CPU Upgrade Question, Please help

[HansGroenewold]

I'm planning an upgrade on my PC and wanted to know what a good CPU upgrade would entail including a video card. I want the best bang for the buck like everyone else and your input would be deeply appreciated. I would like to stick with AMD and another Nvidea video card. Listed below are my current system specs.

3500+ AMD Athlon CPU
Asrock 939 Dual Sata 2 Socket 939 Motherboard, will take both AGP and PCI cards
2 GB RAM
6800 Ultra 256MB Video Card

Sincerely,
Hans Groenewold

 

[cmdrdredd]

If you can find a X2 for socket 939 BUY IT. I can't list specific models because they're hard to find. Maybe a Opteron 165 or similar.

For your video card I would wait for the 8800GTS/GTX. Not to buy them, but because the other cards at the high end like the 7950GX2 and 7900GTX etc will fall in price as well. This presents itself as a excllent time to upgrade. This is all asuming you aren't installing Vista on day 1 and want to play every DX10 game that comes out.

Almost every game will still use DX9 or OpenGL, and those that don't will probably only use a few DX10 features and work under DX9 anyway. The only games that may not have support for DX9 at all include a few games like Halo2 (Vista only btw).

Although the 8800GTS/GTX will definately be faster than even a 7950GX@ and quite possibly 2 of those cards in SLI, the performance of the 7900GTX and other cards in the line are well documented and won't change. To me, having the fastest card and highest FPS isn't all that interesting, but getting a good value with a smooth gaming experience IS.

Everyone's different, but you sound like you are looking for the value since you're not upgrading to a C2D or AM2 right away.

edit: Does that board accept PCI-Express? Or just PCI...I don't have the specs in front of me and you said PCI.

 

[HansGroenewold]

Yes, it takes both PCI and AGP. I've been out of the loop regarding PC's since I bought this gaming PC almost two years ago. I don't use it that much but would like to upgrade it without breaking the bank. What do you mean by X2? Sorry if the question is lame.

Thanks,
Hans

 

[myocardia]

In answer to your question, cmdrdredd, yes, it has both AGP and PCI-E slots. And Hans, this is an X2: link.

 

[imported_stev]

I'm playing the X2 buying game myself. They've been selling out recently, especially the 4800 and now prices have risen a little. I'm new to this so I don't know if they're going to go back down once the gotta-get-it-before-it's-gone hype goes down.

 

[myocardia]

Most likely not, stev, since we're only 6 or 7 weeks from AMD's projected EOL on Skt. 939. It's possible that they'll start popping back up again, but don't hold your breath, I would think.

 

[imported_stev]

Originally posted by: myocardia
Most likely not, stev, since we're only 6 or 7 weeks from AMD's projected EOL on Skt. 939. It's possible that they'll start popping back up again, but don't hold your breath, I would think.

Could sometime tell me what the previous generation's X2 equivalents were? By that I mean what were the high-end chips that started around $800 or so and then went down to the $200-$300 range as they neared EOL? Did they disappear altogether shortly after a major price drop or did a late price increase after a major drop keep them in stock for awhile? Maybe this would give me a better idea of what was going on, but maybe since dual core is in play in this generation and not the previous one, the previous generation isn't relevant to this situation. I know that I'm getting an X2 to help power the unplanned HTPC duties that my main computer is now responsible for

 

[myocardia]

Well, there's really no way to know what will happen in the future. There are ten's, if not hundred's of thousands of Skt. 939 motherboards in America alone. I'd have to assume that the owner of nearly every one of them wants a dual-core for it. But, there's no way to know whether they'll all sell out, or whether the price on them will just go way up. Chances are very good that it will be one of those two choices, though, I would think.

 

[imported_stev]

Good point on the huge stock of unsold mobos out there. I'll just have to keep tabs on things. I would be very happy if I could get a 4600 for $200 or less, but maybe I'll snap it up sooner...in fact I should anyway. A $40 difference shouldn't mean much to me for a system that I will keep for a number of years!

HansGroenewold: Sorry for hijacking your thread. I've been checking out benchmarks at http://www23.tomshardware.com/cpu.html to compare the X2s to my 3500.

The 4200 is like having two 3500s
The 4400 is like having two 3500s with twice the L2 cache
The 4600 is like having two 3500s clocked at 200 MHz higher
The 4800 is like having two 3500s with twice the L2 cache AND clocked at 200 MHz higher

For me, I wanted to upgrade to dual core, but also get a speed boost on single core apps (so it feels like an upgrade, even in single core situations!) and since I'm not overclocking that eliminated the 4200. In the benchmarks relevant to my usage, the 200 MHz clock increase seemed to have a bigger effect on performance than the extra cache and the 4800 is a little more than I want to spend, so I'm zeroing in on the 4600.

Hope this helps.

 

[HansGroenewold]

Stev,

No problem, I get great info here and appreciate the input. I wonder if I should try overclocking my present CPU first and a new video card before trying a new CPU too. Whatcha think? It would be interesting to do one before the other to compare results. BTW, that's a good link.

Hans

 

[imported_stev]

Yes that would be interesting. I don't know jack about overclocking, but others around here do. I have yet to see solid facts on the effects of overclocking on longevity of a CPU (links anyone?). Most people do it for a short-term gaming performance increase. But hey, if it's on a CPU you may end up replacing anyway, why not? I'm interested to know how it goes.

 

[cmdrdredd]

Originally posted by: stev
Yes that would be interesting. I don't know jack about overclocking, but others around here do. I have yet to see solid facts on the effects of overclocking on longevity of a CPU (links anyone?). Most people do it for a short-term gaming performance increase. But hey, if it's on a CPU you may end up replacing anyway, why not? I'm interested to know how it goes.

It's likely that your CPU will last longer than you'll keep it in an overclocked situation provided you have adequate cooling to compensate for the heat output generate from higher frequencies and higher voltages if you do alter them.

 

[imported_stev]

Originally posted by: cmdrdredd

Originally posted by: stev
Yes that would be interesting. I don't know jack about overclocking, but others around here do. I have yet to see solid facts on the effects of overclocking on longevity of a CPU (links anyone?). Most people do it for a short-term gaming performance increase. But hey, if it's on a CPU you may end up replacing anyway, why not? I'm interested to know how it goes.

It's likely that your CPU will last longer than you'll keep it in an overclocked situation provided you have adequate cooling to compensate for the heat output generate from higher frequencies and higher voltages if you do alter them.

And that's pretty much what everyone says. But when I ask to what extent small/med/large amounts of overclocking will affect a CPUs lifespan (or the lifespan of other hardware), no one has any proof. Since there are tons of benchmarking datasets and no data on CPU lifespan versus level of overclock, that leads me to believe overclocking is purely a short-term user's game and of no interest to me on a system where being reliable is much more important than extra performance.

 

[cmdrdredd]

Originally posted by: stev

Originally posted by: cmdrdredd

Originally posted by: stev
Yes that would be interesting. I don't know jack about overclocking, but others around here do. I have yet to see solid facts on the effects of overclocking on longevity of a CPU (links anyone?). Most people do it for a short-term gaming performance increase. But hey, if it's on a CPU you may end up replacing anyway, why not? I'm interested to know how it goes.

It's likely that your CPU will last longer than you'll keep it in an overclocked situation provided you have adequate cooling to compensate for the heat output generate from higher frequencies and higher voltages if you do alter them.

And that's pretty much what everyone says. But when I ask to what extent small/med/large amounts of overclocking will affect a CPUs lifespan (or the lifespan of other hardware), no one has any proof. Since there are tons of benchmarking datasets and no data on CPU lifespan versus level of overclock, that leads me to believe overclocking is purely a short-term user's game and of no interest to me on a system where being reliable is much more important than extra performance.

If you actually plan to overclock a system and are worried about fault tolerance, 100% 24/7 uptime etc...you're going to be called an idiot.

 

[imported_stev]

Originally posted by: cmdrdredd

Originally posted by: stev

Originally posted by: cmdrdredd

Originally posted by: stev
Yes that would be interesting. I don't know jack about overclocking, but others around here do. I have yet to see solid facts on the effects of overclocking on longevity of a CPU (links anyone?). Most people do it for a short-term gaming performance increase. But hey, if it's on a CPU you may end up replacing anyway, why not? I'm interested to know how it goes.

It's likely that your CPU will last longer than you'll keep it in an overclocked situation provided you have adequate cooling to compensate for the heat output generate from higher frequencies and higher voltages if you do alter them.

And that's pretty much what everyone says. But when I ask to what extent small/med/large amounts of overclocking will affect a CPUs lifespan (or the lifespan of other hardware), no one has any proof. Since there are tons of benchmarking datasets and no data on CPU lifespan versus level of overclock, that leads me to believe overclocking is purely a short-term user's game and of no interest to me on a system where being reliable is much more important than extra performance.

If you actually plan to overclock a system and are worried about fault tolerance, 100% 24/7 uptime etc...you're going to be called an idiot.

Precisely why I wouldn't do it. For me, any performance gain wouldn't be worth the possible loss in reliability. Still, the topic of overclocking and the fact that no one ever talks about any downsides is interesting to me.

 

[cmdrdredd]

Originally posted by: stev

Originally posted by: cmdrdredd

Originally posted by: stev

Originally posted by: cmdrdredd

Originally posted by: stev
Yes that would be interesting. I don't know jack about overclocking, but others around here do. I have yet to see solid facts on the effects of overclocking on longevity of a CPU (links anyone?). Most people do it for a short-term gaming performance increase. But hey, if it's on a CPU you may end up replacing anyway, why not? I'm interested to know how it goes.

It's likely that your CPU will last longer than you'll keep it in an overclocked situation provided you have adequate cooling to compensate for the heat output generate from higher frequencies and higher voltages if you do alter them.

And that's pretty much what everyone says. But when I ask to what extent small/med/large amounts of overclocking will affect a CPUs lifespan (or the lifespan of other hardware), no one has any proof. Since there are tons of benchmarking datasets and no data on CPU lifespan versus level of overclock, that leads me to believe overclocking is purely a short-term user's game and of no interest to me on a system where being reliable is much more important than extra performance.

If you actually plan to overclock a system and are worried about fault tolerance, 100% 24/7 uptime etc...you're going to be called an idiot.

Precisely why I wouldn't do it. For me, any performance gain wouldn't be worth the possible loss in reliability. Still, the topic of overclocking and the fact that no one ever talks about any downsides is interesting to me.

Interesting how? We upgrade BEFORE the CPU fails. We use the old systems to run servers or DC boxes after we're done with them in our main rig.

I have a P4 2.6 @ 3.4Ghz that has been running just about 24/7 for going on 6 years now.

If you don't think that's enough...I can't help you

 

[imported_stev]

Originally posted by: cmdrdredd

Originally posted by: stev

Originally posted by: cmdrdredd

Originally posted by: stev

Originally posted by: cmdrdredd

Originally posted by: stev
Yes that would be interesting. I don't know jack about overclocking, but others around here do. I have yet to see solid facts on the effects of overclocking on longevity of a CPU (links anyone?). Most people do it for a short-term gaming performance increase. But hey, if it's on a CPU you may end up replacing anyway, why not? I'm interested to know how it goes.

It's likely that your CPU will last longer than you'll keep it in an overclocked situation provided you have adequate cooling to compensate for the heat output generate from higher frequencies and higher voltages if you do alter them.

And that's pretty much what everyone says. But when I ask to what extent small/med/large amounts of overclocking will affect a CPUs lifespan (or the lifespan of other hardware), no one has any proof. Since there are tons of benchmarking datasets and no data on CPU lifespan versus level of overclock, that leads me to believe overclocking is purely a short-term user's game and of no interest to me on a system where being reliable is much more important than extra performance.

If you actually plan to overclock a system and are worried about fault tolerance, 100% 24/7 uptime etc...you're going to be called an idiot.

Precisely why I wouldn't do it. For me, any performance gain wouldn't be worth the possible loss in reliability. Still, the topic of overclocking and the fact that no one ever talks about any downsides is interesting to me.

Interesting how? We upgrade BEFORE the CPU fails. We use the old systems to run servers or DC boxes after we're done with them in our main rig.

I have a P4 2.6 @ 3.4Ghz that has been running just about 24/7 for going on 6 years now.

If you don't think that's enough...I can't help you

Overclocking in general is interesting to me, for the same reasons people do it. I'm just looking at the good and the bad because I'm probably not going to upgrade as often as many people at AT do.

It is because most people who overclock upgrade before their CPU fails that there doesn't seem to be much information on how destructive (or not) varying levels of an overclock can be. I would say that such a significant overclock on your P4 for so long is encouraging, although I imagine it's not being maxed out as much now as it was when it was in your primary system. If that kind of longevity for an overclocked CPU is common, then I wouldn't have a big problem with doing it. So, thanks for that info.

I'm now going to stop my rampant hijacking of this guy's thread!

 

[HansGroenewold]

That's OK, this is interesting reading and informational as well.

Hans

 

[AndrewL]

Actually the longevity cmd is talking about is very common provided you know what your doing. Check out xtremesystems.org some of the stuff they do is overboard, but theres good threads about people who have had cpus overclocked for many many years. They say the average lifespan of a 24/7 (meaning everyhour of the day usage) cpu is 10 years. They say the absolute most the common overclocker cuts down his cpu life is 50%, that still leaves you with 5 years. You know how useless that cpu will be in 5+ years? Since it is probably the minimum life span youd get if you knew what you were doing you could actually expect 6-9 yrs.