Guden Oden said:
This is complete and utter bunk. It's not going to be water cooled.
The CPU HeatSink will have a water block of some type, this was confirmed in the "OurColony" video. Something similar to the DreamCast heatsink that also had liquid inside. No one is saying it is going to have a water pump and circulate water to an externally cooled water tank.
MS has not demoed or explained exactly what they are implimenting but they have indicated something more than just a solid block with some water inside. My guess is that it is a heatpipe design of some type. Either way, in their opinion it was significant enough to mention.
Anyhow, it is passively cooled (with the case fans used to remove excess heat from the case).
This makes me scratch my head, because the thing IS already cooled by small fans! There's two fans sitting right at the back of the unit, you can even see one of them in the pictures, the one focusing on the power connector.
It's not passively cooled. Just because there's not a fan strapped straight onto the heatsink doesn't mean it's passively cooled. The sink sits in a channel of forced airflow = active cooling.
1. Having a case fan and having a fan directly attached to a heatsink result in TOTALLY different cooling performance.
2. It is commonly accepted that a video card with a heatsink and no fan is considered "passively cooled" even if the PC case has fans. COMMON SENSE tells you that even in passively cooled systems--especially in a small form factor--that there is a need to remove the excess heat from the case.
If it makes you feel any better (since you obviously want to argue about the technicalities):
"If Xenos can comfortably attain 500MHz and run at acceptible temperatures with a heat sink that has NO fan and has air flow from two small case fans, it would seem plausible that with a moderately fast fan ON THE HEAT SINK would improve the cooling properties of the heatsink enough to possibly attain higher frequencies".
There is serious talk of the R520 being clocked at 650-700MHz on the 90nm process. So the process seems capable.
Anyhow:
A heatsink with a fan is going to cool better than a heatsink without a fan.
All my years of computing tells me that sticking a fan directly on the heatsink (even in a system with good air flow) is going to significantly increase the cooling properties of the heatsink. Heat output does not necessarily scale linearly with clockspeed, but a stable 10% overlock (which was the low end of my figures BTW) with a change in cooling method seems possible. I never said for certain, but at least something to ponder.
You want 20% higher performance, you're going to pay MORE than 20% higher price. MS
lowered GPU performance of the original box by only 6.3% because the faster speed would have been too expensive, and here you're speculating wildly about raising it 20% just by adding a fan...
Yets, lets go back over our history.
Xbox1 was going to have a 300MHz processor.
But this got revised all the way down to 250MHz and finally 233MHz when MS realized that 1) they were not going to hit the 150nm process and 2) when yields stunk at the larger process and there was a lot of heat due to the large die size.
The end result: a HUGE Xbox case with loud fans.
Fast forward to today. We have not heard of any yield problems with Xenos (which had working silicone in November of 2004 according to ATI).
Further, it is cool enough to 1) be in the smaller Xbox 360 case and 2) can be effectivelly cooled without a fan directly on the heatsink. The GPU only draws 25W of power according to ATI.
So your analogy to the NV2a 6% in speed is not relevant. Heck, I would argue that the higher clock speeds on the R520 and the fact the GPU does not have a fan indicate that if one were to stick one on the heatstink is a much more persuasive stance. Drawing lines to the Xbox1 is really irrelevant IMO because they are totally different situations (and the end result, loud fans and a large form factor) are obviously not visible with the Xbox 360. This indicates some wiggle room to me.
As for what I said: Of course it is speculation. I never made any claims otherwise.
Besides, if you've had to listen to a god damn PS2 whirring away for hours upon end while watching movies with quiet dialog scenes, I'm not sure you'd maintain your position that more noise = not so bad. It can be quite distracting, because not all games make a ton of noise,
And yet that did not prevent over 90M PS2s from selling did it?
and some may want to use their 360s for listening to music too, and then you simply can't have dustbuster fans in it.
As Tap-In noted MS already noted that when the load is low 2 of the cores poweroff.
