November 16, 2011
Amazon Web Services just announced its most powerful offering yet for supercomputing users that require the power of a large cluster on demand. The newest EC2 Cluster Compute Instance, called Cluster Compute Eight Extra Large (CC2), is aimed at businesses and researchers who require additional HPC capacity in an elastic, pay-as-you-go format.
To highlight the potential of the offering, Amazon created TOP500-class cluster by stringing together 1,064 such instances. The 240.09 teraflop system netted a #42 ranking on the 38th edition of the eminent list announced yesterday at SC11. The company notes that an array of 290 CC2 instances will create a smaller-scale system, with a speed 63.7 teraflops for less than $1,000 per hour.
The new Cluster instance includes 2 Intel Xeon processors, each with 8 cores, for a total of 88 EC2 Compute Units. The 64-bit platform comes with 60.5 GB of RAM and 3.37 TB of instance storage, connected to a 10 Gigabit network. The CC2 moniker comes from the instance type's API name: cc2.8xlarge.
In a blog post, AWS evangelist Jeff Barr, provides additional details:
We've enabled Hyper-Threading, allowing each core to process a pair of instruction streams in parallel. Net-net, there are 32 hardware execution threads and you can expect 88 EC2 Compute Units (ECU's) from this 64-bit instance type. That's nearly 90x the rating of the original EC2 small instance, and almost 3x the rating of the first-generation Cluster Compute instance.
The CC2 instance joins the other 12 instance types available on EC2, each with its own characteristics, such as CPU size, networking, storage, memory, and so on. Like the other instances, CC2 supports a variety of high-performance workloads, including image processing, genome sequencing, seismic analysis, financial modeling and engineering design. Organizations and businesses may choose to use EC2 as their primary computing resource or supplement their on-site cluster as dictated by business needs. For academic users, EC2 can offer a way around the long wait times often associated with departmental clusters.
According to the company literature, the Cluster family of instances provides "proportionally high CPU resources with increased network performance and are well suited for High Performance Compute (HPC) applications and other demanding network-bound applications." In addition to CC2, there are two other instance types in this class: Cluster Compute Quadruple Extra Large (CC1) and Cluster GPU Quadruple Extra Large (CG2). CC1 comes with 2 Intel Xeon X5570 quad-core processors (for a total of 33.5 EC2 Compute Units), 1.690 GB of instance storage in a 64-bit platform. The CG1 instance offers a similar profile, but with the additional processing power of 2 NVIDIA Tesla "Fermi" M2050 GPUs.
The CC2 instance is currently offered as a public beta, available only to users in Amazon's US East Region in Northern Virginia, with the company planning to add other regions throughout 2012. The CC2 instance is priced at $2.40 per hour, although lower a lower total cost may be achieved using Reserved Instances or by bidding on the EC2 Spot Market. Amazon also announced that the pricing for CC1 instances has been reduced to $1.30 per hour.
The AWS staff will be at SC11 through the end of the week. The team is also taking part in a variety of SC11 scheduled activities, listed here.
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?