Trainium gets a bad rap in the AI industry too much. I think it is very valuable for $AMZN. Trainium is excellent hardware in general and can be extremely performant if you use NKI and write everything hand tuned. The Neuron compiler also has very sophisticated code generation. This is something I do not hear talked about much. Neuron compiler can codegen entire Transformers layers. Now I still think NKI is faster, but if you do not need roofline performance you can use the compiler codegeneration. TPU codegen is still much better than Trainium but I would put Trainium definitely up there in terms of codegen quality.

Sponsor news from @awsdevelopers: PyTorch now runs natively on AWS Trainium. See it live at PyTorch Conference NA (Oct 20-21, San Jose) where Robert Hundt, VP Distinguished Engineer at Amazon, will keynote, followed by a live demo from Maen Suleiman, Principal PM, AWS Neuron. Keynote: Demo: #PyTorchCon# #AWS# #Trainium# #OpenSource# #AI#

Morgan Stanley: We see the major customers of EMIB-M being Trainium3, upcoming Trainium 4 and its in-house server CPU. Average EMIB-T capacity is around 5kwpm in 2026, which we expect to further scale to 15-20kwpm in 2027 and 40-45kwpm in 2028. $INTC $AMZN

From torch.profiler to Hardware Cycles: A Practical Profiling Playbook AWS Trainium: designed from the ground up for next generation AI training and inference is now a standard PyTorch device. When you need to optimize, the tools are built in. Esha Lakhotia and Pinak Panigrahi of @awscloud will walk you from torch.profiler capture to cycle-level hardware debugging, with AI-assisted analysis that suggests fixes at PyTorch Conference North America. 📅 Oct 21 | 12:20 PM | LL21ABC Join us in San Jose, October 20-21: #PyTorchCon#

$AMZN just launched TorchNeuron which lets developers run PyTorch AI models on Trainium2 & Trainium3 by changing just one setting instead of rewriting code for $NVDA CUDA. That makes it easier to switch chips & could help drive more Trainium adoption.