Output Capacity

The second equation relates to output rate, $\mathcal{U}_{\rm out}$, capacity:

$$\left(\frac{ N^2_{\rm sta} \cdot N_{\rm sb} \cdot N_{\rm po... ...nt}} \; \simeq \; \mathcal{U}_{\rm out} \quad [\rm {kB/s}].$$ (2)

The first term in the left-hand side is the number of boards, $N_{\rm brd}$, required for the correlation (the full correlator comprises 32 boards). Our current peak $\mathcal{U}_{\rm out}\simeq$1500kB/s. An alternative way to think about the output-rate limit is to remember that the minimum $t_{\rm int}$ for a configuration using the whole correlator is now 1s; configurations that use no more than one-half or one-quarter of the correlator can achieve minimum $t_{\rm int}$ of $1/2$s and $1/4$s respectively. In the future, the Post-Correlator Integrator (PCI) will provide $\mathcal{U}_{\rm out}$ as high as 96MB/s, providing minimum $t_{\rm int}$ for the whole correlator of $1/64$s. However, if recirculation is being used, the minimum integration time would be $\mathcal{R}/64$s. The low integration times, together with the fine spectral resolution afforded by large $N_{\rm frq}$, will provide the possibility to map considerably wider fields of view through reduced bandwidth- and time-smearing effects in the u-v plane. Furthermore, besides increasing the absolute output data rate, the PCI will also provide additional computational capability prior to sending the correlated data to the receiving workstation, which could be used, among other things, to FFT the data into frequency-space or to apply phase-corrections to obtain multiple field-centers in a single correlation pass. This latter capability would appeal to people interested in specific targets within the primary beams of the participating stations, but who wouldn't necessarily want the volume of data ( $N_{\rm frq} \cdot N_{\rm int}$) associated with being able to map the entire area on the sky subtended by them.