CPU suggestions

spudz

Distinguished
Jul 21, 2003
5
0
18,510
I've been wanting to build myself a new system for a while. I had all my parts picked out, and then I got informed of the Pentium 4 Cs. I was originally thinking of getting an Athlon XP 2800+ Barton, but I don't know now.

My question is, which would I be better off getting?

a) XP Thoroughbred 2700+ (~$170 CDN)
b) XP Barton 2800+ (~$250)
c) Pentium 4 2.4C (~$240)

Or something else entirely? The main things I'll be doing on the computer are office work and gaming, but mainly gaming.

The core of the system I had originally planned on was:

Athlon XP 2800+ Barton
A7N8X Deluxe
512MB Dual channel OCZ PC3200
Radeon 9800 Pro 128MB

I've never worked with Pentiums. It's not that I've had anything against them, it's just always been a matter of pricing. Since I was thinking of getting an A7N8X Deluxe to go with the Athlon, what would basically be the equivalent for a P4 mobo? From what I've seen, the C series CPUs pretty much blow everything else away, but I'm not sure on the Pentium motherboards. My budget is limited to about $450 (CDN) for the CPU + mobo. I should also state that I don't plan on doing any overclocking until warranties run out.
 
I would also suggest going with the 2.4Ghz, 800Mhz FSB Pentium 4. It has HT and will not be slower than the 2800+, despite the XP's rating. Plus, a good i865 mobo will support prescott, the next great thing from Intel for desktops.
From what I've seen, the C series CPUs pretty much blow everything else away
Agreed. :smile: The nForce2 equivalent for Intel would be the i865 chipset. It has all features you can possibly think of, good performance, and is very adequately priced.

But hey, the 2800+ should also be one hell of a machine, and is an excellent second choice.

<font color=red><b>M</b></font color=red>ephistopheles
 
HT - just a marketing strategy by intel, though it's that very effective at all except when the program is written for it.

The Ultimate Video Processor:
GeForceFX + Radeon + Parhelia = "Radeohelia FX"
 
jmecor, while a single program might have to be multithreaded to take advantage of HT, many programs can easily take advantage of it. I have access to 4 dual Xeon systems with HT enabled, and I can tell you that HT is not just a marketing strategy, it really increases performance.

One cosmic ray simulation (I work at a physics institute)takes 17 minutes to finish. Two of them, on the same CPU and simultaneously, take 23 minutes, instead of 34 or something...

<font color=red><b>M</b></font color=red>ephistopheles
 
Two of them, on the same CPU and simultaneously, take 23 minutes, instead of 34 or something...
really? That's quite impressive. I can certainly see how HT can help, but I didn't realise it could make <i>that</i> much difference... AFAIK that sort of thing is all CPU really, so I don't know where it finds the 'idle' time from the one thread, to do stuff with the other... Intel have a damn fine bit of kit there..

just out of curiousity, have you tried that on an AXP rig? just wondering how it would compare....

The more I learn, the more I think AMD are doomed - in their current guise at least.... I still hope they prove me wrong with A64 tho.


---
$hit Happens. I just wish it would happen to someone else for a change.
 
Yea 2.4c...i've heard numberous reports that it can hit 3ghz on stock cooling with ease. that is of course with good ram and nice oc'able board.



The customer is not always right.
If they were, they would do my job.
 
Okay. So, from what I'm seeing, I should probably be going for a P4 C. Just a couple more quick questions:

1) What's the best board for the money? (Just disregard the $450 CDN budget thing for a moment.)

2) This is kind of what might hold me back from getting a P4. What's the quality of the onboard sound solutions for the Intel boards (AD1985, for example), compared to NVIDIA's SoundStorm?

I'm looking at the onboard sound solutions because I don't really have the extra cash to blow on an Audigy2 at the moment.
 
Chipdeath, I assure you that I was as impressed as you seem to be. Astonishing, isn't it?... Anyway, no, I haven't had the opportunity to try the simulation on an AXP rig. I was considering asking for acess to the two AMD760-based dual MP 2000+ for an hour or so to run those cosmic ray simulations; however, I still didn't get to it. (the 4 dual xeon rackmounted systems are dual 2.4Ghz, 533Mhz FSB, E7501 chipset with dual channel DDR266 and 3GB system memory... They were not cheap, though.) Seriously, though, I wouldn't expect the dual MP 2000+ to measure up. Keep in mind that we are using Intel's compilers... not GCC.

<font color=red><b>M</b></font color=red>ephistopheles
 
There are a lot of i865PE mobos out there. (Make sure that it's an i865PE because there are also a number of i865P boards that you'll want to avoid.) Abit's IS7 line is supposed to be good, so long as you get the revised version with the quieter fan on the northbridge. (The initial versions had a REALLY loud fan appearantly.) Personally though I bought the Asus P4P800 Deluxe for my new PC because I wanted the IDE RAID1 option that came with the deluxe model and I wanted the fanless northbridge that Asus uses in their P4P800 mobos. I'd imagine that the plain P4P800 would be just as good and cost considerably less. It just doesn't come with Firewire and IDE RAID.

And the onboard 5.1 sound on my Asus P4P800 Deluxe is just awesome! So don't regret not getting an Audigy2 because a lot of mobos have really good onboard sound these days. :) And really, it's the speakers that matter more than the sound card in my opinion.

"<i>Yeah, if you treat them like equals, it'll only encourage them to think they <b>ARE</b> your equals.</i>" - Thief from <A HREF="http://www.nuklearpower.com/daily.php?date=030603" target="_new">8-Bit Theater</A>
 
What about the Intel Pentium 4 2.4GHz 512k socket 478 Hyper Threading 800MHz FSB at $170.00 and MSI 865PE NEO2-LS P4/FSB 800/DDR400/8X AGP at $100.00. Total of $270.00 at pricewatch.com. Tomsharware just did a review on that motherboard.
 
Are you guys building a Cosmic Ray Gun? While your at it.
Hmm....

With a fast enough processor, some custom software, and an array of home-built USB devices that run powerful electromagnets you might be able to build a rail gun. :) ...Or at least a gauss rifle. That'd be pretty neat.

"<i>Yeah, if you treat them like equals, it'll only encourage them to think they <b>ARE</b> your equals.</i>" - Thief from <A HREF="http://www.nuklearpower.com/daily.php?date=030603" target="_new">8-Bit Theater</A>
 
I live in canada too. My choice was:

AMD barton 2500+ <b>127$</b>
or
Intel 2.4C <b>246$</b>

I chose the cheap road. AMD all the way! They might crap out and lose everything, but not before I get atleast another upgrade or two out of them.


And if they did ever crap out. They'd liquidate they're inventory, and I could get a hammer for like 50$
 
**ROFL** I don't know what's worse, that we got onto this topic or that I've actually considered building one at home in the past. If I had a testing range and the money to blow on parts I'd probably give it a go. I mean people build catapults to lob pumpkins, so why not build a rail gun?

The way I figure it, it'd be best to design it as a two stage system. The first stage is controlled by USB and the electomagnets are powered directly. Here millisecond timing isn't quite as essential because the velocities involved will be lower in this stage. What's really needed is just raw magnetic power to get the projectile moving. So you just need to play around with the code until you get the timings right for the strength of the magnets and the projectile to maximize it's acceleration through the first stage.

The second stage is more of an electro-mechanical marvel. Here the accuracy of the timing is more important, so it's best to leave the PC out of it. The magnets will need to power down just as quickly as they power up, so here you'll want them to run on capacitors instead of run a straight line from a power source. Charge capacitors for each magnet seperately. Trigger the magnet using a cheap optical sensor as the switch with a recessed LED as a light source. Set up an array of these as seperate independant devices and you're good to go. :)

The hard part is the calibration of the second stage. The timings would be adjusted by the placement of the optical sensors, and each magnet would have to be timed seperately. I'd imagine that at some point you'd have to move the optical sensors further and further before the magnets because as the velocity of the projectile increases you'll have to start accounting for lag in the circuitry.

And the construction should probably be out of a ceramic compound for a number of reasons. You don't want the barrel getting magnetized and throwing off the calibrations. You want something smooth, hard, and heat resistant. I can't think of anything better than ceramic for that.

Sadly, the rail gun is probably the easy part. It's building the mecha to mount it on that's hard. 😉

"<i>Yeah, if you treat them like equals, it'll only encourage them to think they <b>ARE</b> your equals.</i>" - Thief from <A HREF="http://www.nuklearpower.com/daily.php?date=030603" target="_new">8-Bit Theater</A>
 
Hehe these are the things i never thought i would see on thg...

Make sure you are far away from your computer when you launch it...the EMP could destroy some data

So easy to fry yet tastes so good...
Silicon: The other, other, other white meat :evil:
 
Sounds good. So how much would it cost to build that thing?
Beats me. I haven't really checked prices.

Simple optical sensors are getting pretty cheap though, and LEDs have always been cheap. The magnets you could possibly even make yourself. Large capacitors and the USB electronics and microcontrollers are probably the expensive parts. Well ... other than the ceramics.

My thought was to start out small at first, with maybe a projectile the size of an arrow head. (Not hunting size but just regular target shooting size.) I would imagine that for a grand or two you could build a small one.

Maybe it'd cost as much as ten grand if you wanted to take the project seriously and build something more dangerous than just a regular old hunting rifle. If you wanted to make it portable ... I'm not sure. Once you start talking portable you then have to worry about the power source. Even then, I doubt man-portable would be possible without some serious cash.

"<i>Yeah, if you treat them like equals, it'll only encourage them to think they <b>ARE</b> your equals.</i>" - Thief from <A HREF="http://www.nuklearpower.com/daily.php?date=030603" target="_new">8-Bit Theater</A>