“The more you buy, the more you save,” Huang said ahead of Computex, an annual technology exhibition held in Taiwan. “It’s called the CEO’s calculations. It’s not exact, but it’s correct.”
Huang explained the concept by describing why companies should invest in both graphics processing units (GPUs) and central processing units (CPUs). Both processors can operate autonomously, reducing the time needed to complete a task from “100 time units to 1,” he explained.
So the more you buy, the more you save. That sounds like a good selling point for a CEO who sells processors.
The combination of the two processors is already standard practice in the personal computing sector. “We add a GPU, a $500 GPU, to a $1,000 PC, and the performance increases dramatically,” he said. “We’re doing this in a data center. A billion-dollar data center, we add $500 million worth of GPUs, and all of a sudden it becomes an AI factory.”
Huang then presented a diagram showing that when companies use both, their speed increases by 100, at only 1.5 times the cost.
In March, Nvidia revealed the Blackwell B200 GPU, a $70,000 chip that it claims is “the most powerful AI chip in the world.” It integrates the chip into larger designs like the GB200 NVL72, which combines 72 GPUs and 36 CPUs and is aimed at “the most compute-intensive workloads” and reduces costs and power consumption by up to 25 times .
In recent months, the chipmaker has made headlines as a key player in the AI boom. He raked in more than $22 billion in the fourth quarter of 2023. Tech executives from Sam Altman to Mark Zuckerberg now depend on its chips to fuel their AI ambitions.
businessinsider
