It’s been a week since Nvidia topped, then closed above the $4 trillion market cap mark, and I’m still having a hard time wrapping my mind around “a graphics card company” becoming the first to achieve this milestone. Some analysts are already anticipating a $5 trillion market cap.

When Nvidia hit $3 trillion—just a year ago!—I linked to a story detailing their near-collapse in 1995. I remain awed by their resilience, but also skeptical of the AI frothiness that’s driving their valuation. 

Tripp Mickle, in his New York Times coverage, notes:

Nvidia’s rise is reminiscent of dot-com era titans like Cisco and Juniper Networks, which built the equipment that ran communications networks for the internet. Cisco’s shares increased more than a thousandfold between its initial public offering in 1990 and 2000, when it briefly became the world’s most valuable company.

What Mickle doesn’t say is that Cisco, valued at more than $550 billion at the height of the dotcom era, is worth only $270 billion today.

Nor does he mention that Juniper Networks was at the time valued at over $40 billion, with a share price over $244. It was acquired earlier this month by Hewlett Packard for $14 billion, or $39.95 a share.

I’m not predicting a similar trajectory for Nvidia—that would be foolhardy—but this unprecedented rise from $1 trillion to $4 trillion in two years gives me pause.

Of course, I also didn’t buy Nvidia at $45 like I was told to, so perhaps don’t come to me for stock advice.

No doubt I’ll be writing a “$5 trillion valuation” piece within the year.