Mark 5A
1. Conduant believes it has fixed the problem of write pointers going 
spontaneously to "0".
2. Due to press of Mark 5B, we have not yet addressed the FS<->Mk5 
communications problem, but it is on our list.
--
Mark 5B
1. Mark 5B DIM and DOM have both been pretty thoroughly checked, though 
software support is still developing.
2. Assembled Mark 5B I/O boards for distribution to Mk5 consortium members 
are expected at Haystack next week; checkout will take several weeks, but 
will undoubtedly be slowed by holidays.  Hope to be have them completed 
sometime early January.  Mark 5B I/O boards will be available for purchase 
from Conduant.
3. Using new Amazon StreamStor board from Conduant, we achieved sustained 
recording speed of 2048Mbps onto 8 SATA disks.  Primary purpose of this 
test was to verify FPDP2 interface between Mark5B I/O card and the Amazon 
card. A PATA version of the Amazon board is being developed by Conduant; 
the 2Gbps version of Mk5B is being designated Mark5B+.  Software support 
for the Amazon boards is still primitive and will likely not be mature for 
several months.  Mk5B+ system will play back only at 1Gbps.  Correlation of 
2Gbps data on Mark 4 correlator will be at 1Gbps.
4. Correlator software support for Mark5B and Mark5B+ is underway, but will 
take several months to complete.
5. Phase-cal extraction system has been tested in both DIM and DOM; 16 
tones extracted from each of 16 channels.
-
e-VLBI
At last week's SC05 meeting in Seattle, we demonstrated real-time 
correlator at 512Mbps from Westford, Onsala and GGAO.  Data flows were 
solid for hours on end.  We were not successful in adding Kashima, which is 
complicated by need for on-the-fly data translation from K5 to Mark5A 
format.  Data flowed was sustained at 512 Mbps, but correlator could not 
recognize it, so probably problem with format; we are 
investigating.  Kashima transfer is first actual demonstration of 
VSI-E.  We were unable to achieve sustained 512Mbps flows from Westerbork 
or Jodrell.