible. Maybe it is time to readjust expectations. Outstanding image | Nvidia In Xataka | I have used Claude 3.7 for hours. It is the closest to a human brain that I have felt with an AI

Some GeForce RTX 50 graphics cards are missing rop units. Nvidia has explained why

by

The family of graphics cards GeForce RTX 50 developed by NVIDIA continues to occupy the Center for Care in the field of PC hardware some more than a month and a half After your presentation. On this occasion we do not propose to talk about how difficult it seems to be to achieve one of these graphic solutions; We invite you to investigate A manufacturing defect which affects a small amount of GPUs installed in these graphics cards. Shortly after their launch some users observed that the number of rankization units, identified in English as ROP (Raster Operations Pipeline either RASTER OUTPUT PROCESSOR), integrated in some GPU GeForce RTX 50 was lower than that indicated by NVIDIA in the specifications. These units are responsible for executing the operations required by the rendering by rankization with the purpose of delivering the images formed by the final pixels to the monitor. Nvidia explained what is happening Several buyers They assure that their GeForce RTX 5090 and 5070 TI have a number of functional ROP units lower than the one indicated in the specifications. Days later a user of Reddit He denounced this same problem on his graphics card with GPU GEFORCE RTX 5080. In the latter case the graphic chip should integrate 112 ROP, but that particular card incorporates only 104 ROP, which has a downward impact on its performance. “Affected consumers can contact the graphics card manufacturer to obtain a new one. This production anomaly has already been corrected” Unfortunately it is not possible to solve this problem by resorting to an update of the BIOS or the controllers because it is a hardware defect. Nvidia has confirmed Tom’s hardware that, in effect, the problem exists, and that It is a manufacturing defect which affects a small amount of GEFORCE RTX 50 family chips. According to the company led by Jensen Huang, the amount of affected graphics cards is somewhat less than 0.5%. “We have identified a rare problem that affects less than 0.5% of the GeForce RTX 5090, 5090D and 5070 TI. They have a lower ROP than specified (…) The average impact on graphic performance is 4% and does not affect the workloads of artificial intelligence (AI) and calculation. Affected consumers can contact the graphics card manufacturer to obtain a new one. This anomaly in production has already been corrected “, explains the official Nvidia statement. The ideal in these circumstances is that all players who have bought one of the GeForce RTX 50 that can presumably be affected by this anomaly in the manufacturing process check if Your GPU has appropriate rop units using Techpowerup Gpu-Z or other similar diagnostic application. If your GPU is one of the defective ones they have the right to claim the change of their graphics card in the store in which they have bought it. An important note: some Founders Edition cards sold directly by Nvidia are also affected. Image | Nvidia More information | Tom’s hardware In Xataka | Nvidia “has broken the deck”: it is leading the creation of a memory standard for the PCs with AI

Nvidia has a very strong and not quite unexpected ally against Depseek: TSMC

by

Deepseek has convulsed the industry of the artificial intelligence (AI). The irruption of this free and open source Chinese model has questioned the need to use in training and inference processes Chips for very powerful and expensive. Like those designed by Nvidia, which leads with an unappealable roundness the hardware market for Ia. Even so, just ten days ago Deepseek caused Its market value fell abruptly. Since then the hardware used by this Chinese company has generated a lot of distrust. The Deepseek responsible argue that the infrastructure they have used to train their agglutin model 2,048 chips H800 of Nvidia. And that this process with 671,000 million parameters has cost 5.6 million dollars. However, some analysts defend that these figures do not reflect reality. The very juicy report prepared by SEMIANALYSIS He maintains that, in reality, the infrastructure used by Deepseek to train his AI model approximately 50,000 NVIDIA GPU with Hopper MicroAritecture. According to Dylan Patel, AJ Kourabi, Doug O’Laughlin and Reyk Knuhttsen, at least 10,000 of these chips are GPU H100 of Nvidia, and at least another 10,000 are GPU H800. The remaining chips, according to these analysts, are the cuts cut H20. The TSMC cowos packaging is a very strong support for Nvidia As we have just seen, at the current situation it is reasonable to have doubts about the hardware that Deepseek has used in the training of its model (in inference seems to be using the GPUs Ascend 910c of Huawei). And also about the real cost of this process, which could be much higher than the officially announced by this Chinese company. In any case, Depseek has poured uncertainty about the AI ​​market, and this is the reason why so many US technology companies have lost value. Whatever Nvidia has just received a very strong support from TSMC, which is the semiconductor manufacturer that produces its GPU. And is that, according to Digitimes Asiathis Taiwanese company has decided launch an expansion plan five years long for integrated circuit manufacturing capacity using its advanced cowos packaging technology (Chip-on-Wafer-on-Substrate). According to Beth Kindigof the I/O Fund consultant, this technology will monopolize between 50 and 60% of the market in 2025 compared to 15% it supported during 2024. In 2024 TSMC he officially announced that he was building two cowos packaging plants in the town of Chiayi, housed in southern Taiwan The high demand for GPUs for AI with Blackwell MicroAritectura de Nvidia is largely responsible for the implementation of this plan. The company led by Jensen Huang can respond better to the needs of its customers and will see how its competitiveness is increased in a phase in which Depseek and other Chinese companies represent a challenge. In March 2024 TSMC officially announced which was building two cowos packaging plants in the town of Chiayi, housed in southern Taiwan. However, this is not all. He also shuffled the option to put a plant more specialized in this advanced packaging technology in Japan, presumably on the island of Kyushu, in which This company is building at the moment two semiconductor production plants of avant -garde. In any case, there is something else. And it is that Chiayi plants will be trained to work, in addition to the packaging cowos, with advanced Info and Soic technologies (System on Integrated Chips). It is evident that TSMC wants to cover your back well and look to the future to prevent its production capacity from being threatened by a bottleneck. An interesting note: currently the Cowos packaging is being used with the AMD Instinct Mi250 chips and with the A100, H100, H200, B100 and B200 NVIDIA GPUs, as well as in its derivatives. The review used in these last two chips, the B100 and B200, is known as Cowos-L. In 2025 TSMC will be able to process no less than 60,000 wafers per month using its advanced packaging technology. Image | TSMC More information | Digitimes Asia | Yahoo! Finance In Xataka | Samsung is preparing to give TSMC a bars where it hurts most: the manufacture of the chips for ia

take off in the stock market while Nvidia and Microsoft still do not lift their heads

by

Several days have passed since Nvidia lost more than 400,000 million dollars in market value After the emergence of the new Artificial Intelligence Models (AI) of Deepseekwhich stand out for their power and efficiency. Despite a slight recovery in the price of its shares, the company led Jensen Huang faces another day in red. Its titles fall more than 5% compared to the previous day. The bearish trend extends to other American giants Microsoft, Apple and Google. But not everything is bad news. ASMLone of the most important technological firms in Europe, is living a very encouraging moment. Their shares quoted at $ 683.35 yesterday Tuesday at the close of the market. This Wednesday, at the time of writing this article, They are negotiated at $ 713.98which translates into a growth of 4.48%. We are facing interesting fact if we take into account that the titles of the lithography chipos manufacturer also collapsed at the end of last week. Hopeful future for the European company There is still an hour for the closure of the Nasdaq stock market, but everything seems to indicate that ASML will close on Wednesday in positive. Is that the company based in the Netherlands has presented Its financial results for the fourth quarter which ended last December 31 and exceeded analysts’ estimates. Its turnover was 9,260 million euros (9,070 million euros were expected) and the net profit reached the 2,690 million euros (2,640 million euros were expected). ASML’s actions in front of American companies this Wednesday, January 29 | Image: Yahoo Finance The manufacturer of machines such as Twinscan Exe: 5000 or exe: 5200high -tech piece necessary to make avant -garde chips, has maintained its sales expectations for the current year. He is confident that he will get income from 30,000 and 35,000 million euros For 2025. It is an increase compared to 28,262 million euros invoiced in the last four quarters. The popular Chinese Deepseek does not seem to be a threat to ASML growth plans. Christophe Fouquet, CEO of Asml Current AI models have been created with computing power. Huge data centers running at full speed to train algorithms. Deepseek’s arrival has raised a scenario where the development of this industry It does not depend as much as the avant -garde hardware. Since Nvidia is the main supplier of specialized graphics for Data Centers, the company has paid the price with the collapse of its shares and, consequently, its market value. The CEO of ASML, Christophe Fouquet, was optimistic at the arrival of low -cost models such as Deepseek. In statements to CNBC He said he hopes that the emergence of this company will generate more demand for semiconductors, instead of reducing it. “A lower cost of AI could mean More applications. More applications mean more demand over time. We see it as an opportunity for a greater demand for chips, “said the executive. We will see if we are right. Images | ASML In Xataka | We already know the secret of the extreme efficiency of Deepseek: it has dodged the Cuda de Nvidia standard

Nvidia loses a record amount of $ 400,000 million per Depseek

by

Before the presentation by the emerging Chinese company “Deepseek” of a new low -cost artificial intelligence modelthe technological sector received a strong blow to the stock exchange generating record falls as in the case of Nvidia with more than $ 400,000 million dollars. The giant of the microprocessors who at the time led the market as one of the most sought -after companies in the world, this Monday, January 27, he was threatened with Depseek’s emergebut with a cost of $ 5.6 million, which caught the attention of investors since it is less than the billions of dollars than US companies invest in artificial intelligence. For Oliver Blackbourn, portfolio manager in the Janus Henderson multi -active team “the appearance of a potentially more efficient approach in AI processing questions the need for billions of dollars of investment planned in infrastructure and intellectual property.” While David Bahnsen, Investment Director of the Bahnsen Group commented that “which makes the massive sale of technology this Monday so discordant is that the valuations of many of these AI companies and technology do not offer margin of error. Excessive assessment always becomes a problem over timebut fundamental news becomes a major problem when combined with excessive assessment, ”Insider told Business. For their part, the markets in general were presented downward, for the operations of this Monday the S&P 500 fell 1.4%, the Nasdaq collapsed 2.3%and the Dow Jones until the moments did not record changes. Blackbourn says that “being more exposed than ever to stock markets, There is a danger that wider negative feedback loops are generated if a loss of trust occurs”, He mentioned to El País. Continue reading: –Jeff Bezos brothers invested $ 10,000 in Amazon in 1996: How much money is today?–I collapse in the Stock Exchange after the announcement of the Fed on interest cuts cuts–What are the 10 richest families in the world and where does their fortune come from (Tagstotranslate) Stock Exchange (T) NVIDIA

Nvidia has lost 400,000 million in market value. The lace has been given by China Depseek

by

It is the news of the day. And, perhaps, of the week. The model of artificial intelligence (AI) Open Source Deepseek R1 is causing an earthquake in American technology. And is doing it Due to its open nature. However, your business model is not the only thing that It represents a threat For AI and US semiconductor companies. The most surprising thing is that the infrastructure that Deepseek is relatively modest. To understand with some precision what we are talking about we are interested H100 of Nvidia. The company led by Jensen Huang He is already delivering The first units of his successor, the platform B200as expected, on paper is even more powerful. However, sanctions approved by the US government prevent Nvidia from selling to Your Chinese clients are GPU. Here largely resides the Deepseek rupturist capacity Deepseek’s efficiency and his open nature are convulsing Silicon Valley Chinese companies that are dedicated to developing and training AI models have not been another option to exacerbate ingenuity. We know that many of them continue to buy the most advanced GPUs in NVIDIA through intermediaries and in parallel markets, but possibly they are not doing so in the amounts they need. If we stick to Deepseek according to Financial Times The infrastructure used to train this agglutin model 2,048 chips H800 of Nvidia. And training with 671,000 million parameters has cost 5.6 million dollars. These figures are very restrained. In fact, if they really are reliable, and they seem to be, they would put an unappealable fact on the table: Depseek engineers would have managed to point An extremely competitive AI model with very lower costs than those needed by Openai and Google to develop a model of comparable AI. The H800 GPU is largely responsible for this circumstance. And it is because it was Nvidia’s response to the prohibitions of the administration led by Joe Biden. Nvidia engineers chose to cut the benefits of the H100 GPU with the purpose that the Commerce Department would allow them to sell it in China When the US government prohibited Jensen Huang’s company from giving its Chinese clients its most powerful GPU at that time, the H100 chip, Nvidia engineers chose to cut their benefits with the purpose that the Department of Commerce allowed them to sell it In China. The result was precisely the H800 GPU, which is nothing other than a simplified review, and, therefore, less powerful of the H100 chip. Everything was complicated again on November 16, 2023. And that day the US government approved New sanctions to China that, among other prohibitions, they prevented Nvidia The H800 GPU. Presumably at that time Depseek engineers already had in their hands the H800 chips they needed, although Some analysts defend that, in reality, its infrastructure brings together 50,000 GPU H100 bought through intermediaries. If so, it is evident that the tension held by the US and China would prevent Depseek from recognizing that thousands of illegal chips have in its possession. Whatever the truth is that NVIDIA QuotationMicrosoft, ASML and other large technology companies are falling in a very pronounced way. In fact, the company led by Jensen Huang has lost 400,000 million in market value Given the possibility that Deepseek demonstrates that to put a vanguard IA model, it is not necessary to resort to the most powerful GPUs of NVIDIA or other companies. If this has really been trained only with 2,048 chips H800 OpenAi, Google and other companies will crack. And this industry will give optimization and efficiency the importance they have. We will see what happens finally. Image | Nvidia More information | Financial Times In Xataka | China is closely monitoring the United States movement with Stargate. And your answer has already prepared

NVIDIA postpones Tesla’s great promise to the next decade

by

Autonomous driving. That eternal promise that is supposedly worth billions of euros, that we are always brushing with our fingers and that, in the end, never seems to arrive. The last to express their doubts have been those responsible for NVIDIA. “It’s super difficult”. The words are from Ali Kani, head of NVIDIA’s automotive division, who spoke with the magazine Coach about the future of autonomous driving. “It’s something typical of the next decade. We are not close. It is super difficult,” Kani assured journalists. The question referred to “truly autonomous cars” of which the head of NVIDIA. It is clear that “we will not see them in this decade.” Kani pointed out that cars currently do not have the power or the technology necessary to implement autonomous driving in the short term. A new way of working. Regarding how they work at NVIDIA, Kani assures that what they are doing “is very different” from what they did a year ago. He assures that they are “working on large language models such as GPT Chat with video.” And he emphasizes: “no one was doing this three years ago.” The problems. Basically, we do not have enough computing power to guarantee autonomous driving in the short term. “This type of model needs a lot more computing power, a lot more memory bandwidth. You need more LiDAR sensors and radars, redundant algorithms to make sure it’s safe. All of that has to work in parallel, which means more computing.” Furthermore, NVIDIA is clear that it is not just a matter of raw power. It’s also about giving a good image to potential customers and showing that you can travel safely. “The industry needs to go slow with this. If one company makes a mistake, the entire industry is set back a few years. So we have to act in the most responsible and don’t take any shortcuts. You can only do it when you have proven that it is really safe,” concludes Kani. A different approach. If we take what NVIDIA claims as a reference, the company works in a completely different way than Tesla tries to do. Technology makes it clear that radars and LiDAR sensors continue to be essential, components that Tesla wants to eliminate to trust everything to the use of cameras and recorded images. What Elon Musk’s company defends is that It has a huge fleet of vehicles already on the streets and that everything recorded by them allows their algorithms to learn faster than the competition. They hope, therefore, to spend less money and time to go beyond where Waymo or Cruise have gone. In 2027. The perspectives that come to us from NVIDIA are also very different from the promises that Elon Musk has made about his Tesla Cybercab. According to the owner of Tesla, his fully autonomous cars should be on the streets in 2027 despite his claims “Be unoptimistic with deadlines”. Then he said that next year he will be manufacturing his robotaxis. A promise that, as on so many other occasions, doesn’t seem too realistic. However, Musk is confident in his approach to Donald Trump to open the hand with the tests of autonomous vehicles and deploy their services more quickly. The eternal promise. The truth is, whether we’re talking about Tesla or any other company, the promise of the fully autonomous car always seems to be about to arrive. And it never seems to end. Waymo keeps going but Cruise has fallen by the wayside and has joined a long list of failed attempts. The truth is that Cruise has managed to launch the service in limited spaces but it is also true that the behavior of its cars it was easy to manipulate and who has lived in a eternal controversy of accidents and encounters with San Francisco emergency services. During all this time, Tesla has not managed to make its Autopilot a truly autonomous service and requires human attention. Ford’s BlueCruise also needs it, although in this case can be driven without hands on previously mapped roads. Mercedes does not require it but your system is limited to very specific circumstances. And a bag of millions. Despite everything, there are billions at stake with autonomous driving. Or that’s what they promise us. One of the most optimistic forecasts is that of Tasha Keeney, director of investment analysis at ARK, who quantified in one of his analyzes collected by The Wall Street Journal The value of a robotaxis service like the one Tesla is looking for could account for 60% of the company’s income in 2029 and raise its valuation to $800 billion. Photo | tesla In Xataka | Tesla promises a robotaxi without a steering wheel in 2026. General Motors already tried it with Origin and canceled its project this same year

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.