The server sizing comments are contrary to my experience in competetive POC situations.
Netezza simply size the server based on the prospects data storage requirements.
Teradata have design center guidelines to size their servers, but they will typically break their own guidelines and deliver an under-configured server to try and match Netezza on price. Of course, if they are successful the customer will very soon be faced with a large upgrade cost when they find the server doesn't deliver the performance they need !
Sorry, haven't popped in for a while. By price point, I simply mean the end selling price. If any of you have been a buyer you know that you have to dig through the "marketing" to get to the actual "apples" to "apples" configurations that determine the final purchase price.
NZ commonly proposes a system much larger than required for the data space needed so that it is cpu heavy and hence performs well in the benchmarks, If you make them size it for the data space you are requesting, performance isn't near as good as the new TD boxes, which, are smaller and cost less.
Don't know of any NZ replacements, but I do know of a few companies where NZ lost the business to the new TD "appliances" after benchmarks were run.
Conversely, I do not know of any TD "replacesments" by NZ, although I do know of a few that have both systems.
The Oracle appliance is a marketing scam...
It does not perform in the same league as NZ. As usual, the big bad software company with a huge marketing and product development budget "released" a product that is intended to freeze the market and not intended to deliver any immediate significant value.
Couldn't agree more...Oracle touts 10X performance increase...NZ gets 100X compared to that and you don't have all the administration that you are still gonna have with Oracle/HP solution!....when factoring in TCO over any span of time and the performance level NZ has now (not counting what they have coming down the pipe)Oracle will still have what they always have...smoke and mirrors marketing and positioning selling by and large inferiror performing products that have been cobbled together via aquisition...
Yes, NZ will come out on top and, no, I am not an employee.
Did anyone actually listen to what Larry said? He said the box has 64 Intel cores (probably 16 quad cores) on the database side, 112 cores (again 28 quads) on the storage side, etc.
Translation: same old Oracle software running on same old HP hardware, and oh, it sits in its own commodity rack now. It must be good!
Give me a break. First, notice that he clearly indicates that there is a "database server" and a "storage server." Hence, they are not one unit like the NZ solution. While the processors in an NZ box are PowerPC's, there a a LOT more than 64 of them for database processing, even on the low end. Aside from that, the NZ solution is custom built hardware running a custom FPGA that scans records *at each disk*... not in the "storage server." Notice Larry makes no reference of that.
Simply put, Oracle wont compete when it comes to raw performance. Try to scale the Oracle solution...or can you? Need more storage with NZ (last check was in excess of 100TB)? Simply add more racks.
This is the same old Oracle game...blow smoke to make it look like youre a competitor. Remember when Larry touted Oracle as the next big server virtualization player? Puh-lease! Larry heres a hint...anyone can rebrand Xen or even wrap Oracle software in a nice pretty HP wrapper. Doesnt mean that its anything new or worth considering.
Cant wait to see NZ spank Oracle in the benchmarks. But then again, the street will believe what they want to. So in terms of price, I have no comment.
But I do foresee some rethinking on the Oracle/HP side. When push comes to shove, perhaps Oracle will throw in the towel and simply do what they shouldve done a long time ago...swallow NZ.
Who knows? But apples to apples this is not.
I agree, Oracle DB is based on "Indexes" (for Key searches) I'm sure once NZ & Orcl speed and search benchmarks prove that Netezza is faster on searches for data.
Oracle DB is old by today's standards and the King can't claim he's the only Oracle of DB when the proof is shown, Once proved that Ocacle is 2nd in this segment of the market NZ will come out on top.
Anyone that knows anything about Oracle DBs know that they suck in terms of performance....HP tried for four years to pull off Neoview...NO ONE bought it...they gave it away to WalMart and I don't think they hardly use it...Idiot Executives that get wined and golfed up by Oracle sales people that demo well and talk well and dress well but don't know crap when it comes to data modeling, implementaion etc..Ah..you see Moron CFO it is the FUSION that will lead you to the promised land of analytics...what a load of crap...that crap is cobbled together with so many acquisitions the sales idiot doesn't know anymore than the bafoon executive eating it up...The best technology often loses to superior marketing and ignorance...But since Larry Ellison the great said that they are getting into the application business the "smart money" is getting out...Sad really...