Latest Benches from China - Conroe vs FX62

[UglyChild]

Here are some becnhes of Conroe XE vs FX-62

http://www.hkepc.com/hwdb/x6800vsfx62-10.htm

[Bios24]

holy Super-Pi batman, 17 seconds?! Two please :)

It's obvious the Conroe is awesome, but I'm eager to see the price line-up next to the A64/FX/Opteron series. That's what it'll come down to for me.

[Albaholic]

:drool: I want a conroe

[Recursion]

Looks good to me. Cant wait

[w00tstock]

Cherry picked, All i can say about the previous benchmarks i saw. I think its amazing how conroe when from 15% magin of victory with a slower clock, -200mhz to maybe 15% with a 133 mhz advantage. Was going to hold off on an am2 system to see what conroe had. Going amd now. I just think that AM2 system at 65nm and core improvement will be faster than conroe, not to mention k8l. Also if you look at the temp reading it seems as if conroe runs quite hot for only 75w, could be the cooler but i dont know why an after market cooler would be worse than the intel sock one.
I am very disapointed, but atleast intel sparked a price war.

[snvpa]

^^^^^ Are you serious? so beating the top end FX-62 that costs 1K with a speed that the $330 e6600 can achieve without any effort isn't good enough? What do the AMD fanboys want? What is it that you are dissapointed with, I'm just curious?

[Krowa 02]

^^^^^ Are you serious? so beating the top end FX-62 that costs 1K with a speed that the $330 e6600 can achieve without any effort isn't good enough? What do the AMD fanboys want? What is it that you are dissapointed with, I'm just curious?

I think what he means is that the price cuts by amd will give a better price to performance ratio, especially when 65nm amd chips come out. AM2 will be around for awhile, K8L will probably work with most am2 boards so you can buy now and upgrade later.

Even so, its clear that conroe eats amd right now.

[shaking_ground]

Cherry picked, All i can say about the previous benchmarks i saw. I think its amazing how conroe when from 15% magin of victory with a slower clock, -200mhz to maybe 15% with a 133 mhz advantage. Was going to hold off on an am2 system to see what conroe had. Going amd now. I just think that AM2 system at 65nm and core improvement will be faster than conroe, not to mention k8l. Also if you look at the temp reading it seems as if conroe runs quite hot for only 75w, could be the cooler but i dont know why an after market cooler would be worse than the intel sock one.
I am very disapointed, but atleast intel sparked a price war.

while i disagree that the results of conroe are dissapointing, it could be well argued that for 99% of what a normal person does with a computer, a $149 a64 3800 on a $90 board would be a way more intelligent choice than sinking $300 on a proc and $250 on a board.

His point is valid. AMD will still be a great choice for a budget conscious system that still wants good performance.

Though when 930's drop down to 160 or so in july, it will be a very very very affordable processor......which i plan to buy!

[hUMANbEATbOX]

there will be Conroes that are MUCH cheaper than $300.

there will be decent mobos for WAY less than $250.

not to mention that for a budget systems, the 9xx line will still be around, don't forget the 805 (905 coming?) that will be less than $90.

[]-[itman]

Looks to be around a 10%-15% advantage on average for Conroe, hmmm, who could've predicted that ;)

Anyway, the thing I like about Conroe is two fold,

1)Much cooler running chip for the performance than anything available now, AMD probably won't have an answer for this until 2007 (in large supplies anyway). It will be interesting to see how Conroe yields turn out, some say they are strong, but intel keeps sticking to its 20% by 07, which isn't great, but could be to try and sell off a lot of previous netburst stock.

2)The clocks this chips can reach on air, I mean I've seen 3.4-3.6ghz on air, and this is the first shipping revision, will be interesting to see how good these things can get. Although I do expect 65nm A64's to match this, possibly beat it, again, we won't see them available until 07 and a new Conroe spin may be able to match that.

[shaking_ground]

there will be Conroes that are MUCH cheaper than $300.
there will be decent mobos for WAY less than $250.
not to mention that for a budget systems, the 9xx line will still be around, don't forget the 805 (905 coming?) that will be less than $90.

Oh yeah, i know the prices on the upcoming line, don't get me wrong. All i was saying was that AMD will still offer a very powerful product for the price. a nice budget AMD gaming core (mobo, chip) could be had for a mere $200 bucks if you look around. For people with zero affinity towards either company, who are on a budget, who like gaming, who dont use dual core, may still see AMD as a very competitive company.

I'm personally an intel guy, and plan to stay one, but putting myself in a budget minded non biased position, i could be pursuaded either way, that's all.

-[itman']Looks to be around a 10%-15% advantage on average for Conroe, hmmm, who could've predicted that ;)

Anyway, the thing I like about Conroe is two fold,

1)Much cooler running chip for the performance than anything available now, AMD probably won't have an answer for this until 2007 (in large supplies anyway). It will be interesting to see how Conroe yields turn out, some say they are strong, but intel keeps sticking to its 20% by 07, which isn't great, but could be to try and sell off a lot of previous netburst stock.

2)The clocks this chips can reach on air, I mean I've seen 3.4-3.6ghz on air, and this is the first shipping revision, will be interesting to see how good these things can get. Although I do expect 65nm A64's to match this, possibly beat it, again, we won't see them available until 07 and a new Conroe spin may be able to match that.

AGREED, these new chips will yield ungodly performance if you are willing to sink about.....800-1k on a whole new system.

[greenmaji]

Looks to be around a 10%-15% advantage on average for Conroe, hmmm, who could've predicted that

Anyway, the thing I like about Conroe is two fold,

1)Much cooler running chip for the performance than anything available now, AMD probably won't have an answer for this until 2007 (in large supplies anyway). It will be interesting to see how Conroe yields turn out, some say they are strong, but intel keeps sticking to its 20% by 07, which isn't great, but could be to try and sell off a lot of previous netburst stock.

2)The clocks this chips can reach on air, I mean I've seen 3.4-3.6ghz on air, and this is the first shipping revision, will be interesting to see how good these things can get. Although I do expect 65nm A64's to match this, possibly beat it, again, we won't see them available until 07 and a new Conroe spin may be able to match that.

We've seen nearly 5Ghz benched conroe's do you expect 64nm AM2's to be able to match that +10-15%

btw. you did mention the 10-15% several times before.

And 20% of what yeilds? Total yeilds? Or desktop processor yeilds? No one has answered that question when Ive asked it before.

[w00tstock]

Umm well considering they compared the 1000$ intel proc to the simularly priced amd one. Remeber they compared the x6800 EE. PLease take your intel fanboyisms elseware, I based my decisions based apon upgrade path, price and performance. Sure the x6800 beat the fx, but the fx is an eol chip the x6800 isnt. People in other forums have worked out the difference clock for clock to be between 5 and 6 %. I am taking the stance that amd will be able to gain both clock rates, and performace per clock in their next upgrade, ie 65nm, and that these changes will be more than enough to make amd the better value.

Also on a side note show me a fx 62 and ill show you 3.6ghz too. Show me a 165 and I could show you 3ghz. My point is why the heck are you comparing what a chip could overclock to with out considering what the competitions lower end chips can get to as well?

[shaking_ground]

Umm well considering they compared the 1000$ intel proc to the simularly priced amd one. Remeber they compared the x6800 EE. PLease take your intel fanboyisms elseware, I based my decisions based apon upgrade path, price and performance. Sure the x6800 beat the fx, but the fx is an eol chip the x6800 isnt. People in other forums have worked out the difference clock for clock to be between 5 and 6 %. I am taking the stance that amd will be able to gain both clock rates, and performace per clock in their next upgrade, ie 65nm, and that these changes will be more than enough to make amd the better value.

Also on a side note show me a fx 62 and ill show you 3.6ghz too. Show me a 165 and I could show you 3ghz. My point is why the heck are you comparing what a chip could overclock to with out considering what the competitions lower end chips can get to as well?

Fair enough, a comparison between a 165 oc'd on air hasn't been done (to my knowledge) against a 6400. Now, in all fairness, how do you think said test will come out. If you are willing to spend 300 bucks on a chip and between two chips, one gives you say +10% performance....no offense but that's the one i'll buy. That isnt fanboyism, that's smart consumerism....and hey, if you want to waste your money on an inferior product, be my guest. I've done it before, i can't fault you:beer:

Seriously though man, considering i tried to be VERY fair minded, you calling me a fanboy makes me pretty angry. why don't you go do something useful.

[w00tstock]

I think we all know the intel will win, but the thing that gets me is that amd will have a upgrade path with actual core improvements in addition to clock speed advances. No i odnt think your a fan boy I think svnpa is.

[greenmaji]

AMD will have an upgrade path for a ageing chipset mobo, if you want the top performance out of K8L (it won't be a cheap processor so I surely would want a kick a$$ mobo for it if I was to buy) the same holds true for 65nm AM2 but the chip wont be as much so I could see putting it on an old mobo as an option a bit better. In other words, why not get whats best when you want to upgrade (I said want not you have to upgrade when something killer comes out ;) )
btw.. this line of discussion is for the general CPU section :)

[hUMANbEATbOX]

I think we all know the intel will win, but the thing that gets me is that amd will have a upgrade path with actual core improvements in addition to clock speed advances. No i odnt think your a fan boy I think svnpa is.

he's not a fanboy either. a e6600 will do between 3.6-4ghz on air. show me any AMD chip that can do the same.

[greenmaji]

he's not a fanboy either. a e6600 will do between 3.6-4ghz on air. show me any AMD chip that can do the same.

Or an instance when AMD shrunk there fab and the chips clocked that much better!!! :mad:

Im getting tired of hearing that 65nm chips are going to clock like mad demons, history says they won't!

K8L on the other hand is a differnt story, quite a few nice arch. upgrades :D

But 45nm CSI cores from Intel will be out soon after K8L!!! :eek: It looks like Intel has really got the ball rolling this time :drool:

[ochungry]

Here are some becnhes of Conroe XE vs FX-62

http://www.hkepc.com/hwdb/x6800vsfx62-10.htm
Thanks for the link.
I averaged out the gaming benchmarks @ 1600x1200, ratio'ed in the 2.8ghz Fx62 vs 2.93ghz Conroe's EX and the result is 7% better performance in favor of Conroe. Although, this is a great achievement by Intel's new Core architecture, the performance is not as great as we have led to believe. Certainly not 40%, or even a 10% better than FX every one were claiming.
What it comes down to is: When the conroe is released to retailers/etailers, and testing matured in a month or 2 after release, one can decide which platform serves him/her better than the other.
If money is not an issue you will choose the best money can buy, which is Conroe for time being. If you are on budget and price/performance is the key, IMO, AMD will give you better deal. You can always buy a 3800 x2 and a good AM2 mobo for about $300 and overclock it to an FX level, and be done w/ it, . 7% or 10% better than the other in today's computing is dismal because your machine does any job just as good.
But If you are in WR and care to show off your system, you can spend as much as you can afford (or go hungry) and buy a conroe system and do SP1 1M 24/7 on DICE- No AMD fanboy can touch it, not till K8L. Even then no one knows if that's going to do it or not.
I am glad Intel can compete w/ AMD (or the other way around), because it creates price war and both companies will work harder for our money.
Let's just hope that both companies have both good(s) and not so good(s). This way we can decide what is good to buy and what's not depending where you fit-in as described above.
Furthemore, take what said by me only as an "opinion", and let your own logic resolve.

[MikeyLikesItSI]

the result is 7% better performance in favor of Conroe. Although, this is a great achievement by Intel's new Core architecture, the performance is not as great as we have led to believe. Certainly not 40%, or even a 10% better than FX every one were claiming.


I think what it REALLY comes down to is end user. How this chip feels to for every day use. avg of 7% in gaming benchmarks definitly foreshadow performance between the 2, but that is such a small fraction of what a computer can actually do. besides, if you really want killer game performance why not save the extra money on the chip/board combo and spend it on CF or SLI?

I think where conroe is super impressive is when comparing it to its netburst counterpart. it REALLY blows that out of the water!! and that right there shows intel is bettering itself between power consumption, and performance gains.

[snvpa]

No doubt AMD still has a competitive offering especially with the price cuts. I guess fanboy might be proper since I do like intels new offering but my current rig and the past 5 have all been amd's because they offered competitive performance for less money. Thats price to performance ratio, intel is the better choice this time around. AMD is only dropping prices on the small 512K cache chips and 512x2 dual cores, the 2x1meg chips will be dropped (FX excluded) and the opties will remain the same price.

http://overclockers.com/tips00980/

So an AM2 165 optie will be the same price as the E6600 roughly $300. The E6600 by all accounts should hit 3.4 on air or water with little effort. Show me a 165 that can hit that at all. Not to mention the conroe will still be faster clock for clock. While AM2 might provide a good upgrade path at 65nm and K8L I have not seen any definitive statement that current boards will work nor do I have much faith in AMD's ability to get either out on time.

Finally while a 150$ 3800x2 sounds pretty damn good a 130$ 930 is an equally attractive budget upgrade path. I must say W00tstock at least you back your argument up with concise and logical statements I can't fault your logic just a difference in opinion. :)

[ZGOZZ]

The Prices for the Xtream Conroe CPU Box set for Intel resalers will be under 1000 dollars.

So it should be in the high 950 ish area....
This is coming from my Intel Rep..

[traldan]

Mmm, can't wait to get my hands on a Conroe. :D

[ochungry]

No doubt AMD still has a competitive offering especially with the price cuts. I guess fanboy might be proper since I do like intels new offering but my current rig and the past 5 have all been amd's because they offered competitive performance for less money. Thats price to performance ratio, intel is the better choice this time around. AMD is only dropping prices on the small 512K cache chips and 512x2 dual cores, the 2x1meg chips will be dropped (FX excluded) and the opties will remain the same price.

http://overclockers.com/tips00980/

So an AM2 165 optie will be the same price as the E6600 roughly $300. The E6600 by all accounts should hit 3.4 on air or water with little effort. Show me a 165 that can hit that at all. Not to mention the conroe will still be faster clock for clock. While AM2 might provide a good upgrade path at 65nm and K8L I have not seen any definitive statement that current boards will work nor do I have much faith in AMD's ability to get either out on time.

Finally while a 150$ 3800x2 sounds pretty damn good a 130$ 930 is an equally attractive budget upgrade path. I must say W00tstock at least you back your argument up with concise and logical statements I can't fault your logic just a difference in opinion. :)
I dont think you can persuade anyone to go for netburst anymore.
There is a poll @ Hexus.com regarding the likelihood of people buying which and so far: 25% for conroe, 34% for AM2, and 0% for netburst. About 27,000 voted last time I checked.

[Cheator]

Or an instance when AMD shrunk there fab and the chips clocked that much better!!! :mad:

Im getting tired of hearing that 65nm chips are going to clock like mad demons, history says they won't!

K8L on the other hand is a differnt story, quite a few nice arch. upgrades :D

But 45nm CSI cores from Intel will be out soon after K8L!!! :eek: It looks like Intel has really got the ball rolling this time :drool:

Show me a newcastle that clocked 300mhz above stock without some good cooling. Winchester's die shrink was a HUGE improvement. I got 2.5 ghz on a 3200+ when people were getting something like 2300 mhz on the same 130nm cpu. And then Venic came out, blowing away the old overclocks. 2.7 ghz on stock volts with stock cooling, Opteron 165s doing the same (with 2 cores, that aint bad at all), etc. It made a huge difference, much like it did for Intel. Hell, 90nm to 65nm for them was a very big step. 5 ghz celeries to be exact.

AMD's die shrik will again show more and better overclocks, simply because its inevitable. Less space, less heat, more performance. I look forward to seeing what kind of performance we can get out of it, and am waiting for them to come out this winter (Late Q4). Thats the time to switch to AM2.

[]-[itman]

Or an instance when AMD shrunk there fab and the chips clocked that much better!!! :mad:

Im getting tired of hearing that 65nm chips are going to clock like mad demons, history says they won't!

K8L on the other hand is a differnt story, quite a few nice arch. upgrades :D

But 45nm CSI cores from Intel will be out soon after K8L!!! :eek: It looks like Intel has really got the ball rolling this time :drool:

http://theinquirer.net/?article=32322

rumored up to 50% improvement clock rate over current 90nm A64's. Granted, this is probably a bit too hopeful, but let's take that number and make it a little more realistic, say 35%-40%, that puts 65nm A64's at 3.7ghz-3.8ghz on air, plenty to atleast match Conroe in pure speed. Also, according to current roadmaps, intel and AMD plan on releasing 45nm chips less than a year away from each other, but you never know, this may change.

[Krowa 02]

Put it this way, conroe is the way to go for all around computing. It dominates in dvd encoding, shrinking, audio econding picutres etc. Thats stuff that im looking for as well as gaming. I am an all around computing type person, so its a very tough choise for me to make.

[AdvanS13]

We've seen nearly 5Ghz benched conroe's do you expect 64nm AM2's to be able to match that +10-15%

btw. you did mention the 10-15% several times before.

And 20% of what yeilds? Total yeilds? Or desktop processor yeilds? No one has answered that question when Ive asked it before.

as far as conroe/merom yields, currenty-even, if not better in some cases then cedar mill and yonah...no worries there, it's arse kicking in the fab:drool:

[w00tstock]

The question about yields is interesting because intel was supposed to release a 3.3 ghz core instead of the 2.93ghz one. What happened to those 400 mhz? Yields or reserve ammo?

[traldan]

The question about yields is interesting because intel was supposed to release a 3.3 ghz core instead of the 2.93ghz one. What happened to those 400 mhz? Yields or reserve ammo?

Reserve ammo methinks.

[Epox4life]

How much will the FX-62 cost when conroe is released? 600$?

[hUMANbEATbOX]

no i don't think there is a price cut on the fx62.

[traldan]

no i don't think there is a price cut on the fx62.

With results like this, AMD sure as hell better drop prices on it.

[Evilsizer]

here are a few things missed. conroe hitting 4ghz on air older step4 took 1.7v and a newer step 5 took 1.3v.
step 4
http://www.xtremesystems.org/forums/showthread.php?t=101796
step 5 (retail stepping)
http://www.xtremesystems.org/forums/showthread.php?t=103381

alittle 975 vs 965 action
http://www.xtremesystems.org/forums/showthread.php?t=103748

[]-[itman]

here are a few things missed. conroe hitting 4ghz on air older step4 took 1.7v and a newer step 5 took 1.3v.
step 4
http://www.xtremesystems.org/forums/showthread.php?t=101796
step 5 (retail stepping)
http://www.xtremesystems.org/forums/showthread.php?t=103381

alittle 975 vs 965 action
http://www.xtremesystems.org/forums/showthread.php?t=103748

Actually the stepping 5 took 1.6v still, cpuz isn't reporting the correct voltage. These are also not very stable, I mean, most people have issues with pushing 1.6v on 90nm chips with only air, let alone 65nm. Don't get me wrong, very cool to look at, but in the end meaningless until you can get them stable and within a reasonable voltage.

[Evilsizer]

-[itman']Actually the stepping 5 took 1.6v still, cpuz isn't reporting the correct voltage. These are also not very stable, I mean, most people have issues with pushing 1.6v on 90nm chips with only air, let alone 65nm. Don't get me wrong, very cool to look at, but in the end meaningless until you can get them stable and within a reasonable voltage.
dont know what you mean by reasonable since, since these are not netburst. As of right now what the higherest voltage we can run 24/7 is still in the air. Im still betting 1.55v on high end air and 1.6-1.65 on water 24/7. Time will tell.

[greenmaji]

Isn't temperature playing a big part on the high V. His temps are two degrees above ambiant on air 77F/25C. Are high volts that big a deal at 25C???? :confused:

BTW. no one asked about Vdroop and how high the V is on load and at idle. Maybe the V is high from fighting Vdroop??

[Evilsizer]

Isn't temperature playing a big part on the high V. His temps are two degrees above ambiant on air 77F/25C. Are high volts that big a deal at 25C???? :confused:

BTW. no one asked about Vdroop and how high the V is on load and at idle. Maybe the V is high from fighting Vdroop??
well the bad axe has a problem reading temps from conroe correctly, probly needs bios update. IT also has a powerslope feature that reduces Vdroop by a large amount. weither we see this on other mobo's we can only wait and see. What you set in the bad axe doesnt vary much with powerslope so i would stab and say maybe .05 v drop underload maybe less. We would have to ask ross if a proggy like mbm5 or speed fan supports the bad axe.

*edit*
here is something else that was a nice comparesion.
http://forums.guru3d.com/showthread.php?t=185555

[]-[itman]

Isn't temperature playing a big part on the high V. His temps are two degrees above ambiant on air 77F/25C. Are high volts that big a deal at 25C???? :confused:

BTW. no one asked about Vdroop and how high the V is on load and at idle. Maybe the V is high from fighting Vdroop??

I believe 77F is the temp of his room, not his cpu, there's no way in heck the chip is producing 1C worth of heat, only if the chip was idle would this begin to be possible, and even then it depends on how idle the chip really is. I doubt the vcore is that high to combat vdroop as it still is not stable, he can't finish a 32m super-pi on air, I've done that with systems which will crash after 5 min of actual game playing.

[>HyperlogiK<]

We've seen nearly 5Ghz benched conroe's do you expect 64nm AM2's to be able to match that +10-15%

We have to differentiate between how good the chips are, and how good the chips are for US. Conroe smokes K8, however it doesn't destroy it as completely as I had expected at stock speeds. It is clear that anybody wanting a serious high end OCing chip right now should steer clear of anything AM2. Fortunately however, the results do not suggest any kind of intel monopoly (especially considering the 'hopefully' ongoing AMD price cuts), because if the results for the FX62 had 1.5x the FPU performance and a few other tweaks they would probably be about even overall, at least for gaming, and Battlefield 2 matters more than SuperPi. I just don't know if even at 65nm with Si-Ge chips whether AMD is going to be able to match that power consumption/overclockability, other than maybe as cherry picked cores.

[greenmaji]

-[itman']I believe 77F is the temp of his room, not his cpu, there's no way in heck the chip is producing 1C worth of heat, only if the chip was idle would this begin to be possible, and even then it depends on how idle the chip really is. I doubt the vcore is that high to combat vdroop as it still is not stable, he can't finish a 32m super-pi on air, I've done that with systems which will crash after 5 min of actual game playing.

He posted room temp 75F and proc. temp. 77F. thats with a monster heat pipe air cooler, Is that possible? Is it BS? I don't know the poster well enough to give an answer to that. And as posted before, aparently the badaxe has issues reporting CPU temps correctly.
I've seen some strangely high (1.6V+) Vcore's to fight Vdroop with Asus+pressler systems (Vdroop mods fixed this) thats why I asked that question.

Battlefield 2 matters more than SuperPi

3dmark scores are higher with Core2 as well http://smiley.onegreatguy.net/read.gif http://smiley.onegreatguy.net/happydrunks.gif

[UglyChild]

Price list at the bottom of the page.



http://www.hkepc.com/hwdb/x6800vsfx62-7.htm

[AdvanS13]

We have to differentiate between how good the chips are, and how good the chips are for US. Conroe smokes K8, however it doesn't destroy it as completely as I had expected at stock speeds. It is clear that anybody wanting a serious high end OCing chip right now should steer clear of anything AM2. Fortunately however, the results do not suggest any kind of intel monopoly (especially considering the 'hopefully' ongoing AMD price cuts), because if the results for the FX62 had 1.5x the FPU performance and a few other tweaks they would probably be about even overall, at least for gaming, and Battlefield 2 matters more than SuperPi. I just don't know if even at 65nm with Si-Ge chips whether AMD is going to be able to match that power consumption/overclockability, other than maybe as cherry picked cores.

i have posted this before, as many don't know. that is if you were referring to what i bolded. but to re state...chips are not manufactured in i.e malay, costa rica, china etc. those are packaging plants. currently oregon, has handled the last few processes; r&d, ramping, and hvm. the processes are then tought to the other fabs in az, ireland, etc...to continue manufacturing.

[greenmaji]

Ive been told by several members that have experince with Intel ES's that the retail chips are always better, I have no reason to not belive this to be true.

[Cheator]

I find it hard to belive. I've never seen an ES that wasn't kick ass. Why give peole reviewing your product terrible ones? That litterally makes no sense.

[ZGOZZ]

Well we just build a laptop with the Duo duelcore.
I am very impressed with the speed and the long battery life.
This laptop witha 1.6 2mb cach duo is great.
Can not wait to see what the DEE does.

[hUMANbEATbOX]

I find it hard to belive. I've never seen an ES that wasn't kick ass. Why give peole reviewing your product terrible ones? That litterally makes no sense.

ES chips aren't made strictly to give out for review. they are samples. that is all.

with p4's, the ES chips were usually average, some of them being complete duds.

[Cheator]

ES chips aren't made strictly to give out for review. they are samples. that is all.

with p4's, the ES chips were usually average, some of them being complete duds.

Often the CPUs given out for review ARE ESs. Not all ESs are reviewed, yes. But when it is a new product that isn't out yet (such as conroe), it would mostl ikly be an ES at that point.

As for the dud P4 ESs, that is rather suprising, and somewhat odd.

[]-[itman]

He posted room temp 75F and proc. temp. 77F. thats with a monster heat pipe air cooler, Is that possible? Is it BS? I don't know the poster well enough to give an answer to that.

He posted that Southern California was 75F and that his in his apartment there is no AC, so 77F. Again, he doesn't say specifically, but that has to be his room temp or else it is the chip at idle, no way that's at full load.
http://www.hkepc.com/hwdb/x6800vsfx62-11.htm
Granted, the cpu cooler probably isn't as strong, but the guy at xs is also running a much higher vcore and a much higher speed. Again, a 1C increase in heat for the chip he is running is impossible unless the computer is in some kind of sleep mode, lol. Even at idle hkepc shows a 10C increase in temps.


And as posted before, aparently the badaxe has issues reporting CPU temps correctly.
I've seen some strangely high (1.6V+) Vcore's to fight Vdroop with Asus+pressler systems (Vdroop mods fixed this) thats why I asked that question.



3dmark scores are higher with Core2 as well http://smiley.onegreatguy.net/read.gif http://smiley.onegreatguy.net/happydrunks.gif

I don't know about all 3dmark scores, but 05 and 06 show at best a 10% improvement clock for clock.

[hUMANbEATbOX]

some more benchmarks, at high resolutions:

http://forums.guru3d.com/showthread.php?t=185555

F.E.A.R:
1600x1200, 4xAA 16xAF, all max:
Opteron 165 @ 3GHz: Min 25 | Avg 43 | Max 172
Conroe @ 3GHz: Min 53 | Avg 86 | Max 310

Oblivion:
Opteron 165 @ 3GHz: Min 15 | Avg 83.993 | Max 689
Conroe @ 3GHz: Min 29 | Avg 132.776 | Max 1546

[AdvanS13]

As for the dud P4 ESs, that is rather suprising, and somewhat odd.[/QUOTE]

you would be suprised how much torture those es's go through here on testing....they are just that for testing, beat to hell, tortured tested etc etc. es's are always hit and miss...the last four ive had from work however have been perfect.

[Mr.Guvernment]

while i disagree that the results of conroe are dissapointing, it could be well argued that for 99% of what a normal person does with a computer, a $149 a64 3800 on a $90 board would be a way more intelligent choice than sinking $300 on a proc and $250 on a board.

His point is valid. AMD will still be a great choice for a budget conscious system that still wants good performance.

Though when 930's drop down to 160 or so in july, it will be a very very very affordable processor......which i plan to buy!


or even more valid is an intel 805 for even cheaper and you then get dual core :D....... dont forget, intel is cutting prices upto %50 on some chips as well as you mentioned.

[Evilsizer]

-[itman']He posted that Southern California was 75F and that his in his apartment there is no AC, so 77F. Again, he doesn't say specifically, but that has to be his room temp or else it is the chip at idle, no way that's at full load.
http://www.hkepc.com/hwdb/x6800vsfx62-11.htm
Granted, the cpu cooler probably isn't as strong, but the guy at xs is also running a much higher vcore and a much higher speed. Again, a 1C increase in heat for the chip he is running is impossible unless the computer is in some kind of sleep mode, lol. Even at idle hkepc shows a 10C increase in temps.



I don't know about all 3dmark scores, but 05 and 06 show at best a 10% improvement clock for clock.
05 and 06 are for the most part gpu bound. it would really take a stronger cpu then what amd or intel has now not to be the bottle neck for those 2.

[Jame Gumb]

or even more valid is an intel 805 for even cheaper and you then get dual core :D....... dont forget, intel is cutting prices upto %50 on some chips as well as you mentioned.
60%

[]-[itman]

some more benchmarks, at high resolutions:

http://forums.guru3d.com/showthread.php?t=185555

First off, those are the only 2 game examples from his list that show any sort of significant advantage to Conroe, second, I have issues with both of those.

1)
http://www.xbitlabs.com/articles/video/display/evga-7900gt_10.html

With a single 7900GT running 27% slower on the core and with a 15% slower cpu, the slower system still ties this guy's SLI highly oc'd system?!? Not only that, but the results go from a 17% Conroe advantage (avg fps) at 1024x768 where the cpu will be a bigger factor, to a 100% Conroe advantage at a higher resolution where the cpu will be significantly more bottlenecked by the graphics. That test is somehow messed up, it just makes absolutely no sense. The only thing I can think of is if he forgot to turn on SLI when doing the high res test, that would fit the results he got a lot better. Perhaps it's the fact that he's running on a modded board to allow for SLI rather than a true SLI board, whether that might make a difference in some cases, I don't know, but it's possible.

2)
http://www.xbitlabs.com/articles/video/display/evga-7900gt_13.html

With the same situation as before with a single gpu with a 27% slower core speed and a 15% slower cpu, only this time there is 16xAF rather than 8x of the faster system. Again, the slower system is able to obtain a 67% faster minimum fps in a similar run, and the faster system leads in avg fps by only 71% (remember this is with a 15% faster cpu in a game heavily effected by cpu power, and a sli solution running 27% faster on each card and 8xAF vs 16x). Again, perhaps it's the board, but something's not right.

[hUMANbEATbOX]

to get SLI to work, you don't mod the board, you mod the drivers AFAIK (or use old ones back when NV supported it on intel chipsets).

he said in his post that the only change he made between the 2 tests was change the res from 1024*768 to 1600*1200. i believe he implied that he didn't even leave the game.

as for the rest, i dunno man. your guess is as good as mine. it IS 2 completely different people running the tests (the link you've supplied vs. the link i posted) so there could be many differences to the test beds.

*edit* also wanted to add that the Source Stress Test is completely useless. you can throw those results out, because that bench tells you nothing really. it isn't representative of gameplay in any Source games, to me its just a tech demo.

[]-[itman]

to get SLI to work, you don't mod the board, you mod the drivers AFAIK (or use old ones back when NV supported it on intel chipsets).

he said in his post that the only change he made between the 2 tests was change the res from 1024*768 to 1600*1200. i believe he implied that he didn't even leave the game.

Doesn't mean something couldn't have gone wrong, review sites have things happen all the time that take them hours to figure out, the only clue they have is that something just doesn't make sense, and this just doesn't make sense. Again, if the SLI drivers goofed and didn't work right at the new res, then that would make perfect sense.

as for the rest, i dunno man. your guess is as good as mine. it IS 2 completely different people running the tests (the link you've supplied vs. the link i posted) so there could be many differences to the test beds.

Both his test and the one I pointed to were done at max quality with eye candy turned on with almost identical systems. Fear isn't the longest game and has many repetitive environments, but even if the faster machine was benched on a more demanding demo, there's no way it could make up for an sli rig vs a slower clocked single card system. All I'm saying is that something is wrong with the test and a %100 improvement is not realistic.

*edit* also wanted to add that the Source Stress Test is completely useless. you can throw those results out, because that bench tells you nothing really. it isn't representative of gameplay in any Source games, to me its just a tech demo.

I agree, *although I would like to point out that basically the same demo was done by intel during IDF and they magically found a 24% advantage for Conroe.


*Note: just a little venting about corporate bs in general, not meant to be a hate post against intel.

[>HyperlogiK<]

I find it hard to belive. I've never seen an ES that wasn't kick ass. Why give peole reviewing your product terrible ones? That litterally makes no sense.

If I remember rightly the PD 9XX series ES chips were pants and ran hotter than the PD 8XXs.