New Paradigm, New Chips, New Data Center, New Cloud: It’s About AI, Stupid
Quotes of the Week
Upgrades and Numbers
Graphene AI Chips
INVNT/IP and CFIUS
From PRP to CHASE-CI
Takeout Window
Larsen C Breaks Up. It’s the Ice, Trump.
The RGB of North Korea
Ethermail
—-
Today, the tech world seems to be focused on two things: Cloud Computing and Artificial Intelligence, or AI (and its semantic variants). Pretty simple, eh? Amazon, Microsoft, Google, Oracle, Salesforce, and many other companies are having wonderful earnings reports due mostly, and sometimes entirely, to the new popularity of cloud computing.
Cloud happens in data centers, and the hottest use of them today is for AI. Data centers are built with many (thousands, usually) servers, increasingly running two kinds of chips: standard central processing units (CPUs), and graphic processing units (GPUs).
So it wouldn’t take a rocket scientist to decide that: a) servers are the most important piece of hardware being sold today; or b) that their sales must be on an explosive growth curve. One could go even further, on the international stage, and suggest that, since the US invented all of this, US companies must be riding a fantastic Sales Tsunami. Right?
Wrong.
In this week’s discussion, we’ll look at the arcane, slow-moving, very conservative market for servers, and at the chips that drive them, to find out what’s really going on, who’s making the money, why they aren’t making more, and what disruptions are under way or just around the corner.
