
IBM announced many more Unix server models this year Oracle also announced new Unix servers. Therefore, you'll end up saving money with IBM servers. And the cost of the IBM server hardware is similar to x86 servers where the IBM server would cost more is for using AIX, however because IBM servers pack a lot more SAPS than x86 you'll end up needing more x86 servers to run an entire SAP environment than with IBM. So, an IBM Unix server with the same type of configuration than x86 provides 50% better performance. If you are purchasing the SAP on VMWare i will also do the setup on your PC.
#For sap installation what is the requirement windows#
SAP ECC6is installed on Windows Server 2008 R2 & Windows Server 2012 R2. When you pay, i will send the link on your gmail account. Go directly to the SAP benchmark results database and search for SAP certification numbers 2013001 for Cisco, 2012035 for the IBM Unix server and 2012031 for NEC. SAP on VMWare is already uploaded to Google Drive. The best Windows result on the same Intel hardware is for a NEC server that generates 36,770 SAPS or 2298 SAPS/core. According to the SAP Fiori apps that you want to use, different components must be available on the front-end server and on the back-end server. It provides 35,680 SAPS which results in 2230 SAPS/core. To enable SAP Fiori, ensure that the relevant back-end and front-end components are available in your system landscape. For example, the best Linux/x86 result is with a Cisco UCS B200 M3 with the latest Intel E5-2690 2.9GHz processor. x86 also supports more SAPS per core than what you say but much less than the IBM server can. That's nearly 2 1/2 times more than what you say Unix can support. That's 3419 SAPS per core (3418.75 to be exact). For example, since you are using 2-socket/16-core servers as your base comparison, I find that a 2-Socket/16-core IBM Power Unix server (which came to market in 2012)running the latest SAP ECC release with Enhancement Package 5 generates 54,700 SAPS. You also say that there are no new Unix servers, which is not accurate either. Rajib, Sorry to say but your statement that Unix and x86 have the same SAPS per core performance is simply not accurate based on certified SAP benchmarks.
