Choosing a processor for specific tasks

[Spacehead]

I'm thinking of upgrading my current processor(500mhz Celeron) to something with a little more speed. I also realize i need to upgrade mobo & RAM too.
Here's the things i am mainly going to be running on it:

Windows 98se(already installed)
- music creation(sotftsynths, recording software, etc.)
- CD burning
- occasional gaming (Quake type games)
- SETIathome

I've been kind of leaning towards an AMD 1800+, but i'm not sure if maybe i should be looking at a PIII(around 1 Ghz).
Anyone think there will be any major difference between either of these CPU's? The music creation is the main thing for this computer.

--------------------------------------------------------------------------------------

I'd also like to get a new computer for mainly websurfing & just general type stuff. Nothing real CPU intensive.

- would probably go with Windows 2K or XP.
- would like to have around 1Ghz CPU in it.
- running SETIathome

Any recommendations for a CPU for this type of application? AMD or Intel(Pentium-Celeron)?

 

[CrawlingEye]

It seems like you're not going to upgrade for a while, I'd say get a P4 1.6a and OC it.

You really can't go wrong with something like that.
The retail 1.6a will give you performance over the 1800+ with 3rd party cooling.

Especially since you want to play things like Quake, as well.

 

[FIFO]

I pretty much have to agree with CrawlingEye. At this point...


However, if I were you, I would wait till at least Sept before making a decision. There is all kinds of stuff going on in the CPU realm that no one can confirm on right now. There are reports that the T-bred (.13 XP) can hit well into the 2Ghz range. If this is true, a t-bred will be a MUCH better dollar spent. Especially if we get a little AXIA deja-vue and you can buy the slowest t-bred and OC it to the end of time.

Beyond that, there is also a lot of talk about the non-SOI Barton that will continue and still have 512 KB L2 cache. If the L2 cache does not limit the OCing and it is just as good as the t-bred, then we could ALL be in for a big surprise.

I was almost going to upgrade from a t-bird 900 @ 933, but I think wisely am going to wait till the end of the summer to see what all of these rumors turn into.

Ultimately in short though:

- P3 and T-bird based Athlon (Socket A non-XP) are basically neck and neck, clock for clock. Athlon sells for less.

- ANY P4 is the slowest clock for clock. By quite a lot in some instances. (No one talk Q3, we all know it flies on any Intel)

- More often than not, if you are just doing mobo, CPU and RAM, AMD is MUCH better dollar for dollar.

- An Athlon XP 1900+ sells for about the same as an Intel P41.6a, so by clock for clock, the P4 NEEDS the huge overclock just to keep up. (AXP 1900 runs at 1.6Ghz)

- If you really want to nitpick balls to the wall, then Intel does have the fastest system CURRENTLY available. Unfortunately, you have to pay more than twice the money for the CPU+RAM+MOBO and you only get maybe 5% to 10% better performance. (There is also a 900 Mhz difference needed by a P4 to pull this off. As well as much higher FSB bandwidth and the, although not by much lately, most expensive memory solution.)


From a personal side, it really comes down to where you are comfortable and how much you are willing to spend. But economically AMD wins every day of the week.

Good luck and have fun!!!
Jason.

 

[CrawlingEye]

Actually, the p4's have been dropping in price very rapidly, making them better deals than the AXP's.

Another thing, if you want to look into Tbred and say it'll scale to 2ghz and still keep it's higher IPC, you're obviously not thinking properly.
The higher you go, the more IPC you lose, that's just basic. At the current rates, if you could compare a 5ghz AXP to a 5ghz P4, the P4 would actually win, clock for clock.

Another thing to note is that AMD's are made to run current code very efficiently, but long-term they lose speed.

(If you compare an AXP 1900+ 3 years or so from now to a 1.6a, the 1900+ will be nearly obsolete and the 1.6a will still work fine)

 

[Spacehead]

Thanks for the replies so far guys.

One thing i should have mentioned before, i don't know anything about overclocking at this point, so any CPU i get will be running at stock speed for awhile.

Price is an issue. One of the things luring me to the 1800 is the price, only a couple of bucks more than the 1700, but about $30-$40 less than the 1900.

As for the Win 98se machine, i've heard to either get a high end P4 or go with the P3 for music apps if your going to go Intel. I've heard alot of good things about both P3 & AMD XP's for music. My gut feeling is that i probably won't go wrong with either.

 

[Damascus]

Originally posted by: CrawlingEye
Another thing, if you want to look into Tbred and say it'll scale to 2ghz and still keep it's higher IPC, you're obviously not thinking properly.
The higher you go, the more IPC you lose, that's just basic. At the current rates, if you could compare a 5ghz AXP to a 5ghz P4, the P4 would actually win, clock for clock.

Another thing to note is that AMD's are made to run current code very efficiently, but long-term they lose speed.

(If you compare an AXP 1900+ 3 years or so from now to a 1.6a, the 1900+ will be nearly obsolete and the 1.6a will still work fine)

Could someone explain this to me?

 

[FIFO]

That is very true Spacehead. Basically, there is no such thing as buying the "wrong" thing any more. They are all very fast and overall, well priced.

Now I have to seriously disagree with that comment that CrawlingEye made in regards to the t-bred and IPC. IPC is constant according to the CPU core design/stepping. If there is no change and you can ramp the operating speed, then there is a linear increase in instructions per second. If your explanation was correct, then the P3 1Ghz should have been no better than a P3 750. Or perhaps you could explain why a P3 faster than 1Ghz can trash a P4 from 1.5 to 1.8 just like an Athlon can?

The only thing that decreases the IPC of a CPU is a core change. Prime example: P4, dropped the IPC so that a 1.5 Ghz was necessary to perform as fast as a 1-1.2 Ghz P3 or Athlon. The ultimate goal was to dupe the masses with a wildly high operating speed and create the illusion of a performance advantage.

Now of course if your FSB bandwidth sucks, then you are not going to have anything to process and the whole thing is a moot point.

Any hoo,
Happy computing!!!

 

[CrawlingEye]

Erm, IPC isn't "instructions per second" first off, it's "instructions per cycle."
If it were "second" then we would pretty much have replaced the PR system.

Second, IPC ALWAYS decreases as you scale higher.

Here's a good read for you: http://www.sharkyforums.com/showthread.php?s=&threadid=137171&pagenumber=1

 

[Damascus]

I still don't get what you're trying to say... the higher a CPU scales, the more IPC it loses?
5Ghz XP would lose to a 5Ghz P4, but wouldn't the P4 also lose IPC according to what you
said, since it is just "so basic"? The crux of what you're saying is that an XP loses IPC as it
scales because it is AMD?

 

[FIFO]

You are right on the money there Damascus. Crawlingeye sounds like an AMD hater/basher!!! He wants cake and to eat it too.

If what you said is tru, then Damascus is right on the money.

The problem is that the link that you sent was firstly a waste of reading and secondly ignored my earlier statement on FSB BW.

When people talk of IPC/IPS.... Let me rant about that for a sec. To get your IPS you need IPC. You go IPC x Mhz/Ghz = IPS. Basically you lipped at me and tried to say that Mph has nothing to do with Fps (feet per sec.) The two are interrelated.

Anyhow, SPECint/fp, they tend to be impacted by FSB BW. Anand used to run an rc5 benchmark to show the raw power of a cpu and that is the best test that can be run to demonstrate what I am talking about. The test fits into L1/L2 cache and therefore does not rely on the main memory. This eliminates the possibility of a poor mobo design from impacting the performance of the CPU. It makes it a perfect level playing field.

Considering Crawlingeye seems to be an Intel zealot, maybe he should consider the following:

P3 beats Athlon in Q3 = Advantage SSE
AXP gets SSE = No more advantage Intel.
Sysmark does not recognize the SSE in AXP = Advantage Intel
Someone "fixes" ths and all of a sudden = No more advantage Intel
Intel hates this, BapCo makes a "new" version and pow = Very disputable and unexplainable drop in AXP performance

With SSE or with it turned off on both Intel and AMD, clock for clock there is no competition for the AMD. If your argument was correct, then you should have seen the Athlon bottom out at 1 Ghz. You hav not. What you have seen is that the Athlon needs more FSB BW if the WHOLE SYSTEM (NOT CPU) is to scale properly. There is a myriad of BMs out there that show that the Athlon could benefit greatly if the FSB wsa "officially" increased.

Intel could have VERY easily kept the very good FPU from the P3 and just "tossed" in SSE2 and a bit longer of a pipeline to allow for an acceptable increase in Mhz/Ghz. They did not. They wanted back what AMD had taken. They wanted the Ghz crown back. There was no concern for the overall performance of the CPU. Even this hyperthreading thing has been shown by anandtech to not really be that big of a deal, considering it is further needed to level the clock for clock playing field between AMD and Intel.

I am now done feeding the troll.

 

[CrawlingEye]

I'll give you another read. Note: That's from AMD fans.

That should explain how the p4's can scale higher than the AXP's and even beat them at the same mhz rating.
If you want a little thought to ponder, consider this:

The AXP 2100+ is 1.7ghz, but the 2200+ will be 1.8ghz.
Their PR rating system has it so it should only be an increment of 66mhz not 100 though, for the 100point increase.

Why does it now take 100mhz?
Gee, maybe their IPC has decreased?

 

[CrawlingEye]

Make sure to read the last 3 pages of the link I gave.

 

[FIFO]

You still don't get it do you???

Here is some CPU 101 for you that I have picked up along the way:

If you have two CPUs running EXACTLY at the same clock speed and one has a longer pipeline than the other, it is a well known and accepted fact that the CPU with the longer pipeline will take longer to start producing results. It takes a full clock tick or 1 hz for the data to move from one stage to the next. EX:

Athlon = 10 stage = 11hz before first results starts to pour out.
P4 = 20 stage = 21hz before the first result starts to pour out.

Now this is just the start. After they both start pouring out data there is no real difference until you introduce branch prediction. When you "guess" what data is going to be needed, you are taking a risk that you may be wrong. When this happens the whole pipeline is flushed and has to start over. That is a minimum of twice the penalty that an Athlon will incurr when a P4 makes a branch mis-predict. Considering that the branch prediction accuracy is awfully close between the two arcitectures, that the P4 will incurr more idle clock ticks due to having to wait for the pipeline to fill back up. In a pound for pound match, there is basically no way that a hyper-pipelined CPU can keep up with a short pipeline design.

Now considering that the P4 is marketed on the premise that the higher the hz the better, please digest the following.

It is arcitecturally given that pipelines must continue to get longer in order for more performance to be had. This is evident in that the Hammer series is having its pipeline extended from 10 (Athlon) to 12(Hammer). The effect of lengthening the pipeline is that you gain the ability to increase clock speed to points higher than what you could previously. The "trick" is to increase the pipe just enough so that you do not come up with a design that due to the limits of silicon, may never be reachable. I read not very long ago a very good article that stated that anything beyond 16 stages in a pipeline is really pushing it. Branch prediction needs to be spot on or the penalties would be severe.

Now with all of this in mind lets look back at your argument here and in the Sharkys forum in regards to "scalability".

You both here and there have been conviently swapping "scale" back and forth between ipc/ips and clock speed. You will never get an argument from me that the K7 Palomino (AXP) arcitecture cannot scale CLOCK SPEED WISE as well as the P4. Now due to the following formula:

IPC x (hz per sec) = Instructions per second

The following numbers have been extrapolated directly from the ALU results of SiSoft SANDRA 2002

Athlon = 2.75 x 1Ghz = 2750 mips
P4 = 1.84 x 1Ghz = 1840 mips

For the P4 to "scale" to the same IPS as the Athlon it needs to do the following:

1.84 x 1494565217 hz (1.5Ghz) = 2750 mips

Even if I look at FPU results...

Athlon = 1.3708 FLOP per hz
P4 = 1.2216 FLOP per hz

We still see the Athlon performing faster than a P4. And to add insult to injury, the P3 is basically the EXACT same perf as an Athlon clock for clock.

Now this is the "Scale" that I am talking about. You cannot dispute it as even with SSE and SSE2, there are TONS of benchmark results that show that a 1.5 Ghz Athlon can and does Destroy a 1.5 Ghz P4. For the P4 CPU to compete it NEEDS to "scale" its CLOCK SPEED up VERY high to exceed the AXP in the final result called FPS in a game or time to complete in rendering software or some arbitrary number generated by a productivity benchmark.

In the end, you really want to stop using the term "scale" for clock speed. That should be more properly referred to as "ramping up" as it NEVER ends up with a linear scaled increase in overall performance for the system that the CPU is in.

This is where the P4 can "Scale". The P4 has a larger amound of FSB BW and that allows more of the clock speed increase to actually be used. This is where the Athlon would suffer if it were to be left at 133 x DDR for an FSB. The reason that this is only becoming a bottleneck now is that even for all of the technical merrit of RDRAM, it cannot beat the low latency of DDR-SDRAM. This battle can be similarly compared to the whole pipeline issue. Does one optimize for most data per clock, or do you shoot for as many clock cycles that you can make happen in the same period of time with less data per clock.

There is nothing fundamentally wrong with Intel choosing the road that they did and AMD theirs. What really burns my a$$ is that people seem to forget that you are not buying Mhz, you are buying FPS in a game or time to complete in rendering software or some arbitrary number generated by a productivity benchmark.

If you can get more in one category by purchasing an AXP and spending less then great. If you are a business and price is secondary to how fast you can/need to get the work done, then you obviously buy the fastest, regardless of clock speed.

My closing words are: Buy what suits your NEEDS best and your budget. Otherwise, ignore the zealots. I used to be one for both sides and I can tell you as an IT professional that has no "choice" when buying servers for work that I will still plan on buying Intel for the forseeable future, even if Hammer is blow your socks off awesome. Stability at the server level is far more important than IPC/IPS/hz.

Jason.

P.S. -- For the REAL mission critical apps we only use HP9000 PA-RISC running HP-UX. x86 and MS be damned.

 

[RaynorWolfcastle]

I don't think I've seen a thread this full of misinformation (on both sides) in quite some time

CrawlingEye:

Second, IPC ALWAYS decreases as you scale higher

That's quite simply untrue... in an ideal world where all the system components increased at the same speed (ie FSB, hard disk, CPU, etc) you'd keep a constant IPC. However in real life, upping a 1 GHz CPU frequency by 100 MHz and changing nothing else will (for obvious reasons) never give you a 10% boost in speed. It may get close to 10% if the application is almost wholly CPU dependent, but it will never reach exactly 10%.
The main reason IPC decreases as CPU clockspeed is upped is because sooner or later another part of the system becomes the bottleneck. That being said, there is no reason the Athlon's IPC should drop faster than the P4's unless there is a bottleneck elsewhere in the system (FSB maybe?).

IPC x (hz per sec) = Instructions per second

hz per second?!?!? You realize of course that Hz is just shorthand for cycles/second right? The correct formula is in fact IPC (instructions/cycle) x Frequency (cycles/second) = Instructions per second.

I'm staying away from any further comments on architecture since it'll probably lead to my being called a zealot of some sort.

spacehead For what you're planning to do (considering you won't be OC'ing), I'd get the Athlon XP 1800+, it'll be plenty fast for what you'll be doing with it.

-Ice

 

[FIFO]

I love how you start out to say that there is a lot of misinformation and all you come up with against me is a SLIGHT formula mistake!!!

Hz = Cycle

therefore

Cycle = Hz

You chose to use the word, I chose to use the symbol (and accitentally get redundant: Hz per sec = cycles per second, per second. I must have thought I was falling Cudos to anyone who gets that joke. ) . The math that you put is EXACTLY the same as mine!!!!!


Now I actually invite you to make as many comments that you want to as long as they are based as close to fact as possible. If I am wrong, please site example in articles that prove so. Do not do what Crawlingeye does and direct us to some other thread of a rant that contains little fact but much confusion.

As you can tell, I love to debate!!!

Why not keep going! It will make us all better for the next time!!!

Jason.

 

[CrawlingEye]

Actually, RDRAM PC1066 is closing in on the latency losses. Just the same, DDR is ramping up for the p4 with the introduction of the SiS 655 chipset that's to be released fairly soon (dual channel DDR333), as well as granite bay.

On a clock for clock basis, the AXP wins, that's been stated. Your point in saying that is obvious, however at 5ghz vs 5ghz, it's quite likely that the p4 could beat it on a clock for clock comparison (yes, the numbers are given intentionally for effect).

The pipeline is intentionally lenghthened, and the predictions are more accurate with the P4.
However, the AXP has more time to correct the mispredictions, where as the P4 takes longer.
It's only obvious that the easy solution to this is to add more bandwidth to correct the mispredictions, so goes why the P4 has better memory controllers and so forth.

If you're trying to argue something specific instead of "AMD's better than Intel because I said so" then make your point.
You seem to be beating around the bush quite a bit.

 

[RaynorWolfcastle]

Originally posted by: FIFO
I love how you start out to say that there is a lot of misinformation and all you come up with against me is a SLIGHT formula mistake!!!

Hz = Cycle

therefore

Cycle = Hz

You chose to use the word, I chose to use the symbol (and accitentally get redundant: Hz per sec = cycles per second, per second. I must have thought I was falling Cudos to anyone who gets that joke. ) . The math that you put is EXACTLY the same as mine!!!!!


Now I actually invite you to make as many comments that you want to as long as they are based as close to fact as possible. If I am wrong, please site example in articles that prove so. Do not do what Crawlingeye does and direct us to some other thread of a rant that contains little fact but much confusion.

As you can tell, I love to debate!!!

Why not keep going! It will make us all better for the next time!!!

Jason.

Ummm.... A Hertz is not a cycle... Hz != Cycle. Hz = Cycle/second (no offense, but that joke was terrible btw )

-Ice

 

[CrawlingEye]

Originally posted by: Damascus

Originally posted by: CrawlingEye
Another thing, if you want to look into Tbred and say it'll scale to 2ghz and still keep it's higher IPC, you're obviously not thinking properly.
The higher you go, the more IPC you lose, that's just basic. At the current rates, if you could compare a 5ghz AXP to a 5ghz P4, the P4 would actually win, clock for clock.

Another thing to note is that AMD's are made to run current code very efficiently, but long-term they lose speed.

(If you compare an AXP 1900+ 3 years or so from now to a 1.6a, the 1900+ will be nearly obsolete and the 1.6a will still work fine)

Could someone explain this to me?

The higher you scale, the longer of a pipeline you must create.
The longer of a pipeline you create, the further your IPC decreases.

Intel's all ready at 2.5+ghz and have a comparable IPC to AMD, who is at only 1.8ghz.
The P4 in a linear comparison very well could beat an AXP if you took it to an extreme comparison like 5ghz vs 5ghz.

 

[FIFO]

Well it looks like icecool83 seems to be awfully short on factual criticism to the point that all he can do is nit-pick about a small slip-up in a term that I used.

And in regards to CrawlingEye..................................

I think that all of my pro intel comments all wrapped up in one little nutshell go like this:

A system based on an i850E and P4 2.53 Ghz with PC1066 RDRAM is the undisputed overall champion of the galaxy. I will not dispute that.


I cannot however STRESS the fact enough that I am speaking SPECIFICALLY ABOUT THE INTERNALS OF THE CPU ONLY!!!!!!!!!!!!!!!!!!!!!!!!!!!

If I jacked the P4 and the AXP to 5Ghz the AXP WILL beat the P4 IF I/O TO THE REST OF THE PC IS NOT AT ISSUE!!!!!!!!!!!!!!!!!!!!!!!!!!!!

AMD is more intent on finding a way to maximize IPC than maximizing Ghz. As a bonus though, AMD has to also make their Ghz higher for marketing reasons. Due to the very high IPC, when the AXP cpus get within even 600 to 700 Mhz of the P4, they match or beat the P4.

Beyond this, if the benchmarks that were done on the 800 Mhz hammer are any indication and AMD can get this puppy to 2 Ghz and beyond, the P4 doesn't stand a chance except in the land of the marketing lemmings.

Like statements I made earlier elude to: Intel may have tried to push the P4 architecture out a little early. It is possible that the P4 may prove longer lasting than some may think, but due to the fact that it is Now and not 2005, that doesn't really matter.


Now on your new comments there Eye:

Quote

The higher you scale, the longer of a pipeline you must create.
The longer of a pipeline you create, the further your IPC decreases.


Intel's all ready at 2.5+ghz and have a comparable IPC to AMD, who is at only 1.8ghz.
The P4 in a linear comparison very well could beat an AXP if you took it to an extreme comparison like 5ghz vs 5ghz.

End Quote.

Your first statement actually refutes your second. The AXP has yet to need to have its pipeline extended. This means that the IPC is not affected whenever AMD produces a higher clocked version. Therefore, if AMD could ever get the current Palomino core to the magical speed of 5Ghz, it would utterly trounce the P4 in an "Instructions per Second comparison". And before you comment again, remember that I am speaking of the CPU. Not the CPU and the Northbridge.

 

[CrawlingEye]

Actually it doesn't. The P4's all ready ramped up and it's current architecture will take it to 3ghz and likely further.
AMD will need to increase their pipeline to compete with p4's at these speeds, giving it even less IPC, while Intel's pipeline remains the same.

In terms of architecture, you can't seperate the northbridge from the cpu since one reflects the other.


Regarding your 800mhz hammer? That's a joke.
Unless AMD's trying to become apple and pretty much release cisc cpu's, then they're not releasing an 800mhz cpu, nor will the hammer perform that well. 800mhz allows significantly smaller pipelines which with good ALU's could perform fairly well. However, it would be obsolete in a VERY short period of time, due to it's lacking legacy code support.

 

[RaynorWolfcastle]

Originally posted by: FIFO
Well it looks like icecool83 seems to be awfully short on factual criticism to the point that all he can do is nit-pick about a small slip-up in a term that I used.
And in regards to CrawlingEye..................................
I think that all of my pro intel comments all wrapped up in one little nutshell go like this:
A system based on an i850E and P4 2.53 Ghz with PC1066 RDRAM is the undisputed overall champion of the galaxy. I will not dispute that.
I cannot however STRESS the fact enough that I am speaking SPECIFICALLY ABOUT THE INTERNALS OF THE CPU ONLY!!!!!!!!!!!!!!!!!!!!!!!!!!!
If I jacked the P4 and the AXP to 5Ghz the AXP WILL beat the P4 IF I/O TO THE REST OF THE PC IS NOT AT ISSUE!!!!!!!!!!!!!!!!!!!!!!!!!!!!
AMD is more intent on finding a way to maximize IPC than maximizing Ghz. As a bonus though, AMD has to also make their Ghz higher for marketing reasons. Due to the very high IPC, when the AXP cpus get within even 600 to 700 Mhz of the P4, they match or beat the P4.
Beyond this, if the benchmarks that were done on the 800 Mhz hammer are any indication and AMD can get this puppy to 2 Ghz and beyond, the P4 doesn't stand a chance except in the land of the marketing lemmings.
Like statements I made earlier elude to: Intel may have tried to push the P4 architecture out a little early. It is possible that the P4 may prove longer lasting than some may think, but due to the fact that it is Now and not 2005, that doesn't really matter.

The truth of the matter is that I don't want to dig up sources to support my claims (so that it doesn't sound like I'm pulling it out of my ass). For the record, a lot what you have said so far is correct FIFO, BUT:
A) When you provide information, provide a source as well

B) Unless you have benchmarks that I am not aware of, we don't know how the Hammer will perform (that Quake benchmark counts for nothing since the string in WCPUID was grayed out and the processor was running at 800 MHz). I'll reserve judgement on the Hammer for when I see production-level silicon reviewed by a well respected website.

C) AFAIK, the Pentium 4 has better branch prediction than the Palomino, not positive about this though. pm may be able to confirm this.

D)

The following numbers have been extrapolated directly from the ALU results of SiSoft SANDRA 2002
Athlon = 2.75 x 1Ghz = 2750 mips
P4 = 1.84 x 1Ghz = 1840 mips
For the P4 to "scale" to the same IPS as the Athlon it needs to do the following:
1.84 x 1494565217 hz (1.5Ghz) = 2750 mips

Your basing your comparison of processor performance on Sandra? Please explain what are these numbers that your multiplying together. Remember, the Athlon usually runs applications designed for the P6 architecture better than the P4, but if the program is properly optimized for SSE2 things change quite a bit (check out Anand's Lightwave 7.5 benchy, you'll find that a 1.6 GHz Willy beats the 1.8 GHz Tbred in one of them ). Bottom line, though your general idea is correct, your example is terrible as the IPS changes for every application (and for that matter every platform).

E)

Intel could have VERY easily kept the very good FPU from the P3 and just "tossed" in SSE2 and a bit longer of a pipeline to allow for an acceptable increase in Mhz/Ghz.

Wow, you should definetly apply at Intel, they could definetly use a CPU architecture genius like you.

They wanted back [...] the MHz crown.

LOL, did it ever occur to you that the P4 architecture was designed well BEFORE AMD took the MHz crown from Intel?

F) You're oversimplifying things a little too much and your "guesses" as to what the performances of future processors is like are laughable since you are essentially basing your assumptions on the scaling of current CPUs and don't (and can't for that matter) predict the effect of architectural enhancements on each processors.

I could probably go on, but I think this proves my erroneous information point well enough.


:frown: I'm done with this thread. Sorry your thread turned into this Spacehead (my CPU suggestion is above).

-Ice

PS: There are certainly flaws in Crawling Eye's arguments, but since he didn't go so far as to imply that I'm a nit-picker who understands nothing, I didn't bother pointing them out.

 

[CrawlingEye]

Go on, icecool. I want to hear the flaws in my statements.

 

[Swit]

As we re-inject the original post in a subtle reminder to avoid thread piracy...

Spacehead,
Either cpu will get where you want to go, I have no doubt. Your more likely to run into problems with specific hardware / driver conflicts on your motherboard and peripherals than any problem inherent to either the AMD or the Intel cpu.
If you already have music hardware/software that you'll be using, check user groups (newsgroups or forums)and hardware/software support sites to see what (if any) combinations give grief. Read these with a grain of salt. Beware of bug reports, accusations of crappy hardware and such disguising a lack of RTFM (Read The Friendly Manual).
With that caveat, the groups will help you avoid trouble.
A bit of background research to augment your list of 'things you want to do', good components and you can make a box that'll make your budget happy and not cause you buyer's remorse.

G'luck

Swit

Originally posted by: Spacehead
I'm thinking of upgrading my current processor(500mhz Celeron) to something with a little more speed. I also realize i need to upgrade mobo & RAM too.
Here's the things i am mainly going to be running on it:

Windows 98se(already installed)
- music creation(sotftsynths, recording software, etc.)
- CD burning
- occasional gaming (Quake type games)
- SETIathome

I've been kind of leaning towards an AMD 1800+, but i'm not sure if maybe i should be looking at a PIII(around 1 Ghz).
Anyone think there will be any major difference between either of these CPU's? The music creation is the main thing for this computer.

--------------------------------------------------------------------------------------

I'd also like to get a new computer for mainly websurfing & just general type stuff. Nothing real CPU intensive.

- would probably go with Windows 2K or XP.
- would like to have around 1Ghz CPU in it.
- running SETIathome

Any recommendations for a CPU for this type of application? AMD or Intel(Pentium-Celeron)?

 

[Migroo]

I'd go with any Athlon XP over a P4 just because its my preference not to need to overclock to maximise performance. I'd definately prefere a stock speed 1700+ over a stock speed P4 1.6.