Latest AI News
This 24-Year-Old From Bengaluru Is Betting Against Expensive AI Drug Discovery
Anindyadeep Sannigrahi founded LiteFold just under a year ago. Now, his startup has released LiteMol-1—a pioneering multi-molecule foundation model.
View
‘AI Can’t Write the TypeScript Compiler’
“Our compiler is also not a typical piece of code. That’s why AI is bad at writing it: it hasn’t seen anything like it in the training set,” reveals Anders Hejlsberg.
View
The U.S. is building barriers around drones and robots, but China has scale to get around them
In July and August, Washington tightened restrictions on foreign-madeadvanced robotic systemsand imposedsteep tariffson imported drones and their components, both moves citing national-security concerns. The drone tariffstake effectin September, with additional component tariffs following in 2027. These moves are part of a broader U.S. effort to restrict foreign technology in strategically important industries. The FCC’sCovered List, established in 2021, initially targeted telecommunications and surveillance equipment from companies including Huawei, ZTE and Hikvision before expanding to foreign-made drones and, most recently, to advanced robotic devices. The latest move comes as Chinese manufacturers have built commanding positions in both drones and humanoid robots, often competing at prices U.S. and European rivals struggle to match. Taken together, the restrictions are raising a bigger question for the global robotics industry: If Chinese drones and humanoids are increasingly shut out of the U.S., where does the competition move next? The restrictions may protect parts of the American market, but they don’t directly address China’s global manufacturing scale and cost advantages. Industry analysts and executives who spoke with TechCrunch said the result may be less a clean U.S.-China split than a more fragmented global market, with Chinese companies expanding elsewhere while U.S. and allied manufacturers compete in markets where security requirements matter more. The U.S. and Chinese robotics industries remain deeply connected, but the two countries enter the competition with very different advantages. Unlike semiconductors, robotics does not hinge on a single technology that one country can easily control, said Ankur Saxena, an investment director at TDK Ventures. China dominates global humanoid robot manufacturing, with global shipments hitting 22,000 units in the first half of this year — the vast majority from Chinese manufacturers — according to areport by Counterpoint. U.S. companies, by contrast, are operating at a far smaller scale, said Soumen Mandal, a principal analyst at Counterpoint Research. The world’s five largest humanoid robot makers by shipments — AgiBot, Unitree, Galbot, UBTECH and Leju Robotics — were all Chinese and together accounted for 86% of global shipments in the first half of 2026, according to Counterpoint. That advantage could compound. Lower prices allow Chinese manufacturers to put more robots into use, generating real-world data that can improve their technology. Higher production volumes, in turn, can drive costs down further, Saxena said. Mandal said Chinese humanoid makers are also pushing costs down by bringing more of the technology stack in-house and drawing on China’s existing manufacturing base. Unitree, for example, is developing more components internally, while automakers such as XPeng can draw on their experience in chips and vehicle manufacturing as they move into robotics. “The United States leads in frontier AI, software and semiconductor innovation,” Saxena told TechCrunch. “China leads in manufacturing scale, supply-chain depth and cost.” That manufacturing edge has let Chinese companies cut humanoid prices faster than most U.S. competitors can match. “You cannot sanction your way around a cost curve. You can only out-build it, and America has yet to begin making the decade-long investment that will require,” Saxena said. The answer may increasingly be outside the U.S. Even if Chinese robotics companies lose access to the American market, they still have a large domestic market and room to expand elsewhere, particularly in regions where demand for affordable automation is growing, Saxena said. Chinese robotics companies are already targeting price-sensitive markets with severe labor shortages across Europe, Southeast Asia, Latin America and the Middle East, said Mandal. Mandal expects humanoid makers to follow a path similar to Chinese electric-vehicle companies: build scale at home, expand into overseas markets, and eventually establish local production. Countries facing labor shortages and demographic decline could become early markets for humanoids, particularly in manufacturing, where robots can take on repetitive work. The drone market offers an early glimpse of what that more fragmented robotics landscape could look like. The industry is increasingly splitting into two ecosystems: a U.S.-led market built around American-made, NDAA-compliant systems, and a China-led market focused on low-cost, high-volume production, said Bentzion Levinson, founder and CEO of Virginia-based drone maker Heven AeroTech. Levinson said Western manufacturers are unlikely to beat Chinese companies in the low-end consumer drone market, where cost remains a major advantage. Instead, U.S. and allied companies could increasingly compete in long-range autonomous systems for defense and critical infrastructure, where security requirements carry more weight. Levinson sees the next competitive frontier shifting from the drones themselves to the technology that powers them and the equipment they carry. “The next battleground is over who owns the next-gen energy and payload architecture,” he said, pointing to battery constraints in particular. As drones become more capable, he added, battery limitations could make power systems an increasingly important point of competition. Agility Roboticswelcomedthe FCC’sdecisionin July, saying it could address security concerns around foreign-made advanced robots before they become deeply embedded in the U.S. market, as has happened in the drone industry. The company pointed to its Digit humanoid, which is designed and assembled in the U.S., while also calling for continued access to the tools and technologies needed to advance robotics research. “The alternative to China isn’t a purely domestic U.S. supply chain; it’s a diversified allied one,” Saxena said. That could create opportunities elsewhere in Asia. Japan has decades of experience in industrial robotics and precision manufacturing, South Korea brings strengths in electronics, batteries and automobiles, and Taiwan is a major player in semiconductors. But none can simply replace China, Saxena said, given how deeply Chinese components remain embedded across the global robotics industry. Asian manufacturers could emerge as a middle ground between lower-cost Chinese robots and more expensive U.S. offerings, Mandal said. South Korea’s Hyundai, which owns Boston Dynamics, and Japan’s Toyota are among the automakers investing in robotics, drawing on their expertise in vehicles, manufacturing and autonomous systems as they move into humanoid robots. Yang Fang of Beagle Technology, a California-based agtech startup that uses AI and robotics software to turn conventional farm equipment into autonomous machines, told TechCrunch that robotics is likely to become more regional as companies design machines for the labor needs, working conditions and customers in their home markets. Chinese robotics companies, for example, may focus on products suited to China and nearby markets, while U.S. companies are more likely to build for industries across North America, he said. The result may not be two neatly separated U.S.- and China-led robotics industries. Instead, the restrictions could accelerate the emergence of regional markets: Chinese companies competing on cost and scale across much of the world, U.S. and allied manufacturers gaining ground where security requirements matter most, and manufacturers in Japan, Taiwan and South Korea trying to carve out space between the two.
View
OpenAI Buys Tens of Thousands of Mac Minis, Studios for Reinforcement Learning: Report
Apple has been caught off guard by the level of enterprise demand for Macs for AI workloads.
View
Musk’s faster path to more gas turbines comes with pollution problem
Elon Musk says he’s found a way to solve one of AI’s biggest bottlenecks by making a hard-to-manufacture turbine part himself. On Saturday, Musk confirmed what a secret foundry SpaceX has been building in Bastrop, Texas, is for — an apparent response to a story that was already closing in on the details. Earlier in the day, The Informationpublished a reportciting job listings that explicitly mention a “blades and vanes foundry,” plus findings from Corey Trinetti, a due diligence specialist who authors detailed reviews ofAI infrastructure sitesin his newsletter and who’d reported that SpaceX had bought roughly 830 acres near its existing Starlink factory in Bastrop between March and June. “SpaceX and Tesla are each building 100GW/year of solar production capacity as fast as possible,” Muskwrote on Xon Saturday, “but natural gas will still be needed to supplement and bootstrap solar for several years. The limiting factor for nat gas turbine production is casting the blades & vanes. By doing in-house casting at SpaceX, we can accelerate nat gas turbines coming online by up to 18 months, which is a profound game-changer.” The “why” of all this goes back to one of the biggest challenges facing the AI industry right now. GPU shortages are still an issue — Nvidia’s newest Blackwell chips are still running lead times of several months, for example — but a second constraint has emerged alongside it, which is the physical power grid. The International Energy Agency projects global data center electricity use will roughly double by 2030, and gas turbine maker GE Vernova says it’s essentiallysold outof production capacity through 2030 due largely to AI infrastructure demand. That shortage is why building private gas-fired plants next to data centers, instead of waiting on the grid, has become a ubiquitous strategy for so-called hyperscalers, includingAmazon,Google,Meta,OpenAI, andMicrosoft. After years of prioritizing wind and solar, they’re all now betting on natural gas to get data centers online faster. As for the casting bottleneck specifically, according to The Information, the blades inside a gas turbine’s hottest section run at temperatures around 3,000 to 3,600 degrees Fahrenheit, which is roughly 800 degrees hotter than the melting point of the very metal alloy they’re made from. That’s only possible because of the blades’ internal cooling channels and thermal-barrier coatings, plus thespecific way each blade is cast. Just four companies worldwide have mastered the casting process well enough to produce them at industrial scale, and all of them are tapped out right now. What makes the whole thing especially difficult is that each blade has to be cast as a single, unbroken crystal, grown slowly inside a vacuum furnace, without the microscopic seams that let ordinary cast metal crack under stress. It’s a tricky process even for the smaller blades used in jet engines; the blades in power-plant turbines are considerably larger, which makes producing them at that scale and without defects even harder. If SpaceX pulls this off — and it’s easier said than done, of course — it would mean a Musk-controlled entity holds a manufacturing capability that every other AI infrastructure builder currently depends on a tiny oligopoly for, giving SpaceXAI an edge that’s difficult for any well-funded but non-manufacturing competitor to copy quickly. But it would also mean more gas turbines coming on fast, and turbines in the ground are already drawing federal lawsuits and peer-reviewed health research over the pollution they emit. In Memphis, where SpaceXAI has run gas turbines to power its Colossus data centers since 2024, the NAACP has repeatedly accused the company of operating turbines without the permits or pollution controls required by federal law. The organization’s concern is that turbines like these emit smog-forming compounds andhazardous chemicalslike formaldehyde, pollutants linked to asthma, respiratory disease, and certain cancers. (The site sits near neighborhoods that already face heavy industrial pollution, and University of Memphis researchers said that in their own admittedly limited analysis, air pollution grew “slightly worse” because of the data center.) But Memphis just happens to be the most visible case. The same fight is playing out anywhere gas turbines have become the default fix for data center power shortages. In Virginia’s “Data Center Alley,” astudy commissioned by the Piedmont Environmental Council, using the EPA’s own COBRA health-impact model, found that emissions from a single facility’s eight full-time gas turbines could reachmore than 2.5 million peopleacross multiple counties — with the heaviest impact landing on already-marginalized communities — and cause an estimated3.4 to 6.5 additional premature deathsa year, translating to $53 million to $99 million in annual health-related damages. The list, and complaints, go on.
View
Caterpillar is bringing to AI deployment what it learned from automating mining
Nearly every company that’s trying to deploy artificial intelligence runs into the same problem: it’s hard to integrate the tech into everyday operations. Industrial heavyweight Caterpillar has spent decades dealing with a version of that problem in the physical world, and now it’s using its experience to deploy AI. Caterpillar’s push into the autonomous space started with mining, where labor shortages and hazardous conditions can make automation particularly useful. Today, it sells automated haul trucks, drilling, underground loaders, dozers, remote-controlled construction equipment, and more. It also offers a software command center, fleet management, and even remote terrain intelligence as part of its autonomous toolkit. “Now we’re in this super exciting time where we can take all of that learning from mining and bring it into much more dynamic environments, jobsites, quarries, and construction sites,” the company’s CTO, Jaime Mineart, told TechCrunch on the sidelines of the Ai4 conference in Las Vegas earlier this month. The industrial giant is now applying AI more broadly, including in tools used by technicians and its own employees. One example is theCat AI Assistant, which lets field technicians standing next to a machine use voice commands to pull up repair procedures, troubleshoot potential problems, and identify parts that may be needed before beginning a repair. Mineart said the tool is now being used by customers, operators and technicians. The assistant draws on Caterpillar’s proprietary data, which spans information generated by its connected machines. Mineart said Caterpillar has about 1.6 million connected assets globally and more than 16 petabytes of structured data. The company is also using AI to power software for scanning sites and generating digital twins in manufacturing to analyze operations, she said. And like nearly every other company, Caterpillar is using AI across its enterprise operations, as well as for software development. “We use AI agents to modernize legacy code, generate and test new software, and identify defects earlier,” Mineart said. But Mineart is quick to point out that building the technology is only part of the challenge, as deploying an autonomous machine is not the same as transforming a site to use AI. Companies also have to rethink how people work alongside the technology and how existing processes need to change. “The hard part about autonomy and about physical AI is incorporating that technology into the customer jobsite and into the workflows,” she said. Mineart said the company leans on experienced operators to help train AI systems, leveraging institutional knowledge built over decades. And as machines become more autonomous, some operators may shift from controlling a single machine to overseeing multiple machines from a remote command center. That transition, however, is creating a new challenge for Caterpillar: training its 118,000 employees. Mineart said the company plans to spend$100 million over the next five yearsto train its workforce in AI, autonomy and robotics. That investment is likely being put towards helping the company make the most of the broader boom in AI infrastructure, which is already helping its top-line. Caterpillar’s quarterly revenue reached an all-time high of$20.5 billionin the second quarter, helped by strong demand for power-generation equipment used in data centers. Its power-generation division saw sales spike 72% to $3.10 billion, and CEO Joe Creed said that “no one is slowing down” when it comes to demand for cloud computing and generative AI infrastructure.
View
AI Is Rewriting Hiring. Humans Still Make the Final Call
Modern ATS platforms already incorporate AI, allowing recruiters to organise applications and rank candidates using customised parameters.
View
“We’re not doing 30 bets a year”: Vijay Pande on betting small after running $4 billion at a16z
It used to be that Vijay Pande was better known in academic circles than investor circles. That changed pretty abruptly a dozen years ago, when Marc Andreessen and Ben Horowitz — who’d spent their firm’s first five years explicitly avoiding healthcare and life sciences — decided the category was worth betting on after all and handed the keys to Pande. At the time, he was a Stanford chemistry professor who was best known for building Folding@home, the distributed-computing project that turned millions of home PCs into a supercomputer for disease research. Over the next decade-plus, he grew a16z’s bet into a practice managing close to $4 billion. So it was somewhat unexpected when in June of last year, Pande walked away from it all to start something much smaller. In fact, his new firm,VZVC, co-founded with longtime investor Zach Werner, is built around a handful of concentrated bets a year rather than dozens, it has no associates, and it relies heavily on AI for its day-to-day operations. To learn more about Pande’s hard pivot, we talked with him this week about why he’s making just a handful of concentrated bets rather than spreading himself thin in the current market — and about one of the more interesting conundrums in AI-driven biotech: unlike text, biological data can’t be scraped off the internet, so nearly every company ends up building its own walled-off dataset. What does that mean for all the advances AI in medicine has promised, and who actually gets access to them? This conversation has been edited for length and clarity. You can also listen to the fuller conversation (below). You’ve said biology is moving from a “science of discovery” to something you can engineer. What does that mean? For a lot of the way drugs have been developed, there was very much a fortuitous aspect to it. I think what’s shifted is that AI and machine learning allow computers to wrap their type of understanding around something very, very complicated… to try to figure out what targets you want your drugs to hit, for specific diseases, to be able to make those drugs, and now even to help in the clinical trials — which are the most expensive part of the process. I thought clinical trials were getting cheaper because drug developers are using more synthetic data, so not as many people are needed for these trials. That’s, I think, very much an aspiration. The cost and time to get to clinical trials has been shrinking, especially with AI, but it could still cost hundreds of millions of dollars to run a trial, which is why drugs are very expensive. The probability of a drug going successfully from the first trial to the end of the third trial is just 20%. If 8 out of 10 fail, and these things cost hundreds of millions of dollars, the amortized cost gets really high. The reason they fail typically is not that the biologist did something wrong; it’s that all the experiments these drugs were designed on were on animal models like mice, and in the end, animal models are just not very predictive of humans. The AI model is not going to be perfect, but it’s going to be way better than any animal model would be, and once it crosses that bar, that’s where it gets really exciting. [The phase after that is]: Is the drug the right drug forme? You mean personalized medicine. . . The jargon here is so-called precision medicine. If you go to a doctor with something not trivial, they have to guess what’s going on, because there’s only so much they can tell. Then they give you a drug — and if that doesn’t work, they give you another drug, then another drug. This happens in cancer, it happens in lots of different areas. We would all be much better off if the first drug was the right one. Typically, your blood test values are compared to population averages. But really, they should be compared to: is this [result] weird for you? What we’re starting to do also on the medicine side is [the ability] to just understand what would be right for the individual. Would you say the path to this moment has been slow and steady, or did it spike more recently? I think it’s lots of different things [coming together]. So for instance, precision medicine for the longest time was based on genomics. But the reality is your genome is kind of like the blueprint for your house on day one, but your house is fairly different now compared with the moment it was built. So there are many other things that people can now measure in proteomics and so on that are much more relevant for understanding disease and where your body is now. There has also been [a lot of] automation in robotic measurements that is naturally tied into AI, and those two go hand in hand really well. Over the last decade, there’s been this steady clip fforin both AI for biology and AI for chemistry. The biology part is like, how can we treat this disease? And then the chemistry part is, how can we come up with a drug to go after that specific protein? There have actually been very significant advances over those 10 years. You mentioned that biology is one of the few places AI can’t just scrape data off the internet. What does that mean for how the field develops? It’s a place where you don’t have any of this data that people can just all train the same thing, and your data can’t be distilled from one model to another. It’s a really interesting play from just the pure AI sense. Doesn’t that echo a familiar problem in medicine, though — doctors operating in [territorial, often competitive] silos? You’re onto something really big here. Let’s say [someone] has some type of cancer, and it’s both an issue in oncology and endocrinology — those two doctors really don’t sync together very well. What is really intriguing about AI is that it can, in principle, be a specialist in everything, and it can start to see things that really any single human being couldn’t. It would be equivalent to having a team of the very best doctors all clamoring together in that moment. But is there enough data sharing for that vision to actually be realized? I understand why founders and investors want to protect their [respective findings], but . . . I think one of the bigger trends is that we’re starting to see a shift toward building these atlases of biological information — which, from a technology standpoint, are typically foundation models. And as they become more common, I think we’ll see the same thing that’s happened with open-source LLMs, which do very well against the corporate ones: open-source foundation models in biology having a very broad impact. You’re involved with Genesis Therapeutics, which came out of your lab at Stanford, and Insitro, the drug-discovery company launched by Daphne Koller, a former colleague at Stanford. You say you’re also incubating a company with a founder you’ve known for 20 years. What are you looking for in founders, and in what areas? There are two areas that I’ve been spending most of my time on. One is AI for healthcare delivery, which I did a ton at a16z as well, and then AI for clinical trials. One of the things that’s most important to me [about founders] is that we can really trust each other — founders that have high integrity, that do what they say they’re gonna do… I’m expecting this relationship to be 5, 10 years plus into, ideally, their next company. I want to work with people who are thinking long term like that. Ideally, these are people who are not just trying to win and beat other people, but really thinking about the question: how do we win together? What have you gotten right and wrong in your investing career so far? When I started talking about AI and machine learning and technology and medicine and bio 10 plus years ago, there was a lot of resistance and a lot of people saying, ‘Oh, that’s never going to happen. That’s never going to be useful,’ and so on. That resistance is largely gone and seeing this arc is very fulfilling. I think it took me some time to really appreciate that as seductive as the coolest technologies are, it really always comes back to go-to-market. I tell my founders, especially the ones who are coming from the science or the product side, for them to take all their brilliance and creativity and really apply it to the go-to-market side, that the go-to-market part is at least as hard or harder than the technology side. Help us understand how you’re designing this new firm differently, compared with what you were running at a16z. Right now, we’re doing something really quite different… VZ is named after me, Vijay, and my co-founder, Zach Werner — he’s the Z. We’re intentionally really quite small… on the investment side, it’s really just the two of us. We were actually intending on hiring associates, but it turned out, with the agents that we’ve built up, not to be something that we need to do. How concentrated is “concentrated”? We’re [not] driving 30 bets per year… we’re talking about probably five, not a lot of investments — very concentrated. Adding a company at a typical fund is like adding a Facebook friend — that’s something you do pretty quickly. For Zach and I, it’s more like . . . wanting to have another child. This is a big deal for us. With that structure, who are you competing against for deals? The funny thing about this model is that typically we’re not trying to compete for a hot round — people make room for us. It’s a very different thing than trying to get the hot Series A or Series B. Largely, people want us as investors because of what Zach and I can do, and how hands-on we can be. When I look at people who are inspirations, I look at someone like Antonio Gracias at Valor — he’s well-known now because of the SpaceX deal, but he’s been doing what he’s been doing for 20 years. What Thrive has done, with a more concentrated portfolio, is also a real inspiration. Obviously, a16z is sort of in my DNA as well, but I think those other ones are new additions to how we think about things. What’s overhyped right now in AI and biotech? The reality is that AI can find insights that we can’t get from just humans alone. The thing that always gets tricky is when there’s this call that AI is going to cure all everything. The reason for hesitance there is not because of any doubt about AI — it’s about doubt of the data. LLMs work because there’s so much data to learn from. When the data is just simply not there, then AI can’t magically solve that problem.
View
Sony Music, Warner sue Anthropic, alleging a “brazen campaign” of intellectual property theft
Sony Music Publishing, Warner Chappell and numerous other music publishers have sued Anthropic and co-founders Dario Amodei and Benjamin Mann, alleging the AI lab conducted a “brazen campaign of illegally torrenting, scraping, and downloading copyrighted works.” The lawsuit, which was filed late Friday in the U.S. District Court for the Northern District of California, was first reported byMusic Business Worldwide. The publishers accuse Anthropic of “blatant theft” by using thousands of copyrighted works to train its AI model Claude. Anthropic could not be reached for comment prior to publication. TechCrunch will update this article if the company responds. This isn’t the first intellectual property lawsuit Anthropic has faced.Some of the same lawyersbehind this lawsuit also represent Concord Music Group and Universal Music Group in a case filed in January and led theBartz v. Anthropiccase, in which a group of authors accused Anthropic of using copyrighted works to train products like Claude. Anthropic was ordered topay $1.5 billionin the landmark Bartz case after a judge ruled that while it was legal for the AI lab to use copyrighted works, it was not legal to acquire that content through piracy. While the cases make similar arguments, there are key differences. This latest lawsuit is particularly broad and builds off the other cases, including by accusing Anthropic of “flagrant piracy” through illegal torrenting to obtain millions of copies of books, including those that contain lyrics and sheet music.
View
Nvidia’s AI advantage is moving beyond the GPU
Before this week, the dominant story about Nvidia went something like this: For the first few years of the AI boom, Nvidia was the only source for state-of-the-art GPUs, which became immensely profitable as the industry scaled out. In the last few years, hyperscalers like Amazon and Google have started building their own chips, and Nvidia is no longer the only game in town, leading many investors to wonder how durable its advantage really is. It’s a compelling story, and mostly true. After growing its market cap 10x between the start of 2023 and mid-2025, Nvidia shares have been on a more modest trajectory for the past year, driven by concerns about GPU competition. A new narrative has taken shape since the company’s earnings on Wednesday and investors are starting to realize that Nvidia’s advantage goes far beyond GPUs. As AI’s compute grows into the gigawatt scale, orchestration has become an increasingly complex task. Not surprisingly, Nvidia has built much of the state-of-the-art hardware needed to handle it, giving the company a huge advantage in the systems that surround the GPU even as it sees increased competition on the GPUs themselves. For all the talk ofcompute as a commodity, it’s still incredibly difficult to operate a megascale data center at peak efficiency — and as deployments get bigger and faster, that challenge is only growing. You can see some of this just by looking at the details of what Nvidia is actually selling. The company is currently rolling out its Vera Rubin architecture, which pairs the Rubin GPU with a collection of other units, including the Vera CPU, the Groq 3 LPX inference accelerator and similar racks for storage and networking. Over the past week, I’ve been talking to folks at Nvidia about what those systems actually do, and the results have been surprising. Like the Rubin GPU itself, they’re extremely specialized systems, but instead of churning through tokens, they’re making sure everything outside the GPU works as efficiently as possible. If the GPU is the engine, these are the rest of the car. The Vera CPU in particular is focused on the problem of orchestrating data. “Vera is important because there’s only so much memory that you can put in a single server or any sort of compute platform,” Jason Hardy, Nvidia’s VP of storage technology, told me. As data centers have scaled up computing power, memory capacity has scaled up too, which is whycompanies like Micronhave gotten rich in the second wave of the infrastructure boom. But getting that data to the GPU at the right time isn’t straightforward — and as companies look to drive tokens-per-watt lower and lower, they’re realizing how important that kind of traffic direction is. “We saw upwards of 3x improvement in these operations, where the Vera CPU is allowing for acceleration,” Hardy said. “So now we can use our flash to its fullest potential, because we can get all that performance out of it without bottlenecking.” You can see versions of the same problem outside of Nvidia. When OpenAI developed its Jalapeño chip, a major focus was avoiding these challenges entirely by minimizing the amount of data that needs to be moved around. “We designed Jalapeño to minimize data movement and communication delays,” the company said in a blog postearlier this month. “Its large domain allows the entire workload to remain within one connected system, minimizing data movement and helping the complete request stay fast and efficient from beginning to end.” It’s a different approach, avoiding data movement entirely by conducting a workload within one integrated chip. But the overall logic is the same, increasing efficiency with smarter traffic control instead of just more processor cycles. That in turn opens up a whole new layer of infrastructure for companies to compete over. This new focus on data orchestration isn’t automatically a win for Nvidia. The company will have to compete with rival chipmakers and hyperscalers just as it has with GPUs. But the competition has moved to a new layer, where building a rival GPU matters less than being able to make the entire system work efficiently. And at least in the early stages, Nvidia looks to have a commanding lead.
View
Gnani AI Launches Sovereign AI Stack With 30B-Parameter Model and AI Agents
Gnani AI says its latest stack is designed to keep sensitive data within an organisation’s own infrastructure while reducing the cost of processing Indian languages.
View
India Has a Frugal Approach to Space Flights. But Cheaper Doesn't Mean Cheap
The challenge increasingly is not simply doing space missions cheaply, but sustaining a more ambitious space programme.
View
