Posted by: Anil | April 20, 2007

UnixBench 4.1.0 GCC vs Entic’s compiler optimization

On one of our v.Dedicated servers, we compared the benchmarks generated with UnixBench when compiled with the default GNU GCC (which is what almost 99% of all hosting/server providers use) vs Entic optimized Solaris native compiler. No tuning was done to UnixBench itself.

Here are the results for GCC run:

                     INDEX VALUES
TEST                                        BASELINE     RESULT      INDEX

Dhrystone 2 using register variables        116700.0  4455048.7      381.8
Double-Precision Whetstone                      55.0      887.5      161.4
Execl Throughput                                43.0      626.2      145.6
File Copy 1024 bufsize 2000 maxblocks         3960.0    49992.0      126.2
File Copy 256 bufsize 500 maxblocks           1655.0    12987.0       78.5
File Copy 4096 bufsize 8000 maxblocks         5800.0   158381.0      273.1
Pipe Throughput                              12440.0   323922.0      260.4
Pipe-based Context Switching                  4000.0    39441.5       98.6
Process Creation                               126.0     1144.9       90.9
Shell Scripts (8 concurrent)                     6.0      164.0      273.3
System Call Overhead                         15000.0   264727.8      176.5
                                                                 =========
     FINAL SCORE                                                     166.5

and (drum rolls please…) for Entic’s optimized results:

                     INDEX VALUES
TEST                                        BASELINE     RESULT      INDEX

Dhrystone 2 using register variables        116700.0  7853579.1      673.0
Double-Precision Whetstone                      55.0     1621.8      294.9
Execl Throughput                                43.0      737.1      171.4
File Copy 1024 bufsize 2000 maxblocks         3960.0    80697.0      203.8
File Copy 256 bufsize 500 maxblocks           1655.0    32700.0      197.6
File Copy 4096 bufsize 8000 maxblocks         5800.0   235523.0      406.1
Pipe Throughput                              12440.0   282546.0      227.1
Pipe-based Context Switching                  4000.0    37091.1       92.7
Process Creation                               126.0     1210.4       96.1
Shell Scripts (8 concurrent)                     6.0      165.0      275.0
System Call Overhead                         15000.0   162554.1      108.4
                                                                 =========
     FINAL SCORE                                                     209.2

Thats more than 20% performance improvement!

The servers you purchase from Entic are all optimized to run the fastest. You’ll be sure to get the fastest performance possible for your web serving.

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Categories

%d bloggers like this: