- Amazon has recently unveiled two next-generation processors that will power AI models with AWS in the future.
- These two processors – the Graviton4 and the Trainium2 – will be produced by AWS.
- According to Amazon, these processors provide 30% better performance and 75% more bandwidth.
Amazon Web Services (AWS) is one of the biggest cloud computing services in the world right now. However, to make cloud computing more accessible, in addition to integrating AI with it seamlessly, AWS plans to create its own commercially available processors for businesses. This is where the new Graviton4 and the Trainium2 processors come into the picture.
In the ongoing AWS re: Invent 2023 event held by Amazon on November 27, they announced their plans to produce their own processors. However, they will not be powerful computer processors like the Intel Core i9 13900KS. Instead, they will be powerful “AI-powered processors” for servers, primarily targeting businesses.
The Graviton4 and the Trainium2 processors will power the next-generation machine learning and language-learning AI models. According to AWS CEO Adam Selipsky, they will provide better performance and efficiency than current-generation processors.
This is what AWS Compute and Networking Vice President David Brown has to say regarding these processors:
Silicon underpins every customer workload, making it a critical area of innovation for AWS. By focusing our chip designs on real workloads that matter to customers, we’re able to deliver the most advanced cloud infrastructure to them.
They expect this processor to deliver 30% netter computational performance with 75% more memory bandwidth. In addition, according to Adam Brown, this is possible due to these processors having 50% more cores than any current-gen processor.
Such advancements will make all future AI models and APIs perform faster, being able to handle higher workloads. In addition, these processors are made to cost less compared to the energy they will require to run.
So what are your thoughts on these processors? Will it revolutionize the way we use AI models now? Let us know in the comments below!