Engineering project hub

Note:Now you can watch this blog on  youtube I think GPUs are aging like milk right now. Every year, there’s a “ new god GPU ” on the market that makes the previous one feel instantly outdated. We’ve already seen this: first the H100 , then the H200 , and now the Blackwell B200 , each bringing insane performance jumps. In my view, buying a new flagship GPU is often a waste of money. Let me tell you why: You buy a flagship card, and 6–12 months later, the resale value is trash because new software demands even more VRAM and TFLOPS again. The situation in data centers is even worse. Companies are pouring hundreds of billions into GPUs that run at crazy power levels all day. A significant chunk of them either die or become “old” within just 1–2 years . That is a lot of silicon going straight to the graveyard. For normal people like us, the lesson is simple: don’t treat your laptop or PC like a mini data center. If you try to run AI workloads 24/7, you will cook your hardware in 1–2...

  I have seen that Nvidia doesn’t want to be just a chips company  Anymore In the past 3 years, it has grown so fast that it became the first company to reach a market value of 5 trillion dollars . But Nvidia is not stopping there. They want to become an ecosystem that no one can shake. We can already see a huge change in their investments. In the last year and this year, they invested in Nokia , one of the biggest network-development companies, to work on 6G . They also invested in  Intel  and as we know, Intel is one of the two major chips used in laptops. Nvidia wants to be inside every laptop or PC in the world. Their only big rivals in chips now are Google and Apple , because they use their own chips for their own products. By making Nvidia chips common in all laptops and PCs, developers will start using Nvidia for research and product building. Now Nvidia is also secretly looking to build their own smartphones , where AI can run offline with full customizatio...