New Xbox 360 production revision hits shelves

I have a somewhat philosophical question: How come it's 'news' when Dailytech reports on unconfirmed speculation from another forum when speculation regarding the same information have been going on, even on your own forums, for days? Because someone else ran with it, so now it's 'their fault' if it's wrong? Does all forum-originated news need some big site to jump on board first before becoming proper news?

Spoiler

If so, where's the news that Gamespot and Joysiq are reporting that Halo 3 is running 640p?

 

Wrong. There have been older chips from Canada.

Try temperature measurements first.

*sigh*

 

Huh? Philosophical? What?

It's news from DailyTech because that's the first place I spotted it. Nothing more, nothing less. Maybe the question should be, "Rys, don't you read your own forums?"

I can answer that one a bit better

 

As in: "This seems to be a trend in online journalism". Just had a discussion about the standard of source citing, and how online media often spread stories by referring to each other in a grand circlejerk fashion that only serves to obscure the actual source of the information. Sometimes they even morph enough along the way to come back around as a different story from the same outlet; like the 'whisper a sentence' game kids play at parties.

This wasn't a jab at you or B3D specifically, just wondering whether it was deliberate.

My point is that it is less and should be more. That Dailytech read some stuff on a forum isn't news, nor is it news that Rys read something on Dailytech. The stuff Daylytech read on a forum might be news, but then the B3D news should be based on that (feel free to credit Dailytech for bringing it to your attention, though).

Fair enough.

 

Temperature has nothing to do with power dissipation if you use different heatsinks. And a die shrink of a part at the same speed from 90nm to 65nm will certainly reduce power consumption.

Also the die size of a 65nm part is half of the same part in 90nm, so the wafer costs are reduced in half if TSMC is not charging more for the same process. Packaging cost and other stuff will not go down of course.

 

Obviously that wouldn't be a fair comparison. You need to test the same heatsink on both chips and measure the temperature. It shouldn't be that difficult to think of considering the person already took off the heatsink.

A die shrink should lead to a reduction in power consumption. There is no confirmation that these parts are 65nm beyond speculation. The temperature test is a far better metric than what Dailytech said:

"The heatsink doesn't have a heatpipe, and therefore must be 65nm! It's confirmed! Less crazy heatsink == smaller process node!"



Which is plain ridiculous. The temperature test would at least give a more concrete idea of any changes. Wouldn't you agree?

Except for the fact that the die isn't that much smaller as shown in the photos. How do you get 50% when the die size appears to have only decreased a small percentage?

If you aren't producing twice as many chips per wafer (with the same yield), you aren't going to get 50% reduction. Even if the 65nm were to have better yields than 90nm initially, the die size being so similar to that of the 90nm part won't give you that high of a price reduction because then the 90nm yields would have to be twice as terrible in the first place.

So again -> to the 50%.