Amazon has introduced the second generation of its EC2 computing on-demand infrastructure as a service cloud service, with higher performance processing designed for larger applications.
The Second Generation Standard Instances offer 50% higher absolute CPU performance and are optimised for applications such as media encoding, batch processing, caching and web serving.
Amazon is also lowering the pricing of existing EC2 instances. The new offerings use the same Elastic Block Storage system that was blamed for the recent US East Coast outage.
Currently available only in the US, Amazon is providing two versions:
The extra large instance (m3.xlarge) has 15GB of memory and 13 ECU (EC2 Compute Units) spread across four virtual cores.
The double extra large instance (m3.2xlarge) has 30GB of memory and 26 ECU spread across eight virtual cores.
Amazon EC2 CPU instances The micro instance (613 MB of memory) is for lower throughput applications and websites The high memory instances (17.1-68.4GB of memory) are designed for memory-bound applications, including databases and memory caches The high-CPU instances (1.7-7GB of memory) are designed for scaled-out compute-intensive applications, with a higher ratio of CPU relative to memory The cluster compute instances (23-60.5GB of memory) are designed for compute-intensive applications that require high-performance networking The cluster GPU instances (22GB of memory) are for HPC applications using general purpose graphics processing units The high I/O instance (60.5GB of memory) provides very high, low latency, random I/O instance performance
The instances are now available in the northern Virginia region of the US; plans to support them in other regions are in place for early 2013.
The extra large instance is charged at $0.52 while double extra large costs $1.16 per hour.
To tie in with the second generation launch, Amazon has also cut the price of first generation standard (m1) instances running Linux in the northern Virginia and Oregon regions by over 18%.
Amazon is coming under pressure with suppliers like Rackspace and Microsoft, which now offers infrastructure as a service on its Azure cloud platform.
The price reduction first generation EC2 instances is good news, although this is limited to the US. CIOs are still wary of public cloud offerings due to data sovereignty concerns, but more businesses are prepared to consider services like Amazon for cloud-bursting CPU and storage capacity when their own internal datacentres are being stretched.
New Amazon EC2 pricing Instance Type New On Demand Price Old On Demand Price Small $0.065/hour $0.08/hour Medium $0.13/hour $0.16/hour Large $0.26/hour $0.32/hour Extra Large $0.52/hour $0.64/hour
