Novell NetWare 6 performance tuning guidelines for ProLiant servers

figure 10. effects of hyper-threading on WebBench performance: quad processors

Performance Effects of Hyper-Threading (HT) on WebBench Test

(Higher is Better)

2,000

4,000

6,000

8,000

10,000

12,000

14,000

16,000

18,000

1 4 8 12 16 20 24 28 32 36 40 44 48 52 56 60

# of Clients

Requests Per Second

HT-ON

HT-OFF

Results: With a workload of 60 clients, the ProLiant ML570 G2 server configured with

four processors provided no statistical difference in performance on average with hyper-

threading disabled compared to when hyper-threading was enabled.

Note: It should be pointed out that we have been able to document modest performance

benefits when hyper-threading is enabled on 1P and 2P servers with several applications

running under Microsoft Windows 2000 Server and Linux operating systems.

Table 12 lists the applications and operating systems used in the test run in figures 7 - 10.

table 12. effects of hyper-threading technology used in different operating systems

Application W2K server Linux 7.x & up NetWare 6.0

SpecWeb99 Yes Yes No data

WebBench 4.1 Yes Yes No

NetBench 7.0.2 Yes Yes No

MS Exchange Yes No Data No Data

Yes - hyper-threading is beneficial or improves server performance

No - hyper-threading is not beneficial or degrades server performance

For more information about hyper-threading technology, visit

developer.intel.com/pressroom/archive/releases/20010828comp.htm