There are currently three services offered by the IDRE Research Technology Group.
Cost – The cost for one terabyte of HPC storage for a period of one year is currently $259.
This rate includes the physical storage space, infrastructure to support it, administration of the users of your storage space, hardware and software upgrades and problem fixes. It does not include backups. See below. This rate will remain in effect until renewed through the campus. The current rate was approved in July 2017. Note that if you are using federal funds these charges will incur overhead (currently 54% as of March, 2014) as they are considered a service and not equipment, which would be exempt from overhead. See: http://www.research.ucla.edu/ocga/sr2/idcinfo.htm#FA1 for current rates.
Backups – No backups will be done on your data unless you specifically ask for this service. Although we have made every effort possible to ensure reliability and fault-tolerance of our storage systems there is no guarantee that you will not lose your data.
Cost – Backup of 1TB of storage for one year is an additional $148.85 for a total of $407.85 a terabyte per year.
To purchase HPC storage, please go to https://rim.idre.ucla.edu/secure/storage.php
For more information on this service see: https://idre.ucla.edu/resources/shared-storage/hpc-storage
Cloud Archival Storage Service (CASS)
- Storage is sold in units of a terabyte (TB*) for a period of one to five years.
- Rates listed below are per terabyte, per year.
- The funding source code determines which rate table to use.
- Pricing is per transaction. This means additional storage purchases are not cumulative with respect to volume or time discounts applied to previous purchases.
- Storage is currently available only to University of California (UC) entities.
- Rates are current as of November 1, 2013 and are subject to change without notice.
- Note that if you are using federal funds these charges will incur overhead (currently 54%) as they are considered a service and not equipment, which would be exempt from overhead. See: http://www.research.ucla.edu/ocga/sr2/idcinfo.htm#FA1 for current rates.
To determine the cost per terabyte, first find the row containing the total amount of storage you require. Then follow across to the column corresponding to the number of years required. Multiply that rate times the total number of TBs and the number of years to determine the total transaction cost.
|1 Year||2 Years||3 Years||4 Years||5 Years|
|1 – 7 TB*||$156.74||$141.06||$134.79||$128.53||$125.39|
|8 – 15 TB||$137.93||$133.23||$128.53||$125.39||$122.26|
|16 – 31 TB||$131.66||$128.53||$124.61||$122.26||$120.69|
|32 – 63 TB||$128.53||$125.39||$121.47||$120.37||$119.12|
* 1TB = 1,000,000,000,000 bytes
For more information on CASS see: https://idre.ucla.edu/cass
The cost of one HPC Compute Node is approximately $6,150 as of January 2018. This node features 2 x 12-core Intel Xeon E5-2650v4 CPUs, running at 2.2GHz, with 64GB* of memory, local SSD storage, and 5 years of support from Silicon Mechanics. Included is full access to these cores through the Hoffman2 Shared Cluster queuing system, administration of the group and users of your compute resources with complete support of the hardware and the operating system as well as connection to the cluster’s Infiniband fabric.
If you are interested in adding additional memory, more cores, larger hard drives/SSDs, or GPUs to your compute nodes please let us know.
* Note that the previous batch of nodes contained 128GB of memory for roughly the same cost. We have dropped back to 64GB because memory prices rose sharply on 2017 and a node with 128GB of memory would run an additional $800. Rather than increase the node cost that drastically, we opted to keep the cost close to what it was and let those that specifically need the extra memory pay the premium. Nodes can easily be upgraded if you need extra memory. We will continue to monitor memory prices and make adjustments to node prices to provide the best configuration we can while keeping the cost reasonable.
For questions or to order compute nodes please contact Lisa Snyder firstname.lastname@example.org
Information and usage policy for the Hoffman2 Shared Cluster is available here: https://idre.ucla.edu/resources/hpc/hoffman2-cluster