Quote:
|
Originally Posted by unregistered
HAL-9000
being not "prone to failure" is quite different from a MTBF of 50,000 hrs
the MCP600 is a good example, it will never have a MTBF of 50,000 hrs because there seem to be a few 'early failures' which cannot be eliminated, handled under warranty - not mission critical
- Swiftech incurred massive expenses related to solving the pump head problem, (which became a net loss when the pump mfgr decided to sell direct to our distributors)
Swiftech anticipates NONE of these kinds of problems because the MCP650 is a proven and validated design (and because we have a contract)
|
I was wondering why you guys ditched the MCP600, its a great pump. Sorry about getting shafted by the mfgr. But the point I was making is that I doubt the CSP750 is prone to being unreliable, and the cost advantage of buying two CSP750's vs. one MCP650. 50,000 hrs MTBF for the motor alone is impressive. Even if the seals are good for "only" 25000 hrs MTBF that's still almost three uninterrupted years of performance. Reliability starts to become measured by other criteria in my mind at this point, like how many times it can be cycled MTBF for instance.
I run a MCP600 pump, have no complaints whatsoever...love the pump. But I'm moving into an Nvidia Ultra 6800 next week for DoomIII reasons, and am going to watercool it. That CSP750 would be a great way to keep the overall flow in the circuit up to par even with the new addition, and with minimal impact and fuss.
That kind of scalability makes those things unique in the business, it definitely has a place in the market. Discreet scalable pumps
do have advantages. Jury's still out in my mind regarding overall performance as the sole solution in a rig; but I don't know if the MCP650 is either better performing or more reliable than a CSP750-X2, because that information really isn't available at this stage on account of both products' histories in watercooling being rather recent, and not tested enough by valid independent reviewers. That's my point.