source: freepik

Investing Lesson: Mismatch, Moore’s Law. Nvidia

2 min readMay 23, 2021


Every problem opens an opportunity. That is a business lesson.

Nvidia is a company that sees a problem as an opportunity. Before the booming of AI, none see Nvidia like today.

We may ask, what the h*ll gaming card maker do in the AI business? or, maybe a few yers later, we ask, “Why Nvidia is here in the automotive industry?”

Nvidia's journey of transition begins with Moore Law.

In simple term, Moore’s Law state that due to human and industrial demand of computing power, the computer processor have to double its capability every two years.

At a certain level, this demand can be satisfied — until the chip dimension takes the issue.

The capability of the processor determined by the number of transistors bundled in its chip — more transistors means more computing power. Applying Moore’s Law in 2050, to meet the rapidly growing demand, we have two possibilities:

  1. We have a transistor in a Hydrogen atom size or
  2. Have a laptop as large as your living room

Both of them seem makes no sense and to add another constraint, it’s also increasingly — exponentially expensive — for the manufacturer to keep that pace. So, we have to violate the Mooe Law.

NVIDIA: GPU as a solution, Accelerated Computing

Turn away from CPU, the industry tries to look at another solution — another chip - the GPU.

The GPU — once being used exclusively for gaming purposes — began considered as a computational unit. GPU has a unique architectural structure. It has more semiconductor density than CPU. Meaning that for the same surface, GPU has more transistors. This configuration is optimal for a simpler job. From the user perspective, GPU work in a simple — parallel manner. CPU, the counterpart, operates in a complicated — serial type function.

GPU advantages don’t make this Integrated Circuit replace the CPU. Both have their own advantages, thus, it will be best to employ these two kinds of the chip to work together. GPU will take the burden of simple yet intensive sections of the applications. Giving space for CPU to concentrate the remaining complex sections

NVIDIA(NYSE: NVDA) — the inventor of GPU — calls this process accelerated computing. And in the upcoming AI application in a wide range of industries, the firm has a strong position to catch the opportunity due to its status as a pioneer.