Mac Pro or xServe

[SourceChild]

In an xServer, a user cannot have more tha three PCI cards.
If one is a GPU and one is a Capture Card, there is no slot left for a Fibre Channel card.
In an xServe however, the drives are SaS 15,000 RPM capable with a controller card as effective as the Fibre channel controller on an xServe RAID.
Therefore, if I as a user were to configure an xServe with a System drive and then Stripe two SaS drives to create my media drive, wouldn't I be able to match the performance of a Mac Pro hooked to an xServer RAID.

I have been giving this thought but I was curious to other people's results with this. I can't loose Video input through a capture card but I also cannot give up a fast GPU. I would although prefer a small single rack with xServes in it than a bunch of racks with bulky Mac Pros in them.

Pound for Pound, do the xServes match up to the Mac Pros? Are the xServes more powerful or reliable?

What are the results?

[SourceChild]

I am sill very interested in Feedback on this. I am fixing to propose a $150,000 equipment package but since I have not yet prototyped the new xServes running Catalyst, I am interested in feedback.

[emilianomorgia]

I am sill very interested in Feedback on this. I am fixing to propose a $150,000 equipment package but since I have not yet prototyped the new xServes running Catalyst, I am interested in feedback.

I've no experiance on Xserve , but if I ware you , in case of this big dealing of 150.000 bucks I'll go to an apple reselers talk to him about the deal and ask him to give you an xserve , an macpro for testings . Me personally for 150.000 I'll ask the apple reseller to let me meet the office secretary ( if she is cut ) http://www.sinapsi.us/forum/images/upo/emoticon-0109-tongueout.gif

E

[SourceChild]

Yep! And since you're 6 hours into the future (since we established that on NYE) I'll let you go to the secretary first so that by the time I get there, she will be done with her shower.
Then we can tell her boss she has a thing for Catalyst Guys.:cool:

The big thing I am really curious about is doing a stripe set across two of the 15,000rpm SaS disks to see if this solution out performs or comparably performs with the Fibre from the xServe RAID.

I would trust benchmarks but I have proven enough wrong that I don't trust anymore unless I hear from someone who has tested in the same application I intend to use.

Peace!