• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Read, if you plan to buy new rig soon. AMD vs INTEL

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

NoxioN

Member
Joined
Jun 1, 2003
AMD VS INTEL is the performance worth the price? Which system runs hotter? I hope this thread will help explain a few debates currently going on this forum and will aid you in your decision when picking out a system to build. Please no flaming! I am not bias toward either companies and I am happy with both systems.

Rig 1 $200 chip board combo
DLT3C XP1700 JIUHB AMD CPU
ABIT NF7-S version 2
SLK900u with ~50cfm 92mm fan
Aluminum case with sufficient ventilation

Rig 2 $375 chip board combo
2.4c Intel 800fsb CPU
ASUS P4C800 Deluxe
SLK900u with ~50cfm 92mm fan
Aluminum case with sufficient ventilation

First we look at performance, both rigs are overclocked about 1 GHz and tested. The Intel rig is clearly the winner and don't even comment on the memory bandwidth we all knew that would happen. Not much more to say about performance but then again the Intel rig cost $175 more to build so if it did not outperform the AMD then we would not be buying it. If the Spingdale board was used then the price would be only $75 more for the Intel rig. If the AMD board was an nforce Epox 8RDA then that would lower the price of the AMD rig by $35. The chip and board combo for both are: AMD $190 and INTEL $375
The prices and board options have come down so if you wanted to buy now you could go with an 8RDA for $85 1700+ for $65, 2.4c OEM $175 IS7 abit $100.

Here are some pictures on performance when overclocked about 1 GHz.

AMD DLT3C @ 2.51 GHz XP core

INTEL 2.4c @ 3.4 GHz

So if you're playing to get a new rig and can't decide then this might help. You now know what it would cost for the board and chip combo and which one performs better and what speeds. A 2.4 GHz amd is = to a 3 GHz 533bus Intel, according to Sandra MIPS. If you go the cheapest route for both rigs, it's not guaranteed to hit the same speeds but with some work and maybe a mod you may get similar results.
What will have the biggest impact on system performance and stability is Motherboard, CPU, RAM, PSU and Temps. Skimping on a generic 250w PSU might get you wacked rails and an unstable system so please do some research and think about what parts you are using. Not all chips are created equal, some weeks, Plants, and steppings will outperform others. When a system is unstable at a certain FSB no matter how much voltage you give it and you feel it can do more please check what speed you're ram is running at or try to find the best known bios. The forum is full of great people and information; please familiarize your self with the search button :) as well. GL with whatever Rig you decide to purchase.

Both systems are often talked about and since I have both rigs I know where you guys are coming from. I love both rigs and each has their advantages. The intel rig with HT is a little bit smother with multitask obviously but as far as gamming and performance they are almost identical. I can tell the intel is a little faster but not jaw dropping faster. The only ways I can truly tell is with benchmarks but from everyday surfing or gamming both rigs are similar. I am truly pleased with both rigs and don't regret anything. The only thing I would say is if you are planning to spend more then $175 for a CPU then definitely buy intel 2.4c because the 2800+ can not beat the 2.4c overclocked. The only AMD CPUs I would purchase are the ones under the $150 range, anything above that you just can't beat intel. But anything under $150 for intel just can't beat AMD.





TEMP:

A lot of talk about AMD runs hotter then INTEL. I have ran some test at different speeds and voltage for my rigs and here are the results. The AMD TB core at 1.5v default runs cooler then the Intel northwood D1 chip at 1.525v default. The only way the AMD will get hotter is at a higher vcore like .1v more.

The asus overvolts a little and at load it drops to about default vcore set in bios, the abit board does just the same.

amd @ defuatl core and speed

intel @ default core and speed


Clearly the AMD at default core and speed idles lower then the intel.


amd @ load default core and speed

intel @ load default core and speed

Clearly the AMD at default core and speed even at load has lower temps then the intel.


amd idle @ 1.6v 2170 mhz

amd at load temp 1.6v 2170mhz

intel temp at default speed at 1.6v

intel load temp at 1.6v default speed

Now with a bump in vcore set to 1.6v in bios for both rigs the amd even with a 700 MHz overclock the temps are still lower then the intel at default speed with 1.6v


amd at 1.8v 2400 MHz idle

amd at 1.8v 2400 MHz load

intel at 1.6v 3480 MHz load


intel at 1.6v 3480 MHz idle

Now with both rigs overclocked with a decent MHz and decent vcore we see closer idle temps. The 1.8v for the amd will idle slightly higher then the intel at 1.6v but at full load will be much cooler then the intel rig. This concludes my test with temps and vcores. The amd appears to run cooler even with a little more vcore. Both rigs were idling for 10 mins before I took reading and same time for prime. The test would be better at 2hrs but 10min should give you an idea of results.
 
Last edited:
superb post!! this should be a great guide for everyone contemplating a new rig.

sticky?
 
As usual Noxion, your threads are well thought out, and unbiased.
Seems everyone is flip-flopping between Intel and AMD these days, so this offers a great amount of information to those considering a move to either one.
 
Very nice post....I'm just curious. Those P4 numbers are higher in all those benchmarks..... but I'm just curious how that equates into percentages. I agree with you. I think that if you ran those temperature tests a little bit longer, you'd probably get better results. I'd like to see more of our members do a test like this if they have both AMD and Intel rigs. I think this would provide our members with a great comparision chart for future rigs. This could be the beginning of a nice little data thread like there are in the Video fourms. I think this could possibly be made a sticky if more people posted their results. Cheers and :thup: :thup: :D


*edit* - Doh! yall beat me to the first response!:p
 
thanks matt and congradulation on your marrage :) Im going to a my cousins wedding on the 5th of next month myself.
 
One little question......ALL P4s 2.4GHz D1 will do an average off 3300MHz.........DO ALL of 1700+ do an average of 2400MHz.???......Thanks..... :)
 
hipro5 said:
One little question......ALL P4s 2.4GHz D1 will do an average off 3300MHz.........DO ALL of 1700+ do an average of 2400MHz.???......Thanks..... :)
NO. my D1 2.4c does 290fsb 3480mhz. the only xp1700 that does 2.4 on average is the DLT3C JIUHB that can be bought at xcaliber or SCV other then that most 1700 does 2.1-2.3 range.

its like the magic C1, the only 1.8a that can do 3.4ghz were the C1 and the only SL6RZ that can do 3.4-3.6 were the costa rica good weeks and some malays

the DLT3C JIUHB really is an amazing chip for $65 and so is the week 52 SL6RZ costa rica.

I wouldnt say all, i have seen some threads saying why is my crappy 2.4c only at 2.9ghz? I haven't labelled it as a bad chip just yet cause I dont know the level of overclocking knowledge the user has but some people are having trouble with their D1s so don't say all.
 
I have never seen a single 1700+ DLT3C not do 2.4ghz that was limited by cooling or voltage options. Most, in fact, seem to be doing past 2.5ghz.
 
Gautam said:
I have never seen a single 1700+ DLT3C not do 2.4ghz that was limited by cooling or voltage options. Most, in fact, seem to be doing past 2.5ghz.

Stable?????........For 24/7 or just for screenshots and benchies???
 
hipro5 said:


Stable?????........For 24/7 or just for screenshots and benchies???
yes prime stable, i can prime it all night at 2.4ghz if you wish :)
 
Gautam said:
I have never seen a single 1700+ DLT3C not do 2.4ghz that was limited by cooling or voltage options. Most, in fact, seem to be doing past 2.5ghz.
you honestly don't believe that do you?
 
I meant "wasn't" in case that's what you were referring to, and in that case, no I have not. Some motherboards only allow voltages to around 1.8v or so, whereas most steppings require in the ballpark of 1.9 in order to do 2.5. A few worse ones, like mine, need 1.85 to do 2.4ghz. Most need only around 1.7. With that said, any 1700+ revision B DLT3C will be able to do 2.5ghz as long as they have a decent power supply(even my $27 Fortron will do,) a motherboard that allows a large range of voltages(Abit NF7 series) and a heatsink with a thermal resistance of .25 C/W or lower.($20 Thermalright SK7) While I'm sounding specific, most people have similar combos. As long as these limiting factors are removed, 2.5ghz is a breeze.
 
1700's at 2.4GHz I believe, but doing 2.5 and beyond is a stretch.. Most 1700 DLT3C's are doing between 2.2 and 2.3GHz., or if you're throwing lots of voltage at it, 2.4 is possible.
 
Clearly you are misinformed. "Lots of voltage" is only around 1.85v or so for the worst DLT3C's to do 2.4ghz. The better ones achieve this with only around 1.7v. 2.5ghz needs over 1.9v 99% of the time. This is a high voltage, but hasn't posed any problems thus far for those who use it, as long as they have high quality power supplies and midrange to high end heatsinks. I'm trying to be reasonable, but saying that most 1700+ DLT3C's will only do 2.2-2.3 is like saying that most 2.4C will only do 2.9-3.0ghz, both of which are entirely false statements. As long as all limiting factors are considered and dealt with, 2.5ghz is easy for 1700+ DLT3C's as is 3.4ghz for 2.4C's, at least AFAIK; I don't know much about Intels, so I dont want to make any false assumptions. Perhaps you guys should take a peek in the AMD forums if you really believe that 2.5ghz is a big deal. It has become pretty much the norm.
 
Gautam said:
With that said, any 1700+ revision B DLT3C will be able to do 2.5ghz as long as they have a decent power supply(even my $27 Fortron will do,) a motherboard that allows a large range of voltages(Abit NF7 series) and a heatsink with a thermal resistance of .25 C/W or lower.($20 Thermalright SK7) While I'm sounding specific, most people have similar combos. As long as these limiting factors are removed, 2.5ghz is a breeze.
i try not to make blanket statements like these because there is always an except to the rule.

fyi your 2.5ghz may be a breeze but i guarantee you its not for most 1700+ DLT3C users.

btw i just noticed you are on water cooling. therefore your results shouldn't be considered at all. 2.5ghz is alot easier if you have cooling that is 10-15 degrees better than air cooling.
 
The has been one, and only one person that has reported not being able to cross 2.3ghz stable. Everyone that buys a 1700+ DLT3C is essentially guaranteed 2.4ghz. Those with higher end heatsinks can reach 2.5ghz. Its as simple as pumping in the voltage. There is no way I'm going to convince you, so again I'll draw upon my analogy. How would you respond if I said that getting 3.1ghz is a rarity for 2.4C's?

EDIT: I remember of one rare stepping that performed quite poorly in comparison to the rest, and only one. The code was 0312TPXW, if anyone's interested. These were purchased from GameVE.com, as opposed to Excaliberpc.com or svc.com, where most people buy them. I hardly think that this single outlier can be used against my generalization of 2.5ghz not being difficult.
 
I'm running an AMD setup right now, but not after tomorrow.

I'm currently using a Soltek SL-FRN2-RL, with an SLK-800a and YS-Tech 45cfm 80mm fan and a 2700+.

I got my 2700 athlon to 2.4GHz, but that's only after I increased the voltage to 1.8v, which is higher than I'm comfortable with using air cooling - water cooling is entirely different.

Right now I'm at 11 x 210 for 2.3GHz, but that's with 1.7v - a much more temp friendly setting for me personally. -Idle temp is 36c vs about 41c with 1.8v.
 
Well I really would like to know what people are using to get the temps for the above tests. Did you do the test with a hole drilled to the center of the HS like Joe C. does

Just cause a chip runs cooler than another doesn't mean its heat output is lower.
 
However, your combo clearly allows for 1.9v and 2.5ghz. You are the limiting factor :D. I understand where you're coming from though. 1.9v is certainly a high voltage to be using on either are or water. However, I haven't seen any Thoroughbreds have any issues with voltages such as these. 2.1v is very dangerous, but you wont be getting SADS with only 1.925v. It will reduce the life of your processor, however most likely not by enough to even effect you; you would've thrown out your rig years before the processor's death. If you bit the bullet and used 1.925v or so, 2.5ghz wouldn't be a problem. Many do, and I've repeatedly said that this amount of voltage will be necessary, but will still remain 100% stable with an SLK800.
 
Back