No. of Recommendations: 1
I don't disagree with the overall conditions, but Cisco picked up Splunk cheap which also helped, $1.3 is < 10% their revenue, and based on a cursory search their 2000 GAAP EPS were $.36 and they were $2.54 in the last fiscal year. They are a bigger and more important company than 25 years ago and are a bad bubble indicator.
The hyperscalers don't use CSCO in terms of net dollar spend as CSCO would have captures back then. You have a lot more SDN (or software defined networking) and custom hardware and silicon.
https://techblog.comsoc.org/2025/09/01/hyperscaler... is a fun read I found. OCP gets mentioned a lot (Open Compute Stack) and that has additional interesting reading if you follow up on it.
The self-dealing between the real suppliers seems to be orders of magnitude greater than CSCO's slice of the revenue. NVDA's OpenAI investment really has invited overdue scrutiny.
I sometimes think about the intrigue around OpenAI's leadership transition and transition from a nonprofit and the people involved in that episode.
