TeraGrid Allocation Requests
The TeraGrid is now accepting Research Allocation Requests for the allocation period, July 1, 2011 to June 30, 2012. The submission period is from March 15 until April 15. Please review the new TG systems and important policy changes (see below) before you submit your allocation request through the POPS URL or through the Allocations tab of the TG User Portal.
- Allocation Request Submission Deadline: April 15, 2011 midnight local time
The TeraGrid will host two teleconferences to provide step-by-step instructions and answers to your questions on how to write and submit a successful proposal.
- Allocation Request Guidelines Telecon: March 16, 2:00-3:30 CDT (central)
- Allocation Request Guidelines Telecon: March 25, 2:00-3:30 CDT (central)
Call 1-866-740-1260 and enter Access Code 8229741.
To view the presentation slides online during the talk, visit readytalk during the call and enter the same access code in the Participant Login box.(For our records, please enter your name, email and institution when requested.)
NEW TeraGrid Resources:
See the Resource Catalog for a list of TG compute, visualization and storage resources, and more details on the new systems.
- The new TACC Dell PowerEdge (Westmere) cluster, Lonestar, has been in production since Feb.1, 2011. System Info: 302TFLOPS, 1888 nodes of dual-CPU 3.3GHz 6-core Westmere processors, QDR IB, 1PB Lustre file system.
- The new SDSC Appro Cluster, Trestles, has been in production since Jan. 1, 2011. System Info: 100TFLOPS; 324 compute nodes, 10368 cores (4 8-core 2.4 GHz AMD Magny-Cours processors/node), 20.25TB of RAM (64 GB/node, 2GB/core), and 38TB of flash disk (120GB/node) and a 140TB Lustre file system; QDR IB.
- The NCSA Abe and LONI Queen Bee machines of the AQS pool will retire in late April and July 31, respectively, and only the Purdue Steele machine will be available as a separate allocable system (Steele).
- The new PSC Altix UV System, Blacklight, went into production Jan. 1, 2011. System Info: 512 Nehalem 2.26GHz 8-core processors, 4,096 cores total; 8 GB of memory per core totaling 32 TB; NumaLink5 Shared Memory Interconnect implementing two Single System Images with 16 TB of shared memory each. Storage includes local scratch, Albedo, and access to TeraGrid replication services. Pople will finish its production service on July 1, 2011.
- NICS increased the capacity of its Cray XT5 Kraken system by 144 TFLOPS with an additional 1,152 nodes in January. The 166 TFLOPS Cray XT4 system, Athena became available Oct. 1, 2010 as a co-allocated TG resource with Kraken.
- NCSA: The Lincoln Tesla-GPU system will be replaced in June by a 150TF (peak) Fermi-GPU system named Forge. The new system will have 32 nodes (Dell C6145 servers) with dual 8-core Magny-Cours processors and 48GB per node. Each node supports 8 C2070 Fermi GPUs. A 600TB GPFS file system provides an aggregate IO bandwidth of 15GB/s. (Final configuration may vary.)
- NCSA: The new NCSA SGI Altix UV, Ember, replaced the Itanium based SGI Altix Cobalt system with roughly twice the performance per core. System Info: 16 TFLOPS; 4 Intel SGI Altix UV shared-memory systems each with 384 2.6 GHz Nehalem-EX cores and 2 TB of memory; QDR IB.
- NICS: Nautilus, a TeraGrid XD system for data analysis and visualization, features SGI's UltraViolet large-scale shared-memory system. (See also TACC's Longhorn system.)
- SDSC: Dash, a precursor to the Gordon Track2 system, is available for evaluation of flash memory and vSMP software.
- Data Replication Service: TG-wide data management through iRODS, supports 1-100TB storage allocations.
Allocation Request Procedures and Policy Changes:
Please see Section 6.3.2 of the Allocations Policy document for the latest changes in document requirements.
A new, easy to use upload interface keeps track of each type of submission document uploaded. POPS now insures that PIs have uploaded all required documents before a final submission is accepted.
Section 6.4.1 of the Allocations Policy document explains the new priority policy for NSF/non-NSF funded projects for awards when resources are oversubscribed.
POPS now features auto-fill for renewal/supplement/justification/progress_report
requests. PI, co-PI, and Supporting Grant form entries are
automatically populated from the previous request. For renewals, please
make sure you update any changes in PI/co-PI biography, remove expired
grants, include new grant funding, and update values for the proportion of funding supporting your TG computational projects.
A guide for writing requests, and references to examples are available at the TG website. The TG Allocations Policy document explains the allocations process, procedures and policies.