Locator: 51914TRAINIUM.
The bubble [again]: again, unlike the dot-com boom / bubble, today's companies are paying for real hardware / real products with real money. In fact, what's separates the current AI revolution with the dot-com bust? Three things:
- real money is being generated; real money is being spent; earnings are blowing past estimates;
- there are only a handful of Fortune 500 companies involved; it's not that hard to track; and,
- the moats of the current Fortune 500 / top AI companies have very, very large moats; money is just one moat.
Next example, link here. Note: from GPUs (Nvidia) to Trainium (Amazon
Would Decort have been a good fit for Google / YouTube?
*****************************
Trainium
Previously posted:
Locator: 50719AWS.
[This was posted May 5, 2026, link here.]
There's a gazillion things going on
right now in AI, but if I had to list the top three that investors need
to watch, these are the three:
- memory;
- accelerators; and,
- LDCs.
As an aside, the above --
- memory: MICRON (MU);
- accelerators: AWS: and,
- LDCs: Texas.
This post is on accelerators: link here.
A reminder, Graviton, link here. August 31, 2025:
Amazon chips: wiki. Must-read. Annapurna Labs -- an Israeli company now fully owned by Amazon.
In November 2024 Annapurna announced their second generation Trainium 2 intended for training AI models. Based on their internal testing, Amazon are claiming "a 4-times performance increase between Trainium 1 and Trainium 2".AWS Graviton - an ARM-based CPU developed by Annapurna Labs for exclusive use by Amazon Web Services.Note: some sources call them "Trainum 2" chips and other sources call them "Trainium2" chips. In AI, "spaces" are "tokens."AWS (Amazon): new chips -- Ocelot (quantum computing); competing with Nvidia -- Trainium, Inferentia. World's largest supercomputer to be completed in 2025 (?) -- Project Rainier -- partnered with Anthropic, AI --> Claude.
Beth: Amazon's chip business, run rate -- Graviton, Trainium, and Nitro on EC2, is now $20 billion and growing triple-digits y/y. Link here. Link here.
Now, today, link here.







