Visualization Resources
To support the world-class research being performed on our high performance computing systems, TACC provides advanced visualization resources and consulting services, which are accessible both in-person and remotely. These resources include the state-of-the-art Visualization Laboratory in the ACES building; Spur, a system for large-scale remote interactive visualization; Colt, a small visualization cluster in the ROC building; and Stallion, one of the highest resolution tiled displays in the world. Click on any of the thumbnail images to see a larger image of the system in a new window.
The ACES Visualization Laboratory ("Vislab"), located in ACES 2.404a, has been completely transformed, and is now open to all UT faculty, students, and staff, as well as UT System and TeraGrid users. The lab includes updated systems and top-of-the-line equipment, such as Stallion, a Dell visualization cluster with 24 nodes and a 15x5 307 megapixel tiled display, and a Sony SRX-S105 9M pixel flat projection system driven by a high-end Dell workstation. A layout of the Vislab, along with brief descriptions of each of the different sections and equipment is given below.
Dell Visualization Cluster and 307 Megapixel Tiled Display (Stallion)
The Stallion cluster provides users with the ability to perform visualizations on a large 15x5 tiled display of Dell 30-inch flat panel monitors, for 307 megapixel resolution. This configuration allows for an exploration of visualizations at an extremely high level of detail and quality compared to a typical moderate pixel count projector. The cluster allows users to access to over 36GB of graphics memory, 108GB of system memory, 19TB aggregate local disk storage, and 100 processing cores. This configuration enables the processing of datasets of a massive scale, and the interactive visualization of substantial geometries. A large, shared file system is available to enable the storage of terascale size datasets. Detailed Stallion configuration follows.
- 2 Intel Xeon 5355 quad-core processors (8 cores total)
- 16GB of RAM
- nVidia 8800 GTX (G80) Graphics Card, 768MB memory
- 1.5TB local disk
23 Dell XPS Compute Nodes, each with
- Quad Core Q6600 Intel 2.4GHz CPU
- 500GB local disk
- 4GB Ram
- Dual nVidia 8800 GTX (G80) Graphics Cards, each with 768MB memory
75 Dell 3007WFP Widescreen LCD Monitors
- Max Res 2560x1600 per monitor (4.096 Mpixels)
Network Filesystem
- Dell PowerVault NFS500
- 6TB Disk Space (6x1TB disks)
Network Interface
- InfiniBand 4x & 1Gb Ethernet
- Native InfiniBand throughput 10Gb/s
Sony SRX-S105 (9M Pixel) Projection System (Bronco)
The flat screen area gives users a 20 ft. x 11 ft., 4096 x 2160 resolution display, which is driven by a Sony SRX-S105 overhead projector and a high-end Dell workstation. This configuration provides users with the added flexibility to run a wide variety of applications, as only one workstation is required to drive the display. The projector gives exceptional brightness and a high resolution, 9M pixel viewing area. In addition, it is configurable to accept inputs from up to four simultaneous video sources, allowing for a hybrid display of multiple systems. As a result, the setting is ideal for ultra-high-resolution visualizations and presentations. Bronco configuration details follow.
Dell Precision Workstation
- 2 Intel Xeon 5355 quad-core processors (8 cores total)
- 16GB of RAM
- 1.5TB local disk
- nVidia Quadro FX4600 (G80 core) and 768MB memory
- nVidia Quadro Plex 2200 D2 (4 DVI out)
Sony SRX-S110 Projector
- 4k x 2k = 4096x2160 or 4 1080P quadrants
- Total: 8.847 Mpixels
- Screen size 20'x11'
- Single continuous display
Dell Workstations (Horseshoe 1-4)
The workstation area consists for four high-end workstations, complete with large LCD displays. Each workstation provides a single user with the tools for visualization development. Users may develop their visualizations here, and then transfer to one of the two larger display systems.
Horseshoe 1 & 2: Dell Precision 690 Workstation
- Intel Xeon 5150 2.66 GHz dual-core processors (4 cores total)
- 16GB of RAM
- 200GB local disk
- Dual nVidia 8800 GTX (G80 core) Graphics Cards, 768MB memory each
- Windows XP x64 Operating System
Horseshoe 3 & 4: Dell Precision 690 Workstation
- Intel Xeon 5150 2.66 GHz dual-core processors (4 cores total)
- 16GB of RAM
- 200GB local disk
- Dual nVidia 8800 GTX (G80 core) Graphics Cards, 768MB memory each
- Linux Operating System (Kubuntu)
Collaboration Room (Saddle)
The collaboration room offers the opportunity for small groups to work together on developing and exploring visualizations. The display is provided by a single high resolution projector which is driven by a high-end Dell workstation. Users may develop their visualizations on the workstation, and then easily transition them to one of the two larger display systems in the main lab area at a later time. This is an excellent environment for developmental work and collaboration with the TACC visualization staff when large audiences are not present. Details of the Saddle configuration can be found below.
Dell Precision 690 Workstation with
- Intel Xeon 5150 2.66 GHz dual-core processors (4 cores)
- 16GB of RAM
- 200GB local disk
- Dual nVidia 8800 GTX (G80 core) Graphics Cards with768MB memory each
- Linux Operating System (Kubuntu)
The ROC Visualization Laboratory (ROC 1.524) is located at the Research Office Complex building of the Pickle Research Campus. The primary visualization resource is Colt, a Dell visualization cluster with 10 nodes and a 3x3 37M pixel tiled display. Each node contains two dual-core Intel Xeon 5150 processors clocked at 2.66 GHz, 16 GB of memory, and an NVIDIA GeForce 8800 GTX GPU with 768 MB dedicated video memory. All nodes are connected via InfiniBand network fabric, and share a 5TB file system.
The TACC DELL/NVIDIA Visualization & Data Analysis Cluster, Longhorn, is a hybrid CPU/GPU system designed for remote, interactive visualization and data analysis. In addition, Longhorn supports production, compute-intensive calculations on both the CPUs and GPUs via off-hour queues. The large, per-node memory is intended to support serial and parallel visualization and analysis applications that take advantage of large memories, multiple computing cores, and multiple graphics processors. Longhorn is an ideal companion resource for working with large data sets created on Ranger, since Longhorn can directly access Ranger's Lustre parallel file system through a 10 GigE network link.
The system consists of 256 dual-socket nodes, each with significant computing and graphics capability. Total system resources include 2048 compute cores (Nehalem quad-core), 512 GPUs (128 NVIDIA Quadro Plex S4s, each containing 4 NVIDIA FX 5800s), 13.5 TB of distributed memory and a 210 TB global file system. Longhorn configuration details can be found below.
128 NVIDIA Quadro Plex S4s, each with
- 4 NVIDIA FX 5800 GPUs
- 16GB Graphics Memory (4GB per GPU)
- 2 independent graphics busses, one per GPU pair
240 Dell R610 Compute Nodes, each with
- 2 Intel Nehalem quad-core processors (8 cores) @ 2.53 GHz
- 48GB RAM
- 73GB local disk
- connected to 2 dedicated NVIDIA FX 5800 GPUs via Quadro Plex graphics bus
16 Dell R710 Compute Nodes, each with
- 2 Intel Nehalem quad-core processors (8 cores) @ 2.53 GHz
- 144GB RAM
- 73GB local disk
- connected to 2 dedicated NVIDIA FX 5800 GPUs via Quadro Plex graphics bus
Mellanox QDR InfiniBand Interconnect
14 Dell PowerVault MD1000 Direct Attached Storage Arrays (210TB global file system, managed by the Lustre Parallel File System)
Longhorn is current running in a friendly user mode and is scheduled to enter full production on Monday, 4 January 2010.
Allocations on Longhorn are available through the National Science Foundation TeraGrid project: www.teragrid.org.
Spur, TACC's Terascale Sun Visualization Cluster, contains 128 compute cores, 1 TB aggregate memory and 32 GPUs. Spur shares the InfiniBand interconnect and Lustre Parallel file system of Ranger, TACC's Sun Constellation Linux Cluster. Thus, Spur acts not only as a powerful stand-alone visualization system: it also enables researchers to perform visualization tasks on Ranger-produced data without migrating to another file system and to integrate simulation and rendering tasks on a single network fabric.
Please see the Spur User Guide for additional information.
To facilitate the use of our visualization resources, TACC staff provide training on the various technologies, programming languages and software packages that will enable researchers to get the best results from their efforts. Please see our training calendar for more information. If you would like to be trained on the use of the systems in the ACES Vislab, please contact info@tacc.utexas.edu, or submit a ticket via the TACC Portal.
For access information, please visit the Allocations Page
