![\"\"](\"https:\/\/www.technologyinvestor.com\/wp-content\/uploads\/2024\/05\/NVDA-One-Year-again.jpg\")
<\/p>\nWell, what now?<\/p>\n
After a stupendous run in the last 12 months, Nvidia is now the third most valuable company in the U.S.<\/p>\n
<\/p>\n
1. Microsoft Corporation with a market cap of $2,988 billion<\/p>\n
2. Apple Inc. with a market cap of $2,612 billion<\/p>\n
3. Nvidia Corporation with a market cap of $2,183 billion<\/p>\n
4. Amazon.com, Inc. with a market cap of $1,802 billion<\/p>\n
5. Alphabet Inc. with a market cap of $1,640 billion<\/p>\n
Will Nvidia’s run continue? My answer is Yes. To wit:<\/p>\n
+ 10 for 1 stock split coming.
\n+ New faster chips coming. Blackwell.
\n+ Huge demand, especially from data centers.
\n+ Jensen. There’s no CEO with the charisma and vision of Jensen Huang.<\/p>\n
Says Friday’s Wall Street Journal:<\/em><\/p>\n The ride is far from over” in reference to Nvidia’s continued success in selling its older chips while anticipating the release of its new AI product line. .. Wall Street expects Nvidia’s annual data center revenue to exceed $100 billion over the next year, up from nearly $60 billion in the previous year.<\/span><\/p>\n Here’s the Wall Street Journal’s chart.<\/p>\n If you’re not thoroughly bored with my enthusiasm for Nvidia…<\/p>\n Here’s The Economist’s latest piece:<\/p>\n Can Nvidia be dethroned? Meet the startups vying for its crown<\/strong> \u201cH<\/span>e who<\/small>\u00a0controls<\/small> the GPUs, controls the universe.\u201d This spin on a famous line from \u201cDune<\/a>\u201d, a science-fiction classic, is commonly heard these days. Access to GPUs, and in particular those made by Nvidia, the leading supplier, is vital for any company that wants to be taken seriously in artificial intelligence (AI). Analysts talk of companies being \u201cGPU-rich\u201d or \u201cGPU-poor\u201d, depending on how many of the chips they have. Tech bosses boast of their giant stockpiles. Nvidia\u2019s dominance<\/a>\u00a0has pushed its market value above $2trn. On May 22nd it reported that its sales for the quarter ending in April grew by 262%, year on year (see chart).<\/p>\n GPUs do the computational heavy lifting needed to train and operate large AI models<\/a>. Yet, oddly, this is not what they were designed for. The initials stand for \u201cgraphics processing unit\u201d, because such chips were originally designed to process video-game graphics. It turned out that, fortunately for Nvidia, they could be repurposed for AI <\/small>workloads.<\/p>\n Might it be better to design specialist AI chips from scratch? That is what many companies, small and large, are now doing in a bid to topple Nvidia. Dedicated AI chips promise to make building and running AI models faster, cheaper or both. Any firm that can mount a credible threat to the reigning champion will have no shortage of customers, who dislike its lofty prices and limited supplies.<\/p>\n Ordinary processing chips, like those found inside laptop and desktop computers, are in essence designed to do one thing after another. GPUs, by contrast, contain several thousand processing engines, or \u201ccores\u201d, which let them run thousands of versions of the same simple task (like drawing part of a scene) at the same time. Running AI models similarly involves running lots of copies of the same task in parallel. Figuring out how to rewrite AI code to run on GPUs was one of the factors that triggered the current AI\u00a0<\/small>boom.<\/p>\n Yet GPUs have their limitations, particularly when it comes to the speed with which data can be shuffled on and off them. Modern AI models run on large numbers of interconnected GPUs and memory chips. Moving data quickly between them is central to performance. When training very large AI models, some GPU cores may be idle as much as half of the time as they wait for data. Andrew Feldman, the boss of Cerebras, a startup based in Sunnyvale, California, likens it to the gridlock in a grocery store on the day before Thanksgiving. \u201cEverybody\u2019s in a queue, so there are blockages in the parking lot, there are blockages in the aisles, blockages at the checkout. That\u2019s exactly what\u2019s happening with a GPU.\u201d<\/p>\n Cerebras\u2019s response is to put 900,000 cores, plus lots of memory, onto a single, enormous chip, to reduce the complexity of connecting up multiple chips and piping data between them. Its CS-3 chip is the largest in the world by a factor of 50. \u201cOur chip is the size of a dinner plate\u2014a GPU is the size of a postage stamp,\u201d says Mr Feldman. On-chip connections between cores operate hundreds of times faster than connections between separate GPUs, Cerebras claims, while its approach reduces energy consumption by more than half, for a given level of performance, compared with Nvidia\u2019s most powerful GPU offering.<\/p>\n Groq, another startup, is taking a different approach. Its\u00a0ai<\/small> chips, called language processing units (LPUs), are optimised to run large language models (LLMs) particularly quickly. In addition to containing their own memory, these chips also act as routers, passing data among the interconnected LPUs. Clever routing software eliminates the variation in latency, or time spent waiting for data, allowing the whole system to run in lockstep. This greatly boosts efficiency, and thus speed: Groq says its LPUs can run big LLMs<\/small>\u00a0ten times faster than existing systems.<\/p>\n Yet another approach is that taken by MatX, also based in California. GPUs contain features and circuitry that provide flexibility for graphics, but are not needed for LLMs, says Reiner Pope, one of the firm\u2019s co-founders. The GPU-like chip his firm is working on gets rid of such unnecessary cruft, boosting performance by doing fewer things better.<\/p>\n Other startups in this area include Hailo, based in Israel; Taalas, based in Toronto; Tenstorrent, an American firm using the open-source RISC v<\/small> architecture to build AI chips; and Graphcore, a British company that is thought to be about to sell itself to SoftBank, a Japanese conglomerate. Big tech firms are also building ai<\/small>\u00a0chips. Google has developed its own \u201ctensor processing units\u201d (tpu<\/small>s), which it makes available as a cloud-computing service. (It unveiled its latest version on May 14th.) Amazon, Meta and Microsoft have also made custom chips for cloud-based\u00a0ai<\/small>; Openai<\/small>\u00a0is planning to do so as well.\u00a0amd<\/small>\u00a0and Intel, two big incumbent chipmakers, make\u00a0gpu<\/small>-like chips already.<\/p>\n One danger for the newcomers is that their efforts at specialisation could go too far. Designing a chip typically takes two or three years, says Christos Kozyrakis, a computer scientist at Stanford University, which is \u201ca huge amount of time\u201d given how quickly\u00a0ai<\/small>\u00a0models are improving. The opportunity, he says, is that the startups could end up with a chip that is better at running future models than Nvidia\u2019s less specialised\u00a0gpu<\/small>s are. The risk is that they specialise in the wrong thing.<\/p>\n Having previously worked at Google, which developed the currently dominant \u201ctransformer\u201d architecture used in\u00a0llm<\/small>s, Mr Pope of MatX is confident that his firm has \u201ca somewhat good crystal ball\u201d. And if a new approach comes along\u2014\u201cstate-space models\u201d are the latest thing\u2014its chip is versatile enough to adapt, he says. Mr Feldman says all modern AI is still just \u201csparse linear algebra\u201d under the hood, which Cerebras\u2019s chip can do very quickly.<\/p>\n Is greatness a transitory experience?<\/p>\n Another challenge is that Nvidia\u2019s software layer for programming its GPUs, known as CUDA, is a de facto industry standard, despite being notoriously fiddly to use. \u201cSoftware is king,\u201d says Mr Kozyrakis of Stanford, and Nvidia has a significant advantage, having built up its software ecosystem over many years. AI-chip startups will succeed only if they can persuade programmers to rejig their code to run on their new chips. They offer software toolkits to do this, and provide compatibility with the major machine-learning frameworks. But tweaking software to optimise performance on a new architecture is a difficult and complex business\u2014yet another reason Nvidia is hard to dislodge.<\/p>\n The biggest customers for AI chips, and the systems built around them, include model-builders (such as OpenAI, Anthropic and Mistral) and tech giants (such as Amazon, Meta, Microsoft and Google). It may make sense for such companies to acquire an AI-chip startup, and keep its technology to themselves, in the hope of besting the competition. Instead of trying to compete with Nvidia, chip startups could position themselves as acquisition targets.<\/p>\n Mr Pope says MatX is targeting the \u201ctop tier\u201d of the market, which suggests that it hopes to sell its chips\u2014if not the whole company\u2014to the likes of OpenAI, Google or Anthropic, whose AI models are the most advanced. \u201cWe would be happy with many kinds of exit,\u201d he says, \u201cbut we think there is a sustainable business here as a standalone company.\u201d That remains to be seen. Cerebras, for its part, is said to be preparing for an initial public offering. So far none of the startups has made a dent in Nvidia\u2019s dominant position. Plenty of people are hoping that one of them will. \u25a0<\/span><\/p>\n I am seriously worried about Google<\/strong><\/span><\/p>\n + Perplexity and ChatGPT\/Chat.40 are much better search engines for research and information.<\/p>\n + Amazon is\u00a0 a much better search engine for finding the products you want to buy.<\/p>\n And Google? You seen anything new and exciting to come out of it recently? I’ve basically given it up for Perplexity. I go to Perplexity first. My son Michael does too.<\/p>\n How Exercise Strengthens Your Brain<\/strong><\/span><\/p>\n I’m obsessive about playing aggressive single tennis every day. Will I live longer, healthier and be more alert? The studies say I will. But that’s not the motivation. My real motivation is that it simply makes me feel good.<\/p>\n Segue to today’s New York Times.<\/em><\/p>\n Physical activity improves cognitive and mental health in all sorts of ways. Here\u2019s why, and how to reap the benefits:<\/p>\n The brain benefits really start to emerge, though, when we work out consistently over time.\u00a0Studies show<\/a>\u00a0that people who work out several times a week have higher cognitive test scores, on average, than people who are more sedentary. Other research has found that a person\u2019s cognition\u00a0tends to improve<\/a> after participating in a new aerobic exercise program for several months. …<\/p>\n Physical activity also\u00a0benefits mood<\/a>. People who work out regularly report having\u00a0better mental health<\/a>\u00a0than people who are sedentary. And\u00a0exercise programs<\/a>\u00a0can be effective at treating people\u2019s depression, leading some psychiatrists and therapists to prescribe physical activity. The Centers for Disease Control and Prevention\u2019s\u00a0recommendation<\/a>\u00a0of 150 minutes of moderate aerobic activity or 75 minutes of vigorous aerobic activity per week is a good benchmark.<\/p>\n<\/div>\n<\/div>\n Perhaps most remarkable, exercise offers protection<\/a>\u00a0against neurodegenerative diseases. \u201cPhysical activity is one of the health behaviors that\u2019s shown to be the most beneficial for cognitive function and reducing risk of Alzheimer\u2019s and dementia,\u201d said Michelle Voss, an associate professor of psychological and brain sciences at the University of Iowa.<\/p>\n What kind of exercise is best for your brain?<\/strong><\/p>\n The experts emphasized that any exercise is good, and the type of activity doesn\u2019t seem to matter, though most of the research has involved aerobic exercise. But, they added, higher-intensity workouts do appear to\u00a0confer a bigger benefit<\/a>\u00a0for the brain.<\/p>\n Improving your overall\u00a0cardiovascular fitness level<\/a> also appears to be key. \u201cIt\u2019s dose-dependent,\u201d Dr. Wrann said. \u201cThe more you can improve your cardiorespiratory fitness, the better the benefits are.\u201d<\/p>\n![\"\"](\"https:\/\/www.technologyinvestor.com\/wp-content\/uploads\/2024\/05\/Plugged-in-again.jpg\")
<\/p>\n
\nA new generation of AI chips is on the way<\/p>\n\n![](\"https:\/\/www.economist.com\/cdn-cgi\/image\/width=1424,quality=80,format=auto\/content-assets\/images\/20240525_WBC043.png\")
\n\n\n\n\n
![\"\"](\"https:\/\/www.technologyinvestor.com\/wp-content\/uploads\/2024\/05\/Whimpy-Kid.jpg\")
![\"\"](\"https:\/\/www.technologyinvestor.com\/wp-content\/uploads\/2024\/05\/Captainn-underpants.jpg\")
<\/p>\n![\"\"](\"https:\/\/www.technologyinvestor.com\/wp-content\/uploads\/2024\/05\/Seat-Belt.jpg\")
<\/p>\n