What are the other quads after the current intel ones?.. True quad = a lot cooler

[Perfect_Chaos]

anybody know?

i am waiting to see how Barcelona fairs first in games etc compared to C2D, but i hear when true quad cores come they will be a heck of a lot cooler.. thank god.

 

[Mr Paul]

Barca is a true quad core.

 

[Jokester]

Don't assume that just because all 4 cores are on the same lump of silicon that it will magically be faster as a result. It's a whole lot more complicated than that and will ultimately be more dependent on the specific features that each individual core has rather than the interface between them.

Jokester

 

[FreeStream]

Well current quad cores are only 100W TDP... thats still a lot less than any of the old 939 processors and a lot less than any of the top end P4 range.

The quad cores run at 1.25V stock IIRC while Conroes are 1.3-1.35V and thats 10W per core straight away.


Start overclocking and heat output goes through the roof. My 3.4ghz Conroe is > 100W so a quad core is gonna be > 200W and thats a lot of heat to get through a 4cm^2 area...


Yea next gen will be cooler, but don't expect anything hugely beneficial, once you start ramping up the clocks and voltage (even more so at smaller die sizes) the output will still pass 150W atleast, you wont see quad cores doing 100W output for a few years yet I'd imagine. Overclocked that is, Intel/AMD might bring out some ultra-low voltage versions which usually don't clock very well...

 

[Perfect_Chaos]

Well current quad cores are only 100W TDP... thats still a lot less than any of the old 939 processors and a lot less than any of the top end P4 range.

The quad cores run at 1.25V stock IIRC while Conroes are 1.3-1.35V and thats 10W per core straight away.


Start overclocking and heat output goes through the roof. My 3.4ghz Conroe is > 100W so a quad core is gonna be > 200W and thats a lot of heat to get through a 4cm^2 area...


Yea next gen will be cooler, but don't expect anything hugely beneficial, once you start ramping up the clocks and voltage (even more so at smaller die sizes) the output will still pass 150W atleast, you wont see quad cores doing 100W output for a few years yet I'd imagine. Overclocked that is, Intel/AMD might bring out some ultra-low voltage versions which usually don't clock very well...

and the only best way to cool them while clocking is with water, without spending a fortune that is.

i think its possible to get 3.8Ghz no probs with water? isnt it something like 4.2 on phase?, in that case a lot cheaper to go with water i would imagine.. if going water -> phase its gonna cost u like 200-300 quid more for a 300-400mhz boost in clockspeed over water

 

[Kamakazie!]

Well current quad cores are only 100W TDP... thats still a lot less than any of the old 939 processors and a lot less than any of the top end P4 range.

The quad cores run at 1.25V stock IIRC while Conroes are 1.3-1.35V and thats 10W per core straight away.


Start overclocking and heat output goes through the roof. My 3.4ghz Conroe is > 100W so a quad core is gonna be > 200W and thats a lot of heat to get through a 4cm^2 area...


Yea next gen will be cooler, but don't expect anything hugely beneficial, once you start ramping up the clocks and voltage (even more so at smaller die sizes) the output will still pass 150W atleast, you wont see quad cores doing 100W output for a few years yet I'd imagine. Overclocked that is, Intel/AMD might bring out some ultra-low voltage versions which usually don't clock very well...

According to AMD the Barcelona quad cores have a TDP of 120W. AMD always use the MAX theoretical TDP so under normal usage it will probably be 100W or so.

0.1V less voltage = 10watt less power? The CPU's are pulling 100Amps? That seems like a ridiculous amount, or is it calculated in another way to the usual P=VA?

As for the current quad cores having a 100W TDP, i thought they were 125W and that is Intel figures which are usually an average, not a theoretical maximum?!?

 

[Jokester]

i think its possible to get 3.8Ghz no probs with water?

Not likely to be honest, most seem to be stable at 3.6GHz but with high temperatures (70C+) when fully loaded.

As Concorde says, they're pumping out over 200W of heat at those speeds and voltages (1.45-1.5ish load volts).

Jokester

 

[Perfect_Chaos]

Not likely to be honest, most seem to be stable at 3.6GHz but with high temperatures (70C+) when fully loaded.

As Concorde says, they're pumping out over 200W of heat at those speeds and voltages (1.45-1.5ish load volts).

Jokester

are they even stable at those temps?

think ill wait and see if any cooler chips come along, if its worth it or not i have no idea

 

[Gashman]

quad core has little usage at the moment, what you planning on doing with this said quad processor, i cannot see any reason why you'd want quad over dual at this time, i know there better with all these theoretical benchmarks and stuff but real-world performance increase over dual core? is there any at all at the moment?

 

[FirebarUK]

I use my Quad for running many Virtual Machine's at once, aswell as doing other things. Great for multitasking.

 

[Perfect_Chaos]

how much cooler are the quads gonna get?

they only just manage stock with air cooling.. bit stupid imo

 

[Gashman]

Not likely to be honest, most seem to be stable at 3.6GHz but with high temperatures (70C+) when fully loaded.

As Concorde says, they're pumping out over 200W of heat at those speeds and voltages (1.45-1.5ish load volts).

Jokester

that all depends on what sort of watercooling setup you chuck behind it, cause WC is mainly limited by the ability of the radiator to exchange heat with the air (which is insulator), with a highly efficient block like an apogee GT and a pump with a high flow rate plus a pair of 120.3 radiators im willing to say the temperatures would be a lot lower than 70*C, might be wrong though but my dual core 90nm athlon running at 2.8Ghz (1.525V) and watercooling never really goes shocking amount above room temperature, though when on air (CNPS9500) the temperatures frequently went above 40*C even as high as 50*C sometimes

 

[Jokester]

That is with top end watercooling. The difference between 1 and 2 120.3 is pretty much nothing (I didn't notice any difference between a 120.3 +120.2) as the extra cooling potential of the extra rad is offset by the increased flow restriction.

Jokester

 

[Perfect_Chaos]

That is with top end watercooling. The difference between 1 and 2 120.3 is pretty much nothing (I didn't notice any difference between a 120.3 +120.2) as the extra cooling potential of the extra rad is offset by the increased flow restriction.

Jokester

so the only way to get around it is a bigger rad than 120.3? and thermochill dont do pa 120.4 or higher do they?

 

[weescott]

I did a 120.2 x 2 loop a couple of years ago. True the extra restriction hurts but it's nothing a pair of DDC ultra's can't cure. Thermochill can do custom rads but the costs are VERY HIGH, so not practical. IIRC a HE120.3 could cope with c500w of heat. So add some high CFM fans and noise and it could cope with a highly clocked quad core on its own. Only when you add other components does a single 120.3 start to struggle.

 

[HighlandeR]

quad core has little usage at the moment, what you planning on doing with this said quad processor, i cannot see any reason why you'd want quad over dual at this time, i know there better with all these theoretical benchmarks and stuff but real-world performance increase over dual core? is there any at all at the moment?

Ditto tbh check the dual/quad core gaming thread in the gaming forum, theres very little difference between single core vs dual core let alone quad core in games.

Although id say a dual core is wise for now and its inexpensive, give it another 6 months or 12 months before the need for quad core by then they will be all true quad cores and cheaper/faster and more games/software will perhaps be supporting and making better use of it.