Systems that relate to our environment have to be split in two categories. First there are current installations of running experiments. Two, BaBar at the Stanford Linear Accelerator Center (operated by Stanford University and Department of Energy) and CDF at Fermilab (operated by Department of Energy), both placed in the US. Although these experiments are the state-of-the-art, their capacity requirements are orders of magnitudes lower than the ones for the experiments of the Large Hadron Collider at CERN. Second, what concerns cluster technology we would like to mention the ASCI (Accelerated Strategic Computing Initiative) PathForward project, namely the supercomputer installation at Sandia National Laboratories (operated by Department of Energy). Several supercomputing platforms have been established by the ASCI, but access and information is highly classified due to the military applications for which the systems have been designed. We are only able to give a brief presentation of the option RED, the first TeraFlop computer in the ASCI project. The other project that we would like to present is The NCSA (National Center for Supercomputing Applications) NT cluster, based on commodity personal computers and Myrinet, a technology that is also investigated for CMS. It has to be pointed out, the all computing cluster projects aim at increasing the cumulative processing power, generally measured in TeraFlops nowadays. What concerns interconnectivity, there is a strong requirement to increase bandwidth and decrease latency. An ASCI assessment also states that available data communication technology is far behind the necessary level to build well balanced systems. It has however to be emphasised that the current installations take the components that are available. Research concentrates on reducing the software overhead of communication software only. It is accepted what can be achieved with a given setup. This is not the case for high-energy physics data acquisition. There are strict demands what concerns data dissemination capacities, on the other hand, computing power requirements will be fairly low compared to the technology that is available already today. CMS requires in the range of 5 TeraOps, a number that has been exceeded by a factor of at least 6 today.