DAQ Column up to the TDR

DAQ Column performance measurements

up to the TDR.

20-06-2002

The setup with the old XDAQ version runs stably, with a data transfer rate of about 10 MB/s over fast ethernet, limited by the network capacity. Migration to the new XDAQ in progress.

26-06-2002

Started tests with RCMS prototype.

27-06-2002

Preliminary tests to speed up TTCrx readout. By going to direct pointer access, this is no longer the limiting factor, but we are still at 20kHz only. Should be tested in a standalone test program.

28-06-2002

Waiting for XDAQ memory management problems (fedkit-XDAQ integration) to be solved.

08.08.2002

TTCrx disabled, always have triggers.

15.08.2002

Same conditions, some bugs fixed (block sizes), measured below and above sawtooth (4k).

Measurement with fedkit only, no XDAQ (Eric)

Dell compared to SuperMicro

21.08.2002

More bugs fixed (memory leak).
Control messages over Myrinet (GM).

Compare different event sizes [eps] [gif]
Compare SuperMicro with Dell in same setup [eps] [gif]

01.09.2002

Comparison of different block sizes on SuperMicro [eps] [pdf] [gif]
Block size of 5012 bytes [eps] [pdf]
Comparison of Dell and SuperMicro [eps] [pdf]
Simple model fitted to the measurement with SuperMicro [eps] [pdf]
Same model fitted to the measurement with Dell [eps] [pdf]
Using lognormal fragment size distribution [eps] [pdf]
Using lognormal fragment size distribution in the Dell [eps] [pdf]

11.10.2002

Some working plots collected over some time.

Gigabit Ethernet, one or two ports, MTU 1500 or 9000 [eps] [pdf]
Rewrote EVM matching routine. Comparison with previous version. [eps] [pdf]
Compare different operating modes. [eps] [pdf]
Compare bundling level [eps] [pdf]

15.10.2002

Started to do measurements for the TDR, summarised separately.