I was talking to one our executives in the office today, and found out the company is in talks with a few Business Process Optimisation specialists. In another words - people who specialise in cutting jobs. * is actively looking elsewhere * 

Does anyone like Marmite ? My wife loves the stuff, but I think it's the next worst thing to  .... god I can't even bear to say it !

Bon Echo is the codename for FireFox 2.0. So far, I'm not liking it. I think its a biased point of view - since I can't use any of my current extensions to make the software useful.

I made the mistake of installing the installer version. I should have had gone with the zip install, it would have had kept my previous version.

Well - This is the second part in the saga of finding porn at work place. One enterprising programmer at work has been using a TOR client to bypass restrictions on content. I'm quite surprised how he managed to do that, since we though we were blocking it already. I'm investigating his computer for the modifications he made. ^.^ ;;; He got his computer infected with viruses + some reallly nasty spywares, and now I'm cleaning it up ( reimaging - takes about 15 minutes. ) 

I'm really surprised that people would download porn at work. I mean - not only can you get caught & fired, you get branded by it. You can't use your previous job as a reference in your CV. People are strange. I can understand needing porn. I can understand watching in the privacy of your home - god knows a teenaged boy's best friend used to be the family VCR - now for the affuent boy on the move, it's prolly their video ipod. But at work !!?!

So the question for you people : Have any of you done something at work, you weren't supposed to ? Ever think of taking your iporn to work ? Leave an anon message if you want.

I've got to get back to work now.

I'm in the unique position to do some benchmarks. This isn't something I've seen done in the mainstream media. The results I got - has me wickedsly stunned.

I used something called the Unix bench, a modified version of it that I got hold of. I've used this previous to catalog other hardware I've come across in the past.


Dual Opteron 246 with 2GB ram - SCSI 36GB

                     INDEX VALUES
TEST                                        BASELINE     RESULT      INDEX

Dhrystone 2 using register variables        376783.7  9335674.0      247.8
Double-Precision Whetstone                      83.1     1362.7      164.0
Execl Throughput                               188.3     3188.8      169.3
File Copy 1024 bufsize 2000 maxblocks         2672.0    52269.0      195.6
File Copy 256 bufsize 500 maxblocks           1077.0    15172.0      140.9
File Read 4096 bufsize 8000 maxblocks        15382.0   369142.0      240.0
Pipe-based Context Switching                 15448.6   342437.9      221.7
Pipe Throughput                             111814.6  1097560.9       98.2
Process Creation                               569.3    10620.1      186.5
Shell Scripts (8 concurrent)                    44.8      608.2      135.8
System Call Overhead                        114433.5  1780324.8      155.6
                                                                 =========
     FINAL SCORE                                                     172.0


Opteron 170 with 2GB RAM + SATA II 250GB HDD
                     INDEX VALUES
TEST                                        BASELINE     RESULT      INDEX

Dhrystone 2 using register variables        376783.7 10339492.8      274.4
Double-Precision Whetstone                      83.1     1397.7      168.2
Execl Throughput                               188.3     3855.9      204.8
File Copy 1024 bufsize 2000 maxblocks         2672.0    63358.0      237.1
File Copy 256 bufsize 500 maxblocks           1077.0    18679.0      173.4
File Read 4096 bufsize 8000 maxblocks        15382.0   503384.0      327.3
Pipe-based Context Switching                 15448.6   459182.5      297.2
Pipe Throughput                             111814.6  1761058.8      157.5
Process Creation                               569.3    12592.6      221.2
Shell Scripts (8 concurrent)                    44.8      763.0      170.3
System Call Overhead                        114433.5  2213971.1      193.5
                                                                 =========
     FINAL SCORE                                                     214.1



I expected the Dual Core to be less than Dual Processor. I mean, the Dual Processor server has the added benefit of being a scsi system. It's also heavily optimised with running only the kernel modules required to run the system. So I ran it again with the stock Kernel. The values were predictably lower. Before this benchmark , I was on the fence really about whether I would be advocating to my bosses for the next hardware renewal cycle we go for Dual Core servers. Note - these results are part of my personal experiment at work. The official stuff will rely on the Sandra benchmarks with Windows 2003 64bit version. I'm expecting the benchmarks to get higher in difference. In every case before, it's been better. However, there is a grey area in this. I'm not sure how Windows being NUMA aware will affect the results. I am not certain totally if the SMP kernels are NUMA aware.

* drooling .... *

Sarahkitty - any ideas on the Kernels of Unix variants ?

So What does this mean to you my non geeky friends . Well, if you didn't know it already, AMD and INTEL have released ( a little while ago ), chips that are Dual Core. They are basically TWO chips in one. They are wicked fast, and AMD is the fastest of the LOT. The cheaper one is the faster one. It also uses less electricty while running. You'll save about $100 from the electricty savings alone! If you are considering getting a processor - consider going to AMD X2 range. While are the desktop versions of the dual core chipsets I tested. Note : These won't improve your gameplay much. However, if you run a AntiVirus scanner, or any other application in the background, you'll definately see increased smoothness in anything you do.