System Requirements

The tables below specify system specification for different platforms. The requirements may depend on the application so these should only be taken as guidelines and not as absolute numbers. A complex clipping solution might require much more power than a simple collection solution. The recommendations for servers are for one server. The number of servers used for a given application (the size of a cluster) is a completely different matter and should be estimated using methods described elsewhere.

IDE requirements:

	Minimum	Recommended
Windows	Intel Core Duo 1.8 GHz CPU (or AMD equivalent), 2GB RAM, 260MB Free Disk Space	Intel Core Duo 2.66 GHz CPU (or AMD equivalent), 4GB RAM, 260MB Free Disk Space
Linux	Intel Core Duo 1.8 GHz CPU (or AMD equivalent), 2GB RAM, 260MB Free Disk Space	Intel Core Duo 2.66 GHz CPU (or AMD equivalent), 4GB RAM, 260MB Free Disk Space

Server requirements:

	Minimum	Recommended
Windows (32-bit)	Intel Xeon L5520 CPU (or equivalent AMD Opteron), 1GB memory per CPU core + 512MB for OS, 500MB Free Disk Space	Intel Xeon X5680/X5677 CPU (or equivalent AMD Opteron), 1.5GB memory per CPU core + 1GB for OS, 500MB Free Disk Space
Linux (32bit)	Intel Xeon L5520 CPU (or equivalent AMD Opteron), 1GB memory per CPU core + 512MB for OS, 500MB Free Disk Space	Intel Xeon X5680/X5677 CPU (or equivalent AMD Opteron), 1.5GB memory per CPU core + 1GB for OS, 500MB Free Disk Space
Windows (64-bit)	Intel Xeon L5520 CPU (or equivalent AMD Opteron), 1.5GB memory per CPU core + 1GB for OS, 500MB Free Disk Space	Intel Xeon X5680/X5677 CPU (or equivalent AMD Opteron), 2GB memory per CPU core + 1.5GB for OS, 500MB Free Disk Space
Linux (64-bit)	Intel Xeon L5520 CPU (or equivalent AMD Opteron), 1.5GB memory per CPU core + 1GB for OS, 500MB Free Disk Space	Intel Xeon X5680/X5677 CPU (or equivalent AMD Opteron), 2GB memory per CPU core + 1.5GB for OS, 500MB Free Disk Space
Solaris	sun4u Netra T1 200 (UltraSPARC-IIe 500MHz) 512MB memory per CPU core + 512MB for OS, 500MB HD

Real-time data: If you have a solution where users are waiting for results in real-time, CPU speed is normally the bottleneck, and you should buy the fastest CPU available for your hardware platform.

Dedicated hardware: For best performance we recommend that you always run RoboServer on dedicated hardware. That means that you should not run database servers and other services on the same hardware as your RoboServers. If you have a 8 core machine, and only a 2 core RoboServer license, you should use virtualization (like VMWare), or assign CPU affinity to the RoboServer process (and other processes) to insure that they don't "steal" each other's CPU power.

Operating system: Kapow recommends using the Linux operating system as it is currently faster than Windows when running Java applications. In some scenarios we have measured up to 33% more throughput when using Linux.