Not Apples to Apples
IBM price is discounted... NetApp number are list price.
IBM is RAID1... NetApp is Raid-DP(RAID6).
NetApp and IBM have released SPC-1 benchmarks showing IBM’s 10 gig V7000 FCoE system being narrowly beaten by NetApp’s FAS8020 using 16gig FC. Let's rephrase that. A flash-assisted array using faster Fibre Channel beat an all-disk array using slower Ethernet. Quelle surprise. SPC-1 tests a storage array’s ability to perform …
Do not you think that latency in netapp much lower then IBM?
At all stages of load nearly 2 times netapp has better results. Latency is not related with ports bandwith and randum IO as well.
Netapp 192x450x10k disks, IBM 240x300x15k. During expluatation V7000 will eat more electricity throug the yers of expluatation because 15k disk eating much more then 10k. So it'll cost more in future.
Netapp $5.76/SPC-1 IOPS, IBM $5.94/SPC-1 IOPS. Total ASU Capacity, NetApp: 32,219.301 GB , IBM: 24,433.592 GB.
And what is the maximum of nodes for IBM, 2? NetApp - 8. What about file-based access in IBM, 4 nodes? Netapp have 24.
Ironic really IBM OEM'ing N-series and all these Netapp fanbois turning out to take the piss out of IBM's result. Never mind the fact Netapp has all the latest and greatest as well as the assistance of flash. Yet it still can't beat a 5 year old, all spinning disk result, from a now obsolete midrange array.
May I hint, that your SPC-1 IOPS chart is inaccurate !
Missing e.g. Fujitsu's ETERNUS DX200 S3 which delivered 200,500.95 SPC-1 IOPS. The dual-controller entry-class system offers an outstanding price performance level of US $0.77 / SPC-1 IOPS - at low discounts compared to other vendors.
The response time of 0.63 milliseconds at 100 percent load achieved the lowest response time at 100 percent load ever reported in the history of the SPC-1 Benchmark!
These SPC-1 results are current as of December 30, 2013 and can be found here: http://www.storageperformance.org/results/benchmark_results_spc1/#a00139