Differences
This shows you the differences between two versions of the page.
|
projects:meerkat:compest [2016/12/13 12:10] 134.104.31.74 |
projects:meerkat:compest [2016/12/13 12:17] (current) 134.104.31.74 |
||
|---|---|---|---|
| Line 26: | Line 26: | ||
| Summing up the stations now "only" takes 4*N_a*N_b*bw floating-point operations per second (factor 4 is for two polarisations and complex beams). This is typically 90 TFLOPS, but these can be combined with the phase corrections using fused multiply/accumulate operations. | Summing up the stations now "only" takes 4*N_a*N_b*bw floating-point operations per second (factor 4 is for two polarisations and complex beams). This is typically 90 TFLOPS, but these can be combined with the phase corrections using fused multiply/accumulate operations. | ||
| - | |||
| - | Each complex multiply requires 4 floating point multiplies and 2 floating point adds. The actual sum of the antennas requires (N_a - 1) floating point adds. So we have (7 * N_a - 1) * N_b * N_p * bw. For 40 antennas and 400 beams, this is 191 TFLOPS. This goes up to 306 TFLOPS for the full array. Here we have excluded all other operations, including input type casting, detection and time aggregation. | ||
| - | |||
| ==== Incoherent sums ==== | ==== Incoherent sums ==== | ||
