back to article Server makers wrap iron around Nvidia GPUs

Supercomputer buyers don't want to spend months building hybrid CPU-GPU clusters. They want to buy them pre-integrated and ready to start flopping within a matter In the wake of the announcement of the Nvidia Tesla M2090 GPU coprocessor for servers two weeks ago, Marc Hamilton, vice president of high performance computing at …

COMMENTS

This topic is closed for new posts.
  1. Anonymous Coward
    Devil

    I love armchair strategists...

    Quote: By switching to this bigger tray server and by putting in four SL6500 chassis, yields 31.9 teraflops of GPU performance plus the 1.18 teraflops from eight server nodes for a total of 33.1 teraflops of oomph. It is hard to say what HP might charge for this.

    True, as it will have to put in a joint quote with General Atomics for the nuclear powerplant in your backyard.

    The pre-integrated "ready-to-run" HPC market is limited by datacenter power. It cannot exceed (at least by much) the power consumption expectations for an average datacenter rack lineup. It is a drop-in replacement for normal kit which joe average corporate customer who has decided to dabble in HPC can buy, put and run.

    Same goes for cooling, HP actually did the right choice here. Doing more would have resulted in a product with a much smaller potential market.

This topic is closed for new posts.

Other stories you might like