PC Pitstop Optimize

I had it apply the suggested changes, rebooted and went back to SG Speed Test. I used the server in Kuopio, Finland.

Before making the changes, ~5500 kbps down and ~1100 kbps up.

After making the changes, ~14000 kbps down and ~8500 kbps up.

In round figures 2.5 times better down and 7.5 times better up. The system is now at ~15%/~85% of the theoretical speed down/up.

Maybe I spoke too soon but at 15% of the theoretical speed down, it is still more of a test of "the global environment ", I think.

Christer

 

Hi Christer

. Yes, global in the sense of the world

Try different countries/servers - for example I got .....

Southport, UK 552/217 - down/up
San Diego, Ca - 1300/353
Speedguide - 1959/356

Connected at 3.7 Mbps

 

I really should be working in the garden while the sun is out - a wet Holiday weekend is forcast

Anyway I couldn't resist tweaking - quite amazing results if they are to be believed .....

Southport, UK - 552 > 2007 down, 217 > 363 up

San Diego, Ca - 1300 > 1196 down , 353 > 353 up

Speedguide, Fl - 1959 > 2705 down, 356 > 342 up

So a worthwhile improvement, even after taking into account global traffic, etc.

I'll stick with these settings and see if there is a difference in the real world

 

Pete,
I too think that the improvement is amazing but I'm still no way near my maximum down.

The speed slider does not go higher than 20000 kbps which is 1/5 of my maximum. I tried increasing the TCP Receive Window from the "optimal" 1027840 by first doubling and then doubling again (which would represent a speed of 40000 kbps and 80000 kbps respectively) but I noticed no difference. I don't know if my logic is wrong but the TCP/IP Analyzer told me that the value was excessive if my connection is lower than 20000 kbps but it isn't.

I have, prior to starting fiddling with it, seen speeds of approximately 50 Mbps and occasionally higher from Microsoft and Symantec servers but maybe latency is a key factor. TCP/IP Analyzer gives the limitations based on 200 ms and 500 ms latency. Maybe the mentioned servers are much lower in latency. Under all circumstances will I stick with the suggested "optimal" settings and see how it works.

Another question is how big the improvement will be on slower connections, say 1 Mbps and 10 Mbps? Comparing my up speed (limited to 10 Mbps), now running at ~85% of the theoretical maximum at the current traffic situation, to my down speed (full 100 Mbps) which is running at ~15%, indicates to me that users on slower connections will benefit the most.

Another approach is that paying big bucks for a 100 MBps connection is a waste of most of the bucks compared to a 10 MBps connection.

I tested different servers as well. No other comes close to the one in Kuopio, Finland. Some of the others in Europe are in the 8-10 MBps range, both down and up. The servers on the other side of the pond display a great variation, from 1/10 to 1/5 of the speed of "Kuopio ".

Christer

 

I compared the changes applied by SG TCP Optimizer with the changes suggested by PC Pitstop Optimize and they differ a bit for some of the settings. SG TCP Optimizer did not suggest to use HTTP 1.1 on proxy but I tried it and noticed no difference (maybe because I don't know what I'm doing ... ... )

Christer

 

Well I guess that makes two of us

A lot will depend on the speed of the service to which you are subscribed - I am on an 8 Mbps line, but have yet to see more than 4.2 Mbps - distance from exchange, quality of line, etc.

As I said I will leave things as they are - have a backup of the original settings - and continue with the garden (have mowed the lawn since last posting )

 

Dave,
the only thing I got at Speakeasy was a notification that I need Flash 7 or later (got Flash 9). The "Link to Speakeasy Speed Test" took me to a page for downloading buttons and banners to my own computer.

Christer

 

Hi Christer

Just click on a server and the test should run!

 

Pete,
I don't see any servers anywhere. Please, see attachment for what I get.

Christer

 

Pete,
I disabled Nortons Ad- and Popup-blocking and the servers appeared. (When will I learn to read all the "fine print" ... ... ?)

Christer

 

You have a Flash problem ? - this is what I see ....

 

Our posts crossed

 

Well, the results were not better than from any other server in the US. 1500-4900 down and 1700-3000 up. In 50% of the cases, up was faster than down.

They try to sell "full-duplex 1.5 Mbps data transmission" at $299 to me by telling me that my 100/10 Mbps connection stinks ... ... for which I pay approximately $20 a month! I guess that's why the ad-blocker had to be turned off.

Christer

 

I was browsing my "most frequently visited sites" and came to remember one that has a lot of small *.html to load, Warbird Registry. After tweaking ... ... it loads much faster.

Christer

 

Interesting site - I was well into aircraft as a kid - loaded in a couple of seconds I guess. Quick, anyway

 

It was (almost) possible to count them when they loaded before but not now. It is not useful to try to repeat the test unless you have the TIF cache setting very low or clear the cache manually.

Christer

 

I've experienced a lot of stalling, red crosses and "Page can not be dsiplayed" during the last few days. I don't know why and if it is due to the changes.

Anyone else?

Christer