Which is faster? P4 3ghzHT (s478) or Athlon 64 3500+ (OC to 2.3, s939)?

[TurtleCrusher]

@nenforcer: I couldn't even begin to explain how big the difference from 1.8 to 3.2 is. A 1ghz P3 is about the same as a 1.1-1.2ghz P4? Either way, I expect it to be about 6-7 times faster in intensive situations. the clock speed alone makes a immense change. The Hyperthreading will feel like a 50% gain in speed and the fast (for the time) RDRam should make a big difference.

@OBLAMA2009: I don't have a mobo that fits an i3, I don't have money for a DDR3 kit at the moment. The whole platform swap would cost around $330-$350? That's too much for an entry level security supervisor to spare :whiste:.The P4 was free. I have a few DDR kits and some AGP cards. She only does facebook and flash games. The Athlon 3500+ (Oc to 3600-3700+ speeds) should be fine. I was just wondering about the impact of Hyperthreading on daily usage these days.

With the P4 in my houses server, routing throughout the house is FAST. When someone was using the computer it would be unbearably slow. Now we cant even tell when someone else is on the server PC. Thanks to hyperthreading. I cant see my mom doing more than 1 task on the computer at a time, so a 3700+ is certainly fast enough for her needs.

 

[Bill Brasky]

This thread has been so much fun to read. I had an awesome time. :awe:

 

[Cogman]

windows is a piece of crap and the only reason it got anywhere was be being a better run bussines. not a better OS. I've written and I am writing plenty of code. Windows sucks.

Umm, yeah, as you said in another thread, C escapes you... lol.

The windows API is great to work with, and the MSDN has some of the best documentation around, hands down. I've had to work with the X (free and org) library, just to create a simple window is a nightmare. Then you get into the GTK or QT which are just about as good as using the native windows API. (They are a little better as they provide more "features").

Its like .net. Why ? What was so wrong with vb6 ? Wanna cut the performance of your applications but 50% on the same hardware ?? compile in .net.

Dude, seriously? You are trying to advocate vb6 over .net. There is so much fail there I don't know where to begin. As for the speed thing. MOST applications in the business world don't need to be running at uber speed. Computer are fast enough (Oh look, my program displays a button faster than your program!). And for the situations where it isn't, software developers are generally smart enough to go "Oh, I should be using C/C++ or even possibly assembly here". .Net does its job well.

Its that simple.

windows sells hardware with code bloat.

What? I find my windows mouse drivers nice and compact... (Microsoft doesn't sell hardware). Your code bloat isn't from the microsoft driver setup. It is from all the extra stuff that hardware manufactures feel you need with their hardware, along with drivers for every piece of hardware that company has ever made. Legacy sucks, that is why broadband was invented.

I was able to vastly improve my windows 7 install by killing off needless redundant process's and streamling the binary of the kernel in a few places. thanx god I know assembly.

BWHAHAHA Yeahhh... RIIIIIIIGHT. Tell me, where did you tweak the kernel with your masterful assembly skills? In fact, I'm very interested to see what lines you changed and how/why you changed them.

sloppy sloppy code. Its obiously so to becuase the complier optimizations are non existnatn and if you have looked at enough assembly and binary you can see it in the code.

Umm, yeah, no, This coming from a guy that says "C escapes him" I'm sorry, but I just don't accept the fact that you know what clean and sloppy code look like. Couple that with the fact that Compiler optimizations often produce crappy looking assembly, I find it pretty hard to believe that you would be able to look at a piece of assembly and divine how clean the code that compiled into that assembly looks.


I'm serious though, post what you have changed in the kernel. I'm really interested to see what sort of "optimizations" you have made.

 

[evolucion8]

@nenforcer: I couldn't even begin to explain how big the difference from 1.8 to 3.2 is. A 1ghz P3 is about the same as a 1.1-1.2ghz P4? Either way, I expect it to be about 6-7 times faster in intensive situations. the clock speed alone makes a immense change. The Hyperthreading will feel like a 50% gain in speed and the fast (for the time) RDRam should make a big difference.

No way, I had a Pentium 3 1.13GHz Tualatin and I compared it against my brother's in law Pentium 4 at 1.60GHz and they traded blows, but the Pentium 3 was slightly faster in general code like Office, web browsing while the Pentium 4 was faster with stuff like Flash, media encoding and FPU stuff. Pentium 4 had a terrible IPC deficit compared to the Pentium 3 which made it up to with higher clockspeeds, HyperThreading only brought between 5 and 13% performance boost in the best case scenario due to the very narrow pipeline execution of the Pentium 4, the Core i7 shows a 50% boost in the best case scenario (Which is rarely true), and RDRAM has big problems with latency due to its narrow bus width and high clocks. SDRAM still a better solution.

@OBLAMA2009: I don't have a mobo that fits an i3, I don't have money for a DDR3 kit at the moment. The whole platform swap would cost around $330-$350? That's too much for an entry level security supervisor to spare :whiste:.The P4 was free. I have a few DDR kits and some AGP cards. She only does facebook and flash games. The Athlon 3500+ (Oc to 3600-3700+ speeds) should be fine. I was just wondering about the impact of Hyperthreading on daily usage these days.

Certainly, thanks to HT, the Pentium 4 had aged better than expected when compared to those old times having a rage competition with the Athlon 64, I still having a Pentium 4 3.40GHz Northwood and that's the only processor that can go toe to toe with the Athlon 3500+, so having a slower 3.0GHz Pentium 4 will be inevitably slower than the Athlon (Not by a big difference though), but I think that with the proper multi threaded optimization taking advantage of hyper threading, the Pentium 4 3.0GHz should be able to at least match the Athlon 64 3500+ which has a higher IPC.

And about ModestGamer posts;

Didn't knew that Windows was written in assembly? :awe: and AFAIK, assembly language gains little or nothing by using compiler optimizations

 

[evolucion8]

DP