AWS announces P3 instances to accelerate machine learning

Amazon Web Services has announced a new set of powerful GPU instances to speed-up machine learning

Amazon Web Services has announced a new GPU array to support compute-intensive applications that require high performance speeds.

P3 instances, the next generation of the Amazon Elastic Compute Cloud, are aimed at supporting areas, such as machine learning, computational modeling, genomics and driverless car systems.

These are the first instances to include Nvidia Tesla V100 GPUs, which Nvidia claims is the world's most powerful data centre GPU.

HPE adds that, using P3 instances, companies can build and deploy advanced systems with 14 times better performance than its previous-generation Amazon EC2 GPU compute instances.

Often, training machine learning applications can be a lengthy task, but HPE suggested that these instances reduce the time of the process from days to hours.

The company said the instances carry up to eight Tesla V100 GPUs, providing one petaflop of mixed-precision, 125 teraflops of single precision and 62 teraflops of double-precision floating point performance.

Users can also access a 300GB/s second-generation Nvidia NVLink interconnect, enabling high-speed, low-latency GPU-to-GPU communication. And the instances are also supported by 64 vCPUs, which are based on custom versions of Intel E5 processors.

Each comes with 488GB of DRAM and 25 Gbps of dedicated aggregate network bandwidth too. The latter is achieved thanks to an elastic network adapter (ENA).

Matt Garman, vice president of Amazon EC2, said: "When we launched our P2 instances last year, we couldn't believe how quickly people adopted them.

"Most of the machine learning in the cloud today is done on P2 instances, yet customers continue to be hungry for more powerful instances.

"By offering up to 14 times better performance than P2 instances, P3 instances will significantly reduce the time involved in training machine learning models, providing agility for developers to experiment, and optimizing machine learning without requiring large investments in on-premises GPU clusters.

"In addition, high performance computing applications will benefit from up to 2.7 times improvement in double-precision floating point performance."

Online vacation and home rentals service Airbnb is one of the companies already using the instances.

Nick Handel, who leads machine learning at the firm, said: "At Airbnb, we're using machine learning to optimise search recommendations and improve dynamic pricing guidance for hosts, both of which translate to increased booking conversions.

These use-cases are highly specific to our industry and require machine learning models that use several different types of data sources, such as guest and host preferences, listing location and condition, seasonality, and price.

"With Amazon EC2 P3 instances, we have the ability to run training workloads faster, enabling us to iterate more, build better machine learning models and reduce cost."