I think Intel might be better in some cases now! :-(

[ArborBarber]

Well, I am an intel fan, although my biased opinion has no real validity for I've never owned an AMD machine. I do know that heat kills...Both companies have their issues, does one type of system seem to you to be more reliable then the other? From a system builder's point of view, price isn't the only deciding factor(as I'm sure you're aware.) Support is extremely important from a customer's point of view, as is reliability and value. I think within the last year intel has gotten the message...intel seems to be trying alittle harder lately...

 

[MadRat]

Originally posted by: ChefJoe
Sigh... I've been tempted by AMD's cheap price, but the higher thermal output, seemingly greater frequency of memory type changes/sensitivities to memory issues, and greater number of chipsets, games,etc requiring "patches" has kept me with Intel in my machines.

Uh, guy, try AMD products have LOWER thermal output. Athlon XP puts out alot less overall heat for comparable products. The TBred is a huge cut in thermal output in comparison to the Palomino, even more of a drop than they had going from Thunderbird-Athlon to Palomino-AthlonXP.

 

[kazeakuma]

For low end (price wise) AMD can't be beat for value, but for real performance systems Intel all the way. The market seems to have divided itself like back in the days of the Pentium, Pentium II.

 

[AntaresVI]

After reading this, I still find that on my tiny budget, the only way to get a fast processor under $100 is to go AMD. Plus, AMD mobos go from very cheap to very expensive, so there's leeway. I guess my point is that the cost of an Intel processor equivalent to my xp 1800+ would be prohibitive to me.

 

[Coherence]

Based on AnandTech's benchmarks, it seems CPU-intensive games (JKII) run better on P4, but FPU-intensive games (Quake, Unreal) run better on AMD. And since an AthlonXP 2400+ costs the same as a P4 2.4B, it really is a toss-up between to two, IMO.

I think it's becoming less an issue of which CPU to get, and more an issue of which chipset/motherboard to get. Whichever chipset/mobo combo has the features you want, at the price you want to pay, you get the appropriate CPU for it, AMD or Intel.

 

[Gunbuster]

My main problem is I cant run an AMD CPU on a Intel made/brand mainboard

Intel is my #1 choice for mainboards

I think bad\unstable\unsupported boards was what turned me off to AMD

 

[DT4K]

Originally posted by: Gunbuster
My main problem is I cant run an AMD CPU on a Intel made/brand mainboard

Intel is my #1 choice for mainboards

I think bad\unstable\unsupported boards was what turned me off to AMD

It's true that Intel is top-notch for chipset / mobo stability and compatibility but
there have also been some very good stable fast boards for AMD chips. KT266A, NForce2, KT333, are all great chipsets.

I have gone with AMD in the past because they were the best bang for the buck. And that still holds true for the low to mid range chips. The Athlon XP 1600 - 2100 are all great values right now.

But at the upper range, AMD chips no longer have any price advantage. The 2400+ is about the same price as the 2.4B.

As far as performance, it is a good competitive cycle.
A couple of years ago, AMD had the clear lead, with a 1.2 T-Bird outperforming a 1.8 P4 Willamette for a lot less money.
The Palomino furthered AMD's lead.
The release of the Northwood got Intel back on par with AMD.
The step from 400 to 533 pushed Intel ahead of AMD.
AMD's jump from 266 to 333 got AMD almost back to Intel's performance but not quite.
The release of the 3.06 with HT has now given Intel the clear performance lead.
For now.

AMD's pricing on the high end models is surprising. I guess they have to make money, but they are giving up their clear advantage as far as value goes. But maybe this is a long term strategy. AMD may believe that by selling their chips at a price equivalent to Intel, they can eventually get rid of their image as a value chip maker and instead be seen as more of an equal to Intel. This value image is what has kept them from gaining more market share in the general consumer and corporate markets. I think they no longer want to be a minority competitor and instead want to compete on more of a head-to-head level with Intel. If Hammer is successful, they certainly could move more into this role. Let's all just hope both companies can be successful. Competition is good.

 

[mrman3k]

I agree fully with Shanti's comments. That is how it is.

 

[Siddhartha]

Originally posted by: ArborBarber
Well, I am an intel fan, although my biased opinion has no real validity for I've never owned an AMD machine. I do know that heat kills...Both companies have their issues, does one type of system seem to you to be more reliable then the other? From a system builder's point of view, price isn't the only deciding factor(as I'm sure you're aware.) Support is extremely important from a customer's point of view, as is reliability and value. I think within the last year intel has gotten the message...intel seems to be trying alittle harder lately...

Anand addresses the heat issue

Here

 

[MadRat]

Just in case they missed how bull$#@! their argument about HEAT OUTPUT as an issue:

.18-micron
1.33GHz/XP1500+/Palomino = 60W
P4-1.4/Willamette = 55W
1.4GHz/XP1600+/Palomino = 63W
1.46GHz/XP1700+/Palomino = 64W
P4-1.5/Willamette = 58W
1.53GHz/XP1800+/Palomino = 66W
P4-1.6/Willamette = 61W
1.6GHz/XP1900+/Palomino = 68W
1.66GHz/XP2000+/Palomino = 70W
P4-1.7/Willamette = 64W
1.73GHz/XP2100+/Palomino = 72W
P4-1.8/Willamette = 66W
P4-1.9/Willamette = 73W
P4-2.0/Willamette = 75W

.13-micron
1.46GHz/XP1700+/TBRED-A = 49W
1.53GHz/XP1800+/TBRED-A = 51W
1.6GHz/XP1900+/TBRED-A = 53W
1.66GHz/XP2000+/TBRED-A = 60W
1.66GHz/XP2000+/TBRED-B = 61W
1.73GHz/XP2100+/TBRED-A = 62W
1.8GHz/XP2200+/TBRED-A = 68W
1.8GHz/XP2200+/TBRED-B = 63W
2GHz/XP2400+/TBRED-B = 68W
P4-2.0/Northwood = 54W
2.13GHz/XP2600+/TBRED-B = 68W
P4-2.2/Northwood = 57W
2.2GHz/XP2700+/TBRED-B = 68W
2.25GHz/XP2800+/TBRED-B = 74W
P4-2.26/Northwood* = 58W
P4-2.4/Northwood = 58W
P4-2.4/Northwood* = 60W
P4-2.5/Northwood = 61W
P4-2.53/Northwood* = 62W
P4-2.6/Northwood = 63W
P4-2.66/Northwood* = 66W
P4-2.8/Northwood* = 68W
* = 533fsb

You'd have to consider MHz ratings to show the Northwood as ALOT cooler than an Athlon XP - but AMD's PR-rated chips are so close to identical in comparison from Thoroughbred to Northwood that its nonsensical to say its any big difference. In the previous generations, Palomino and Williamette, the Athlon XP put out less heat for its rating of performance. Let's not forget that AMD fanboys can list the 1.53GHz Athlon XP against the 2.0GHz Pentium 4 because AMD marketed this level of performance head-to-head; AMD wins 51W to 54W with their .13-micron CPUs. Again, the difference is lost because the margin is so slim its stupid to argue the point.

Intel chips shown here clearly will put out the same heat as the load increases and will match the output of the Athlon XPs when comparing chips of the same performance rating. Since the data is maximum output and not typical or idle output, we have no idea what the true comparisons would be. Nor does AnandTech give us any idea what they use for measuring the test; do they just use the same program to test both CPU's although each CPU may test different depending on the programming used?

Its not how hot it can get, its how hot that it runs, that matters most. At idle, where our CPUs spend most of their time, the Athlon XPs are able to curtail their heat generation just like the Northwood. The points being made around here that Athlon XPs run too hot are pure FUD.

 

[RanDum72]

You'd have to consider MHz ratings to show the Northwood as ALOT cooler than an Athlon XP - but AMD's PR-rated chips are so close to identical in comparison from Thoroughbred to Northwood that its nonsensical to say its any big difference

So what do you propose, rate heat output according to PR rating? LOL, this just as bad as the PR rating itself. Heat measurements should be based on true clock speeds and based on your lists, a P4 2.4ghz has the same heat output as a 1.66ghz XP T-bred. The difference is almost 800mhz and that is a huge gap. Even if you base it on PR ratings, the 2400+XP (at 2.0ghz) still runs hotter.
You can argue that a few more watts here and there shouldn't reallymake a difference and it may be true. But in terms of arguing the point of which CPU runs cooler, the P4 runs much cooler, plain and simple. Even if you argue by PR rating (which is BS, IMO), the P4 is still cooler.

 

[MadRat]

<<But in terms of arguing the point of which CPU runs cooler, the P4 runs much cooler, plain and simple. Even if you argue by PR rating (which is BS, IMO), the P4 is still cooler.>>

I've already shown you that your statement is wrong; AMD cpus are often cooler than their Intel counterpart. When you measure the overall power draw of the system then the Intel system will end up comparing equal or worse to the AMD system built at a similar performance rating. Add in RDRAM or Dual-channel DDR, the best available RAM solutions for Intel's Pentium 4 system, and you make the Intel system a woodstove in comparison.

I've played around with both P4 and XP systems and they are both low-level ovens. The TBred cores reduce the load dramatically on the AMD system, though, just like Northwood with SDRAM is a nice cool system. AMD gets a bad rep because the Palomino cores are often used in the comparisons to Northwood, which is pure FUD. TBred has made AMD's heat generation a mute issue.

The figures cited by everyone in every comparative article seem to be based on maximum output. According to the AnandTech article it makes it sound like they ran the CPU's under full load but does not define how that was really achieved. Is it a virtual maximum? Is it the CPU under a certain load? Is it the figure that each company released?

We know that AMD chips are designed for a higher maximum temperature, a fact that you can verify from the respective AMD/Intel websites. We know that each CPU reacts differently to programming from the other, and some loads have different overall impacts on the heat generation in comparison to the other CPU designs. (Intel has software cooling - akin to RAIN or CPUCOOL - through logic in the core whereas you have to run a third party program to gain the same result from AMD processors.) We already know that Intel released their thermal characteristics figures in such a way as to undercut the figures released by AMD, although they really were disimilar tests therefore are a poor factor to use for comparison.

 

[RanDum72]

I've already shown you that your statement is wrong; AMD cpus are often cooler than their Intel counterpart

Where exactly is this? Here's an example from your post:

2GHz/XP2400+/TBRED-B = 68W
P4-2.0/Northwood = 54W

At the same clock speed, the P4 consumes 14W less.
Now we go with 'PR' rating:

1.66GHz/XP2000+/TBRED-A = 60W
P4-2.0/Northwood = 54W

even with the PR rating, the P4 consumes 6W less.

Now with the top dog CPU's:

2.25GHz/XP2800+/TBRED-B = 74W
P4-2.8/Northwood* = 68W

The P4 runs 600mhz faster but consumes 6W less power.

In comparison, a similarly clocked P4:
P4-2.2/Northwood = 57W

It consumes 17W less power.

Now, I'm not making these up; these figures are from your own post.

We already know that Intel released their thermal characteristics figures in such a way as to undercut the figures released by AMD, although they really were disimilar tests therefore are a poor factor to use for comparison

Yeah sure. The last Intel CPU I had before the P4 was back in the 486 days. I have used AMD CPU's almost exclusively ( I had some Cyrix CPU's)until recently and even I (without any instruments) could tell which CPU runs hotter. First, its noticeably quieter with the P4. Secondly, my room doesn't feel like an oven anymore regardless of whether the PC is running SETI or just sitting idle.

 

[brian_riendeau]

i switched to intel recently and would not go back to amd while amd and intel are priced almost the same in the mid -> upper range of the market. i strongly feel that intel provides the best chipset availible compared to the via offerings most people with amd systems use. through years of computer use, i feel that intel has always been a more troublefree solution ON AVERAGE. i used to go with amd solely based on price/performance. now that amd lost its price advantage, i went back to reccomending intel to anyone i know building a system i help with. if someone is on a tight budget, amd rules the roost with $50 and $60 ahtlon xp chip though.

 

[CraigRT]

Originally posted by: ai42
I agree with that, while AMD has been given all the press by us ethusiests lately. I have not been too impressed with AMD's recent offerings, and would much prefer to have a P4 right now.

They're not as fast.
I am die-hard AMD again.
I got a P4 1.8 which runs at just about 2.8 and I'd take an XP2400 over it any day.

I've had alot less luck with my P4 rig than any AMD system I've used in recent past.

overall: AMD gets the nod from me... love their stuff..

 

[Mrburns2007]

Current Recommendations:

1. Athlon 1600 and Epox dual channel DDR motherboard for budget system. (they can wait for the price to fall on Athlon 2400 or better and upgrade)

2. Athlon 2400 and Asus dual channel DDR motherboard for performance system.

3. P4 3060 with HT with dual channel DDR for money no object system.

 

[MadRat]

Originally posted by: RanDum72
Where exactly is this? Here's an example from your post:

2GHz/XP2400+/TBRED-B = 68W
P4-2.0/Northwood = 54W
At the same clock speed, the P4 consumes 14W less.
Now we go with 'PR' rating:

1.66GHz/XP2000+/TBRED-A = 60W
P4-2.0/Northwood = 54W
even with the PR rating, the P4 consumes 6W less.

You ignore everything I said about AnandTech's tests being impossible to use as a comparison because they give us no clue as to what was contained in the test. We'd have to know the contents of the test in order to judge its validity. And its not the XP2000+ that AMD put up against the 2.0GHz P4, it was the XP1800+. So by those numbers AMD won 51W to 54W. When Intel further released the 2.2GHz P4 then AMD continued to compare the XP1800+ to it and again AMD wins 51W to 57W when comparing .13-micron chips. If we want to go to the extreme and compare .18-micron chips its 66W (XP1800+) for AMD to the 75W (2.0GHz P4) for Intel.

Originally posted by: RanDum72
Now with the top dog CPU's:

2.25GHz/XP2800+/TBRED-B = 74W
P4-2.8/Northwood* = 68W
The P4 runs 600mhz faster but consumes 6W less power.

In comparison, a similarly clocked P4:
P4-2.2/Northwood = 57W
It consumes 17W less power.
Now, I'm not making these up; these figures are from your own post.

How thoughtful of you to quote me then misrepresent my statements. If the XP2800+ is AMD's top chip then it should be compared to the 3.06MHz P4, not the 2.8GHz P4. The XP2600+ and XP2700+ are the second tier of AMD, both running at 68W - the same level as the 2.8GHz P4.

 

[Justorq]

Hey guys,
I was just wondering why you people still say that the 3.06GHz intel has a "clear" performance lead ??
Just look at some benchmarks and you can "clearly" see that the 2800+ keeps up pretty well with that proc. It beats it in some occasions too ...
I really dunno where you guys get your facts from ... you could build a very stable system with an AMD chip very easily. Even with the retail HSF, the temps for the system i built with the 1700+ where hovering around 45C.
cmon... you people are intelligent ... capable of having good discussions ... but don't go on telling things, cuz it just isn't right anymore.

Justin

 

[tbates757]

Originally posted by: Dr Smooth
I just checked Newegg and they are selling the retail P4 2.8 gHz for $389 and the OEM AMD XP 2800+ for $399. So the AMD chip costs at least $30 - $40 more. Though I am sure the XP2800+'s price will continue to go down.

wait a second, isnt that $10 difference? how do you figure 30-40? lol, huh?!?

 

[Strych9]

Originally posted by: BD231
Intel offers a more advanced cpu so it's obvious it's a better choice if your building a computer from scratch

 

[Siddhartha]

Originally posted by: tbates757

Originally posted by: Dr Smooth
I just checked Newegg and they are selling the retail P4 2.8 gHz for $389 and the OEM AMD XP 2800+ for $399. So the AMD chip costs at least $30 - $40 more. Though I am sure the XP2800+'s price will continue to go down.

wait a second, isnt that $10 difference? how do you figure 30-40? lol, huh?!?

The P4 is a retail unit, meaning it comes with a heatsink\fan. The XP2800+ is an OEM unit, which does not come with a heatsink\sink. I figure a heatsink\fan will cost $20 to $30, though I am sure you could pay more for one.

 

[MadRat]

We'll have to wait for the XP2800+ availability to increase before saying its more expensive.

On the other hand, if Intel has the top dog they should be able to command top dollar. Why don't they?

 

[Siddhartha]

Originally posted by: MadRat
We'll have to wait for the XP2800+ availability to increase before saying its more expensive.

On the other hand, if Intel has the top dog they should be able to command top dollar. Why don't they?

Well, Newegg is selling the 3.06 gHz P4 for $830.......

 

[MadRat]

$830. LOL. I stand corrected - that is commanding top dollar!

 

[Siddhartha]

Originally posted by: Dr Smooth

Originally posted by: MadRat
We'll have to wait for the XP2800+ availability to increase before saying its more expensive.

On the other hand, if Intel has the top dog they should be able to command top dollar. Why don't they?

Well, Newegg is selling the 3.06 gHz P4 for $830.......

And Newegg is selling the XP2800+ for $399.....