SAP on Nehalem-based systems from HP and Sun
Do you remember my blog post about the vast difference between Opteron 8-socket from HP and Sun? Well … history is repeating: There is an interesting difference between HP and Sun Nehalem system in the SAP SD benchmark, again.
According to the SAP SD benchmark certificate 2009005 the Sun Fire X4270 yields 20,300 SAPS at 3,700 users. The system was configured as following:
Sun Fire X4270, 2 processors / 8 cores / 16 threads,Intel Xeon Processor X5570, 2.93 GHz, 64 KB L1 cache and 256 KB L2 cache per core, 8 MB L3 cache per processor,48 GB main memory
The system used Solaris 10 as the OS and Oracle 10g as the database with SAP ERP 6.0, Enhancement Pack 4 (Unicode). The SAP SD benchmark certificate 2009004 discloses that a HP ProLiant DL380 G6 delivers 18030 SAPS at 3100 users.
HP ProLiant DL380 G6, 2 processors / 8 cores / 16 threads,Intel Xeon Processor X5570, 2.93 GHz, 64 KB L1 cache and 256 KB L2 cache per core, 8 MB L3 cache per processor, 48 GB main memory
The System used Windows Server 2008 Enterprise Edition as the OS and the RDBMS SQL Server 2008 with SAP ERP 6.0, Enhancement Pack 4 (Unicode). So the benchmark yielded 12,5 percent more performance for the Sun configuration than for the HP configuration. What´s a good conclusion out of this benchmarks? It´s still problematic to compare the results, as the OS and the RDBMS are different. The SAP SD benchmark isn´t really database bound. I don´t think that the differences between the both systems are that huge, so it´s perhaps a good assumption, that Windows vs. Solaris is the main point leading to this difference. Iit would be really interesting to see a Linux result of the HP system ;)