Thursday, 25 May 2006

    OK so today at London went well. Car journey was good, didn't get there too late. THe presentations were a bit to take in (bloody materials science and condensed matter physics).
    Lunch allowed me to chat to a Dr at the Royal Institute about water simualtions. Needless to say he didn't understand the magnitude of what we are doing.
    After that the farce began, the allocation of the hours for HPCx, a supercomputer we use. We had asked a modest amount an it got cut down. But still we know what we can do with those hours. Fucking loads.

    Car journey back was amusing, my supervisor has a weird and direct sense of humour. Work was discussed and  it seems like I'll inherit alot of research to oversee. Shit! I'm becoming important.

    So now for the weekend and an early start tomorrow to run some neural nets now that they are optimised.



    Above is a quantum topological iso surface of the electron density of the water dimer. Pretty eh!?

3 comments:

  1. What kind of specs does the HPCx have?

    ReplyDelete
  2. The HPCx system comprises 96 IBM POWER5 eServer nodes, i.e. 1536 processors, delivering 9.2 TeraFlop/s peak, or up to at least 6 TeraFlops/s sustained. The system is equipped with 3.2 TByte of memory and 36 TByte of disk

    ReplyDelete