[Beowulf] Network considerations for new generation cheap beowulfcluster

[Jess Cannata]

Mark Hahn wrote:
>>> 5 years ago, the low-end approach was 100bT; now its 1000bT.  the 
>>> prime target for that approach (serial or EP) has simply gotten 
>>> broader;
>>> I don't see this as anything to complain about.  for "real" parallel,
>>> you have to pay for the network you need.  there as well, you now 
>>> get more for your money, no complaints.  complaining that you can't 
>>> get 1 us, 1GBps interconnect for $50/port is just silliness.
>>>
>> I disagree on this last point. Why can't low latency interconnects 
>> become the standard?
>
> because most of computing is not latency-sensitive.  even in HPC, most 
> cycles
> are consumed by throughput and loose-parallel apps (where Gb works fine.)
>
>> It is not just HPC applications that are demanding low latency networks.
>
> what applications are you thinking of?  the only I can think of would 
> be lock/metadata traffic for very large parallel clustered DB/filesystem,
> and even there the argument is weak.
>
I'm thinking of applications like on-line video games (first-person 
shooters, real-time strategy, massively multi-player role-playing, etc. 
benefit from lower latencies) and automatic electronic stock trading to 
name a few.