AWS Launches Graviton3 for Boosting Cryptographic, Machine Learning Workloads!
Amazon Web Services (AWS) is pulling out all the stops to gain an edge in the cloud ecosystem. Embracing the same spirit, the leading cloud services provider launched the third generation of its AWS Graviton chip-powered instances, the AWS Graviton3.
The Graviton3 will amplify the performance of all-new Amazon Elastic Compute 2 (EC2) C7g instances, which are currently available in preview. These Graviton3-powered instances will offer almost 25% faster computing performance and 2x times more floating-point performance than the current generation of AWS EC2 C6g Graviton2-powered instances.
Moreover, the Graviton3-powered instances are 2x times faster at performing cryptographic workloads. They will offer 3x more excellent performance for machine learning workloads than Graviton2-powered instances. The Graviton3-powered instances also support bfloat16 data.
Also Read: AWS Suffers Yet Another Outage!
On the networking side, C7g instances offer up to 30 Gbps of network bandwidth and Elastic Fabric Adapter (EFA) support.
“Graviton3 is up to 25% quicker for general-compute tasks, with two times faster floating-point performance for scientific workloads, two times faster performance for cryptography workloads, and three times faster performance for machine learning workloads,” said AWS CEO Adam Selipsky.
“Furthermore, Graviton3 uses up to 60% less energy for the same performance as the previous version.”
According to AWS’ chief evangelist Jeff Barr, the c7g instances “are going to be a great match for your compute-intensive workloads: HPC, batch processing, electronic design automation (EDA), media encoding, scientific modeling, ad serving, distributed analytics, and CPU-based machine learning inferencing.”
Graviton3 also features a new pointer authentication function that aims at improving overall security.
“Before return addresses are pushed onto the stack, they are first signed with a secret key and additional context information, including the current value of the stack pointer. When the signed addresses are popped off the stack, they are validated before being used. An exception is raised if the address is not valid, thereby blocking attacks that work by overwriting the stack contents with the address of harmful code,” said Jeff Barr.
According to AWS, the Graviton3-powered C7g instances come in multiple sizes, including bare metal.
The C7g instances are also equipped with DDR5 memory, the first time in the cloud industry. Compared to the DDR4 memory used in the current generation of EC2 instances, the DDR5 memory facilitates 50% higher bandwidth and consumes less power.
C7g and the I-family are the two new Graviton3-powered instances announced by AWS this week to help cloud customers better the performance, cost, and power consumption of the workloads running on Amazon EC2.
Alongside Gravition3 processors, Amazon announced Trn1, a new machine learning training instance. It offers up to 800Gbps of networking and bandwidth, making it suitable for large-scale, multi-node distributed training use cases.
In addition, this instance supports training deep learning models on the cloud, including models for image recognition, natural language processing, fraud detection, and forecasting.
Trn1 runs on Trainium, an Amazon-designed processor that offers the most teraflops of any cloud machine learning instance.
More New Articles:
- AWS – Perfect Choice for Disaster Recovery
- AWS Suffers Another Service Disruption in the Same Week!
- AWS Vs Azure Vs GCP
- Global Cloud Market to Reach Nearly USD 800 Billion by 2028: Survey
- AWS Quarterly Revenue Growth Rebounds, Hits USD 16.1 Billion!
- Steering Clear of Barriers to Realize Cloud Full Potential!
- Multi-cloud Strategy Emerging as the New Cloud Reality