Wednesday, November 6th 2024
NVIDIA Surpasses Apple as the World's Most Valuable Company
Valued at $3.43 trillion against Apple's $3.38 trillion at the time of this writing, NVIDIA Corporation is now the most valuable company in the world. NVIDIA has been the hottest tech stock since 2021, as it created the most valuable IP of this decade—AI GPUs, which were ready just in time for OpenAI and ChatGPT to open its doors, jumpstarting the generative AI revolution. NVIDIA's AI GPUs didn't come to being overnight, the company had been investing in the parallel computing and HPC space since the 2000s, with its CUDA programming language that lets application exploit the SIMD parallelism of its GPUs. In the mid-2010s, the company created the Tensor core, accelerating AI, and invested heavily in AI frameworks. Some of the most fundamental research leading up to GPT was done on NVIDIA GPUs. This would lead to the big payoff in the 2020s, helping NVIDIA become a multi-trillion-dollar company. Founding CEO Jensen Huang has been NVIDIA's brains and spine throughout the company's 30+ year journey, and its success could not be possible without him at the reins.
Source:
CNBC
38 Comments on NVIDIA Surpasses Apple as the World's Most Valuable Company
He looks so happy I think he'll give Blackwell away on deep sale.
But if they can enter the CPU market as well, that would be a real beast of a company.
~61% of the x86 CPU market
~80+% on the GPU market
www.techpowerup.com/328422/nvidia-cpus-not-gpus-coming-in-2025
Personally I would love to own an 5090/5080, but wouldn't get an intel CPU in the next 2~3 years. Plans might change, they often do.
I never said, that they are actually going to buy up intel.
There's also a mobile SoC which google calls Tensor G#, which can be found in their phones. These came out first in 2021 with the Pixel 6
How about a well priced new generation of GPU's to celebrate this milestone Huang?
ChatGPT is pretty cool but I'm not sure it's going to remain cool as the ratio of AI-generated stuff to human-curated material on the web starts to affect the accuracy of training. I can only see the potential for a positive feedback loop where AI is trained on AI-influenced content and reinforces its own mistakes.