AMD Shows Deneb (Phenom II) pictures of 6.2GHz CPUz etc

[Tiridum]

Performance/watt is the key, not just power draw in isolation.

perf/watt is very good for nehalem, power draw is not....

[accord99]

perf/watt is very good for nehalem, power draw is not....

It's not bad, given the enthusiast platform it's currently paired with and tri-channel memory support. It only looks bad because it's compared with extremely power efficient 45nm C2Qs.

[Zucker2k]

perf/watt is very good for nehalem, power draw is not....

That's a contradiction; focus, guys.

[STaRGaZeR]

That's a contradiction; focus, guys.

Nah, it is not. You can consume a load of power and be very efficient (perf/watt) in what you're doing. The "problem" here is that while being much more efficient than previous processors Nehalem still consumes more power than those same processors, specially with HT on. But still it is within in the margins of acceptable power consumption, so no big deal. Also, the high consumption is only under load, at idle is the same or even better than Penryn. Nehalem sucks power when needed, and performs in consequence. When it is not required, it sleeps deeply.

So if you have/want a limit of watts, Nehalem can be a problem. But so can be any other CPU, and if you have PC restrictions Nehalem will perform better at any given limit. I don't see any reasons (except price, and maybe max OC compared with Penryn) to get any other CPU. P2 will only have a fight with it in price. Mediocre and within expectations, but we should thank AMD for it. In the end they are the only competition Intel has.

[Jazzman]

perf/watt is very good for nehalem, power draw is not....

Here is how Intel views "power draw".

Figure 3 there shows Intels statistical thermal analysis of MPUs across application workloads ->

Generally, the TDP target is determined as a function of the anticipated thermal stress load the processor will encounter, which in turn is a function of the application software run on the processor. For example, FIG. 3 shows a graph of CPU power vs. number of applications, which is illustrative of a typical power consumption vs. application type distribution. Generally, each application program has its own unique power profile, although the profile has some variability due to loop decisions, I/O activity and interrupts. The graph illustrates a statistical distribution based on averaged application power consumption.

US Patent 7275012 - Automated method and apparatus for processor thermal validation

[BrowncoatGR]

Here is how Intel views "power draw".

Figure 3 there shows Intels statistical thermal analysis of MPUs across application workloads ->





US Patent 7275012 - Automated method and apparatus for processor thermal validation

TDP is a thermal spec. It's not directly related to power consumption.

[Jazzman]

TDP is a thermal spec. It's not directly related to power consumption.

Yep, you are correct. Thanks to informal and Macadamia for this SOURCE.

Overclocking: Whatever Happened to Headroom?

Power consumption of an IC is a function of the number of switching events and the square of the supply voltage, at least in theory. If leakage currents are taken into account, we found that the third power of the voltage provides a better fit for any of the CPUs we have measured over the years. What this amounts to is that a 10% increase in voltage will result in a 21% increase in power consumption and thermal dissipation using the classic square function and a 33% increase using our empirically derived function. If the operating frequency is increased by 30% (knowing that a lot of the Nehalem cores will do run up to 4.1 GHz) using 10% overvolting, the thermal dissipation will reach 177% of the stock value. That is, even using the official TDP of 130W, we are looking at roughly 230W, which approximates a power density of some 100W/cm2. This number is not out of the world but bear in mind that the basis for the calculation was not the maximum power consumption but what a typical user may experience using commercial software.

[accord99]

Yep, you are correct. Thanks to informal and Macadamia for this SOURCE.

You forgot to bold the part where it's the estimated power consumption of an overclocked Nehalem 4.1 GHz.

[Jazzman]

You forgot to bold the part where it's the estimated power consumption of an overclocked Nehalem 4.1 GHz.

Thanks. I added your suggested highlight.