Tatty_OneVery true that, there are some star PII's out there overclocking wise, thing is, with for example the Q9650, you dont even need to be a "hardcore" overclocker to get some serious results on just plain ol air cooling, like almost 4.9gig for example but it's fairly commonplace to get 4.5 - 4.8gig on air, yes I agree, in the extreme the 940BE may overclock higher, in the mainstream though I doubt very much if it could hold it's ground as well.

www.ripping.org/database.php?cpuid=851

mainstream Q9650 wins, extremes 940 wins

newtekie1You make that whole argument, and your own chip can only do 4.14GHz benchmark stable on water.:laugh:

Please, in extreme conditions and cooling, the Phenom II BE might hang with the Core 2 Quads. But everyone but the hardware overclockers and modders will see better overclocking results from an unlocked Phenom II.

whats wrong with the water part? i have seen air cooled 940's do the same? with extreme's they don't just hang they win find me 6.6ghz 3dmark runs on a Q9650 they still win on LN2

Wile EThat's not true at all. The better of the QX9650's see 4.6GHz on water. My crappier one does 4455 on water, and at fairly decent bench voltages. That's 1.625V for mine. Not to mention the QX9650 is faster overall, let alone this 975XE, which walks all over every other cpu available, period.

i had seen alot recently not clocking very well....

Yeah the i7 920 easily is a much better oc'er than the PII 940, much less the XE. Also, this time around intel's XE's benefit from a wider bus bandwidth, not just the unlocked multi.

I wonder if the i7 940 really will be phased out or brought down to earth a la intel's past scheme.

cdawallmainstream Q9650 wins, extremes 940 wins




whats wrong with the water part? i have seen air cooled 940's do the same? with extreme's they don't just hang they win find me 6.6ghz 3dmark runs on a Q9650 they still win on LN2



i had seen alot recently not clocking very well....

Those 6+Ghz clocks don't mean much when a 5+Ghz 975XE outscores them in the cpu tests.

Wile EThose 6+Ghz clocks don't mean much when a 5+Ghz 975XE outscores them in the cpu tests.

i never said the 975XE wasn't better. and the phenom is still holding its own in 3dmark thanks to the better chipset.

cdawalli never said the 975XE wasn't better. and the phenom is still holding its own in 3dmark thanks to the better chipset.

How are any AM2+/AM3 chipsets better than the x58?

cdawalli never said the 975XE wasn't better. and the phenom is still holding its own in 3dmark thanks to the better chipset.

Well, the original statement that sparked the debate made reference to Intel charging so much for the unlocked multi, whereas AMD does not.

Well, that's because the unlocked multi isn't the only factor. Intel can get away with charging so much because their chips are faster.

In case anyone fails to remember, when AMD held the performance crown, they charged $1000 for their unlocked chip as well.

farlex85How are any AM2+/AM3 chipsets better than the x58?

long list or short list? how about spec wise its comparable to a mid range 790X mobo was last year and 790FX has more features and better PCI-e lane control than any intel chipset could imagine. hence why benchmarks show that single card with comparable cpu's the intel wins and multi cards phenom makes up that difference and comps over it.

Wile EWell, the original statement that sparked the debate made reference to Intel charging so much for the unlocked multi, whereas AMD does not.

Well, that's because the unlocked multi isn't the only factor. Intel can get away with charging so much because their chips are faster.

In case anyone fails to remember, when AMD held the performance crown, they charged $1000 for their unlocked chip as well.

yes they did long live the FX57 and FX 62 but during that time intel charged just as much for a 3.73ghz (i think?) P4 that performed worse :nutkick: intel always charges a shitload its intel thats how they roll

cdawallyes they did long live the FX57 and FX 62 but during that time intel charged just as much for a 3.73ghz (i think?) P4 that performed worse :nutkick: intel always charges a shitload its intel thats how they roll

Correct ;) and the FX 57 will live on for yrs to come :pimp:

And yes thats how intel roll, always have, at least now they do have CPU's that do perform so they can sell for a higher price, but back then no one new what AMD was so it also worked for intel back then to.

The day I pay $1000 for a cpu just so I can post a cpu-z pic in a forum.:banghead:
I have been using AMD chips since the day they came out and never ever had a time where I said ,wish I had an intel.:rolleyes:

cdawalllong list or short list? how about spec wise its comparable to a mid range 790X mobo was last year and 790FX has more features and better PCI-e lane control than any intel chipset could imagine. hence why benchmarks show that single card with comparable cpu's the intel wins and multi cards phenom makes up that difference and comps over it.

What 790X board from last year had the ability to do cf and sli and a ddr3 mem controller? The 790FX is really the only board that approaches the x58 in terms of features, but it falls short of allowing cf and sli, and HT just isn't as fast as qpi. I also imagine your blowing smoke w/ the lane control thing, but I'll just go ahead and take your word for it.:D

cdawallmainstream Q9650 wins, extremes 940 wins




whats wrong with the water part? i have seen air cooled 940's do the same? with extreme's they don't just hang they win find me 6.6ghz 3dmark runs on a Q9650 they still win on LN2



i had seen alot recently not clocking very well....

I believe a more accurate statement is that in general Intel's even with the locked multipliers overclocker better than the AMD chips with unlocked multipliers. It is only in a few extreme situations that the AMD chips can compete.

What is wrong with the water part is that Intel chips do the same or better on air, and you are making the argument that the Phenom II's are better at overclocking because of the unlocked multiplier, yet yours can't even manage to match stable Intel clocks without water.

farlex85What 790X board from last year had the ability to do cf and sli and a ddr3 mem controller? The 790FX is really the only board that approaches the x58 in terms of features, but it falls short of allowing cf and sli, and HT just isn't as fast as qpi. I also imagine your blowing smoke w/ the lane control thing, but I'll just go ahead and take your word for it.:D

Well in all fairness, the DDR3 memory controller is on the CPU, not the chipset. The 790FX does allow for more PCI-e lanes than the x58(42 vs. 40), but it doesn't really matter. Considering those two extra lanes don't really help out that much. Both natively support two PCI-e x16 slots and a third PCI-E x8 slot, and neither support three full PCI-e x16 slots.

The 790X statement is just laughable though, with support for only x8/x8 configurations, it doesn't come close to the x58.

farlex85What 790X board from last year had the ability to do cf and sli and a ddr3 mem controller? The 790FX is really the only board that approaches the x58 in terms of features, but it falls short of allowing cf and sli, and HT just isn't as fast as qpi. I also imagine your blowing smoke w/ the lane control thing, but I'll just go ahead and take your word for it.:D

so for the lane control thing go look at Xfire and SLi on AMD and notice how it scales better often giving AMD chips an advantage over intel ones there are an infinite number of cases were AMD has beaten i7 in xfire because intel chipsets suck ass with multiple cards. hence why AMD holds the WR with 3D06. 2x4870X2 on a 6.6ghz 945ES stepping 2 beats 2x4870X2 on i7@5.5ghz any other test the AMD would loose but the high bandwidth that the X2's use show the better control on the AMD chipset than the intel one. hence AMD wins 3D.

oh and the SLi+xfire thing is fine and dandy intel threatened to drop NV chipsets from intel's line completely and NV gave them the ability to put SLi on there chipsets. it was a dirty move by intel. now i have a crosshair II that will actually with the right drivers run xfire and SLi oh and it uses the same BR02 chip to give it more pci-e lanes than X58. (51 total)


you positive QPI is faster? HT is as fast before you break into the fact that its DDR tech. so double that number and you have the total bandwidth or 32GB/s max for intel vs 51.2GB/s max for AMD

wiki]Performance numbers for QuickPath are reported to be 4.8 to 6.4 Gigatransfers per second (GT/s) per direction. Therefore the bandwidth amounts to 12.0 to 16.0 GB/s per directionI believe a more accurate statement is that in general Intel's even with the locked multipliers overclocker better than the AMD chips with unlocked multipliers. It is only in a few extreme situations that the AMD chips can compete.

What is wrong with the water part is that Intel chips do the same or better on air, and you are making the argument that the Phenom II's are better at overclocking because of the unlocked multiplier, yet yours can't even manage to match stable Intel clocks without water.

Well in all fairness, the DDR3 memory controller is on the CPU, not the chipset. The 790FX does allow for more PCI-e lanes than the x58(42 vs. 40), but it doesn't really matter. Considering those two extra lanes don't really help out that much. Both natively support two PCI-e x16 slots and a third PCI-E x8 slot, and neither support three full PCI-e x16 slots.

The 790X statement is just laughable though, with support for only x8/x8 configurations, it doesn't come close to the x58.

and sadly enough with 8x/8x the 790x shows better scaling than X58 does with full X16 slots this has been proven many time with 790GX which has the same logic as 790X and adds a iGPU

cdawallso for the lane control thing go look at Xfire and SLi on AMD and notice how it scales better often giving AMD chips an advantage over intel ones there are an infinite number of cases were AMD has beaten i7 in xfire because intel chipsets suck ass with multiple cards. hence why AMD holds the WR with 3D06. 2x4870X2 on a 6.6ghz 945ES stepping 2 beats 2x4870X2 on i7@5.5ghz any other test the AMD would loose but the high bandwidth that the X2's use show the better control on the AMD chipset than the intel one. hence AMD wins 3D.

I haven't really seen or heard anything to indicate better scaling on 790X chipsets. Aside from this one particular instance in this one bench (WR on 3dmark 06 is hardly telling of an overall trend), are there any other comparisons of the like?

farlex85I haven't really seen or heard anything to indicate better scaling on 790X chipsets. Aside from this one particular instance in this one bench (WR on 3dmark 06 is hardly telling of an overall trend), are there any other comparisons of the like?

here is one in vantage the i7 wins the full score due to being a better cpu but look at the gpu's they are @ the same clocks and the P2 wins


i7
img.techpowerup.org/090320/img.jpg

phenom II
service.futuremark.com/resultComparison.action?compareResultId=730096&compareResultType=19

found that just flipping through hwbot

cdawallhere is one in vantage the i7 wins the full score due to being a better cpu but look at the gpu's they are @ the same clocks and the P2 wins


i7
img.techpowerup.org/090320/img.jpg

phenom II
service.futuremark.com/resultComparison.action?compareResultId=730096&compareResultType=19

found that just flipping through hwbot

Gpu scores are within 100 points of each other, well within margin of error for 3dmark (you can get 100 points difference with two identical runs w/ the same machine). Also, how do you know the clocks for the phenom run? I see 18mhz and 9mhz for core and mem, so I'm not real sure what that means. Lastly, gpu scores on 3dmark are not specifically calculated just by how the gpu performs, different factors seem to effect gpu score subtlely as well. Anyway, I was more wondering if there were some direct comparisons somewhere.

farlex85Gpu scores are within 100 points of each other, well within margin of error for 3dmark (you can get 100 points difference with two identical runs w/ the same machine). Also, how do you know the clocks for the phenom run? I see 18mhz and 9mhz for core and mem, so I'm not real sure what that means. Anyway, I was more wondering if there were some direct comparisons somewhere.

3.8ghz for the phenom.


for your direct comparo if you can wait a couple of days i have a phenom II and NV mobo and freaksavior has an i7 and 2 GTX285's we will bench them both and see how they do

cdawall3.8ghz for the phenom.


for your direct comparo if you can wait a couple of days i have a phenom II and NV mobo and freaksavior has an i7 and 2 GTX285's we will bench them both and see how they do

You need the gpu clocks to make a comparison though.

That'll be an interesting comparison, I'll put my money on them being similar.

farlex85You need the gpu clocks to make a comparison though.

That'll be an interesting comparison, I'll put my money on them being similar.

we both have highend everything from ram to HDD. our next bench session is going to look like this

freaksavior:
2xi7 920
DFI X58
3x2GB pi 1600
3x2GB HCF8's (my gift to him)
300GB vraptor
DICE/water cooling
2xGTX285, 2x2600PRO

me:
phenom 945ES, 9750, 5600 X2
M4A78T-E, crosshair II formula
ton of DDR3, 2x1gb dom D9's PC6400 3-4-3-9, 2x256mb D9DCD's (1160@4-4-4-12), 4x2GB D9HNP
300GB vraptor
DICE
2x2600PRO,790GX,onboard nv8200,2xGTX285

is there a chance that the i7 920 will go down in price? oh and I might of read wrong but I THINK I saw saw someone say x58 dont have 3x PCI-e 16x... but doesn't the P6T6 WS Rev have it?

cdawallmainstream Q9650 wins, extremes 940 wins

....

Well thats good, purely cause there are probably several hundred thousand "mainstream" and just a handful of "extreme's"....... I am happy with that :D

cdawall3.8ghz for the phenom.


for your direct comparo if you can wait a couple of days i have a phenom II and NV mobo and freaksavior has an i7 and 2 GTX285's we will bench them both and see how they do

I have already done that with another member here with a PII 940, I wont say how we did though until your results are up, be interesting to see if they are similar.

if i win the lotto i might get the q8400, but from here to win the lotto :(

Tatty_OneWell thats good, purely cause there are probably several hundred thousand "mainstream" and just a handful of "extreme's"....... I am happy with that :D

now i wonder what i am

Tatty_OneI have already done that with another member here with a PII 940, I wont say how we did though until your results are up, be interesting to see if they are similar.

i wonder how a 945ES compares we will see

cdawallso for the lane control thing go look at Xfire and SLi on AMD and notice how it scales better often giving AMD chips an advantage over intel ones there are an infinite number of cases were AMD has beaten i7 in xfire because intel chipsets suck ass with multiple cards. hence why AMD holds the WR with 3D06. 2x4870X2 on a 6.6ghz 945ES stepping 2 beats 2x4870X2 on i7@5.5ghz any other test the AMD would loose but the high bandwidth that the X2's use show the better control on the AMD chipset than the intel one. hence AMD wins 3D.

oh and the SLi+xfire thing is fine and dandy intel threatened to drop NV chipsets from intel's line completely and NV gave them the ability to put SLi on there chipsets. it was a dirty move by intel. now i have a crosshair II that will actually with the right drivers run xfire and SLi oh and it uses the same BR02 chip to give it more pci-e lanes than X58. (51 total)


you positive QPI is faster? HT is as fast before you break into the fact that its DDR tech. so double that number and you have the total bandwidth or 32GB/s max for intel vs 51.2GB/s max for AMD







and sadly enough with 8x/8x the 790x shows better scaling than X58 does with full X16 slots this has been proven many time with 790GX which has the same logic as 790X and adds a iGPU

One world record doesn't show better scaling. If you want to prove better scaling, you have to take 2 untouched/untweaked Windows installs one on each mobo, then run them with the same cards at the same clocks. And by same cards, I mean the same cards, not the same model. Individual cards respond differently to different settings. That's why these WR benchers go thru so many cards to get the right ones. Each individual card responds differently at that level of overclocking.

hwbot scores don't prove much, with all the different levels of OS tweaking and other tricks some do, while others don't. You have to make a controlled comparison to make these types of claims, otherwise you leave too much to question.

cdawallnow i wonder what i am



i wonder how a 945ES compares we will see

A 945ES? still not as good clock for clock........... oooppppssss I gave the game away :cry:

:p