Welcome to Fractal Forums

Fractal Software => Programming => Topic started by: ker2x on November 15, 2010, 08:58:39 AM




Title: Announcing Cluster GPU Instances for Amazon EC2
Post by: ker2x on November 15, 2010, 08:58:39 AM
http://aws.amazon.com/about-aws/whats-new/2010/11/15/announcing-cluster-gpu-instances-for-amazon-ec2/

$2.10 per hour

The Cluster GPU instance family currently contains a single instance type, the Cluster GPU Quadruple Extra Large with the following specifications:

22 GB of memory
33.5 EC2 Compute Units (2 x Intel Xeon X5570, quad-core “Nehalem” architecture)
2 x NVIDIA Tesla “Fermi” M2050 GPUs
1690 GB of instance storage
64-bit platform
I/O Performance: Very High (10 Gigabit Ethernet)
API name: cg1.4xlarge


Title: Re: Announcing Cluster GPU Instances for Amazon EC2
Post by: cKleinhuis on November 15, 2010, 10:07:07 AM
wow, computing power is reaching massiveareas, i find it funny how the human beings adopt to the limit of ~4Ghz, since the last decade, massive
parallelism has reached us, i think that many of the "classic" algorithms need a redesign concerning parallel throughput, e.g. sorting an array of
numbers

amazing news, thx


Title: Re: Announcing Cluster GPU Instances for Amazon EC2
Post by: ker2x on November 15, 2010, 11:06:23 AM
i'm glad i postponed my purchase of a tesla card (or a bi-gpu clevo laptop, i wasn't sure yet).