Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Strategies & Market Trends : The Financial Collapse of 2001 Unwinding -- Ignore unavailable to you. Want to Upgrade?

To: Elroy Jetson who wrote (12862)	8/4/2024 11:59:04 AM
From: elmatador	Read Replies (1) \| Respond to of 13784

I have been studying this to see how it develops: Yes that is correct: "AI training" operations makes up a lot of the data center cloud workload currently, and there's a great deal of flexibility in what geography that's run in - primarily it's run where there is spare capacity. --------------------------------------------------------------------------------------------------------------------------------------------- Yes that is correct: But this is very unusual and short-lived situation for data center work. Increasingly the AI workload will be inference, and again the primary consideration for where the work is run will be latency - which means the work will be performed where it is requested and needed. -----------------------------------------------------------------------------------------------------------------------------------------------Let's give some examples from the top of my head just to give an idea of the spread: Where there is a hub of activities for a given sector: Aeronautics. The AI work loads will be, geographically, in that sector requires, Colorado (Denver, Colorado Springs) -LM, NG, Sierra Nevada, Alabama (Huntsville) - Boeing, LM, Collins Florida (Central, Jacksonville) - LM Arizona (Tucson, Phoenix) - Raytheon, Collins California (Los Angeles) - Boeing, Raytheon, LM, NG Washington (Seattle) - Boeing Soybeans and Corn. Brazil center west and Corn belt US OIl Gas. Huston Dubai Norway Brazil This is correct Data centers in Kenya will be providing computer work for Kenyans and Kenyan companies and government organizations. If any country need computing power,Cloud providers sell that Infrastrcture as a Service from a local datacenter This is not correct Power and data center costs might be less expensive in Colorado, USA but Kenyans will pay more to have their answers provided in real time from a center in Kenya. Obviously one would buy where it is cheaper It's what the industry you worked in calls edge computing. A processing lag from using a cheaper data center elsewhere in the world isn't acceptable for most uses. If not acceptable, due to costs and latency, it logically follows the workload will be located at the edge near the end point. -----------------------------------------------------------------------------------------------------------------------------------------------