Students taking an introductory computer science course work in the Department's Macintosh Lab in Olin 217. Since Mac OS X is a Unix operating system, students have access to both traditional Mac applications and a wide range of powerful Unix compilers and tools.
The lab contains a number of high-end Apple iMacs in the back and open tables in the front on which students can use either their own laptop or one from our laptop cart. A ceiling-mounted projector is connected to the front computer to allow faculty or students to give presentations to the class.
The room is used as a teaching area during the day and an open lab at night.
Computer science students use the Department's Linux Lab in Olin 219 to carry out most of their applied work beyond the first course. The lab contains 17 machines running Ubuntu Linux. In addition, a ceiling mounted projector connected to the front computer allows faculty or students to give presentations to the whole lab. The workstations consist of both single seat and collaboration stations, giving the room an open feel in which students feel comfortable gathering.
Olin 216 contains about ten computers with a mix of architectures and operating systems. This lab can be reconfigured to meet the needs of specific classes or students working on research projects during the semester or the summer. It is intended specifically for students majoring in Mathematics or Computer Science, who get first priority on the machines, study spaces, and storage lockers in the room.
The Denison math and computer science department hosts its own Beowulf cluster, an architecture of networked computers supporting parallel computation. The hardware resources of the cluster consist of 16 high-end workstations, each providing four CPU cores, eight gigabytes of memory, Gigabit per second network cards, and, new this year, NVIDIA graphics cards capable of supporting additional parallel computation on the General Purpose Graphics Processing Unit (GPGPU). Two additional workstations serve as primary and backup master nodes. All workstations are connected by a high-speed Gigabit per second networking switch.
The cluster includes multiple software architectures for parallel computation. These include MPI, a runtime and set of facilities for message passing between cooperating parallel tasks, Hadoop MapReduce for structured and fault tolerant parallel computation on large data sets, and the CUDA software for use in parallel processing on the GPGPUs.