(Rumor) MS Adding 2nd GPU HeatSink for Better Cooling?

I'd think that cost cutting was to blame as usual. Blaming engineers is usually the wrong direction to go in. It would probably have a sweet carbon nanotube contraption if cost wasn't an issue. :LOL:

In that case PS3 would be the ideal medium ;) Of course, the budget and the time constraint was to blame (their need to launch first and early and cost), but I think the engineers should have done better. But it can't be helped though, since MS is not primarily a hardware engineering company, and likely contract others to do most of their design.
 
It really shouldn't take an engineer to figure out that the more metal you've got for a heatsink, the better the cooling. I mean, look at all that space! ;) (Of course, there is the weight to consider. Adding another few hundred grams and multiplying that for shipping.... it all adds up. :p)


I'd think that cost cutting was to blame as usual. Blaming engineers is usually the wrong direction to go in. It would probably have a sweet carbon nanotube contraption if cost wasn't an issue. :LOL:

I definitely would have looked into CNTs for heat transfer too if I were there. :)

Time is also a factor. It would take forever to make millions of CNT-based heatsinks. Maybe they could have timed it for Japanese sales... ;)
 
It really shouldn't take an engineer to figure out that the more metal you've got for a heatsink, the better the cooling. I mean, look at all that space! ;) (Of course, there is the weight to consider. Adding another few hundred grams and multiplying that for shipping.... it all adds up. :p)

Bigger problem is cost of the heat sink - it certainly is a considerable amount of money more that Microsoft will have to pay for every 360 between now and 65nm SKU. Still, it's their own fault for kind of broken hardware design. And it will cost them less than repairing/replacing all these broken consoles.
 
Bigger problem is cost of the heat sink - it certainly is a considerable amount of money more that Microsoft will have to pay for every 360 between now and 65nm SKU. Still, it's their own fault for kind of broken hardware design. And it will cost them less than repairing/replacing all these broken consoles.

Umm, it's much less than repairing all those broken systems in the first place.

I mean, every time a 360 is repaired under warranty, MS has to pay for shipping both ways just for starters. How much is that, $30? That's just for starters of course, and doesn't account at all for repair costs.

There's been indications that repair costs have been affecting the gaming divisions bottom line negatively lately too. "Higher warranty expenses" have been tabbed in a couple quarterly reports.

The new heatsink is just a flimsy piece of metal. Negligible costs.
 
in my opinion this doesnt make sense. If most of you are coming to the conclusion that an overheating GPU is the cause of the broken hardware message, how is it persisting and breaking so quickly as well as permanently? Most people keep saying they went to turn it on and it simply wouldnt work. Anyone whos played around with computer hardware knows that if something overheats shutting it off 5 minutes allows another period of problem free usage. Also as far as im aware nobody has cured their issues, at least not in mass numbers, by sticking the console next to a decent fan or simply removing the shell, both would do a hell of alot more then a little dinky peice of aluminum and copper heatpipe addition if that was indeed the problem for everyone.

In my opinion if this comes to fruition its to quite simply lower the temperature of the GPU, especially for those people that stick the console in a poorly ventilated area, not to resolve the universal 3 Read Light problem.
 
in my opinion this doesnt make sense. If most of you are coming to the conclusion that an overheating GPU is the cause of the broken hardware message, how is it persisting and breaking so quickly as well as permanently? Most people keep saying they went to turn it on and it simply wouldnt work. Anyone whos played around with computer hardware knows that if something overheats shutting it off 5 minutes allows another period of problem free usage. Also as far as im aware nobody has cured their issues, at least not in mass numbers, by sticking the console next to a decent fan or simply removing the shell, both would do a hell of alot more then a little dinky peice of aluminum and copper heatpipe addition if that was indeed the problem for everyone.

In my opinion if this comes to fruition its to quite simply lower the temperature of the GPU, especially for those people that stick the console in a poorly ventilated area, not to resolve the universal 3 Read Light problem.


Well, there's been a lot of speculation of the type that basically, heat is warping the motherboard, causing it to flex, when that happens the GPU (I think it's the GPU?) loses contact with the motherboard. Hence instant shutoff. There's also some blame on an "X-Clamp" that attaches the heatsink for initiating this flexing, but it all gets very confusing, and I'm not sure how things could be so simple and MS not have fixed it yet. I dont know, I'm just very confused about what exactly is the issue. Like you I've never believed in garden variety "overheating" being the culprit, it's just not very plausible.

Also:

New Pictures of 2nd GPU HeatSink on Xbox 360 - Update: Video
>> =Ikari= sent us these pictures of his Xbox360 he just got back from repair (after RoD problem) ... another confirmation of the new 2nd heatsink/pipe technology we talked about yesterday that Microsoft installs to provide additional cooling for the Xenos GPU. For now it's unknown if Microsoft just installs this new dual-heatsink technology to repaired Xbox360s or also in new units.

So I'm very curious if this new heatsink is in my 360, which was repaired a few weeks ago. It's been running like a champ since, and I've always hoped they put in improved reliability measures, of couse I dont know that.

Anyway, it doesn't seem like the new sink would really make much difference.
 
in my opinion this doesnt make sense. If most of you are coming to the conclusion that an overheating GPU is the cause of the broken hardware message, how is it persisting and breaking so quickly as well as permanently? Most people keep saying they went to turn it on and it simply wouldnt work. Anyone whos played around with computer hardware knows that if something overheats shutting it off 5 minutes allows another period of problem free usage. Also as far as im aware nobody has cured their issues, at least not in mass numbers, by sticking the console next to a decent fan or simply removing the shell, both would do a hell of alot more then a little dinky peice of aluminum and copper heatpipe addition if that was indeed the problem for everyone.

In my opinion if this comes to fruition its to quite simply lower the temperature of the GPU, especially for those people that stick the console in a poorly ventilated area, not to resolve the universal 3 Read Light problem.

This is what I was wondering. I think the problem is more of what happens when it over heats though. I remember reading that the board would warp, and the X clamps couldn't hold them down after a certain period.

I suppose that even when it cools off, they don't go back into place. Hence the screws people are using these days.
 
To my knowledge the problem has to do with the solder connections under the GPU breaking. Whether this has to do with the x-clamps or not the end result is the same. Broken solder connections is actually a common problem with lots of electronics that have parts that get hot. Its amazing how many things you can fix just by finding the broken solder point and re melting them. The hot/cold cycle stresses the solder points until they eventually crack. The hotter it gets the more it stresses your solder points. Getting too hot may also cause the board to flex under the pressure of the x clamps thus putting more stress on the solder points. I think it's pretty obvious that the GPU gets really hot just by looking at the size of the heat sink. It can also handle being abused since the towel trick works.

I don't think the 360's cooling was designed very good to begin with since the dvd drive covers the GPU. If the GPU had a heat sink similar to the CPU I bet we wouldn't be having all these issues. I'm sure the added cooling will help but it does look like a band aid approach. They probably should have made one big heat sink like the PS3 has instead of 2 totally different ones. My guess is though they'll keep this design till the 65nm parts come out and then they'll stick with the same old heat sinks to keep costs down.
 
I think it's pretty obvious that the GPU gets really hot just by looking at the size of the heat sink. It can also handle being abused since the towel trick works.

Its actually pretty small and cheap for a passive heatsink for a GPU of that size in such a small enclosure. "Towel trick" does not work and is not a valid fix in my opinion. Im sure people could get their consoles to come back to life for a bit by dropping them onto concrete from a few feet, doesnt mean its a real fix.

This is what I was wondering. I think the problem is more of what happens when it over heats though. I remember reading that the board would warp, and the X clamps couldn't hold them down after a certain period.

I dont see that little thing removing nearly enough heat away from the board if its a warping problem right under the GPU contact points, and you still have the problem with the heat having no where to go, so the ambient temp within wont change hardly at all.
 
in my opinion this doesnt make sense. If most of you are coming to the conclusion that an overheating GPU is the cause of the broken hardware message.
I don't know that it is. It may be just a factor. Or this may be a step to get a quieter XB360 and reliability isn't affected?
 
Who wants to take bets that Gates had a system die and was highly, highly irritated?

Personally I would be more afraid of a highly irritated Ballmer...

In my opinion if this comes to fruition its to quite simply lower the temperature of the GPU, especially for those people that stick the console in a poorly ventilated area, not to resolve the universal 3 Read Light problem.

Universal? Anyway, they probably figured out most problems were related to the GPU.

I always thought it was heat related, because my mate (the only guy I know who burned through 360's like a pack of fags) had his machines in this really neat but small and closed (glass door) cupboard (sp?). The crashes he described to me looked like graphics problems. Just like the old stories with some of the demo units.

Anyway, it doesn't seem like the new sink would really make much difference.

Why bother then? They'd better done their homework right this time.
 
Last edited by a moderator:
Why bother then? They'd better done their homework right this time.
That's a very key point. I'd expect they're going to this expense for a reason and they researched it first. On the other hand maybe MS deal with hardware like thier software... :devilish:
 
Its actually pretty small and cheap for a passive heatsink for a GPU of that size in such a small enclosure. "Towel trick" does not work and is not a valid fix in my opinion. Im sure people could get their consoles to come back to life for a bit by dropping them onto concrete from a few feet, doesnt mean its a real fix.

I think you misinterpreted my sentence. Perhaps I should have said. You can tell the GPU gets really hot cause the current heat sink is small and cheap. I mentioned the "Towel trick" to point out how resilient the GPU is to getting really really hot. I never said it was a permanent fix. The GPU itself appears to be able to handle really high temperatures thus it itself is really hard to overheat therefore there is no "overheating" problem. BUT just because the GPU can handle it doesn't mean the motherboard/xclamps/solder can. Keeping the GPU cooler will keep the stress on everything else lower thus more reliable.

I just thought of something.. Most broken solder joints caused by heat stresses are due to too little solder at the joint. If the way the 360's are being fixed by Microsoft is just reheating of the GPU to remelt the solder underneath and not adding any new solder to the mix this may be the reason many people are on multiple 360's. Then again I'm no expert when it comes to BGA packages.
 
I always thought it was heat related, because my mate (the only guy I know who burned through 360's like a pack of fags) had his machines in this really neat but small and closed (glass door) cupboard (sp?). The crashes he described to me looked like graphics problems. Just like the old stories with some of the demo units.


That's what happened when my first 360 died; the screen would look like what happens with a video card is overclocked too high, lots of graphic anomalies. Once it would freeze up I could go back after letting it sit and it would work for ten minutes then do the same thing. Eventually it wouldn't turn on at all. It really seemed like a cooling issue then, now my second 360 just died and this one was completely different. It never showed any signs of overheating, it just wouldn't turn on one day. Both times the 360 was in a well ventilated area.
 
Universal? Anyway, they probably figured out most problems were related to the GPU.

universal, as in its been whats been causing the majority of problems. Just dont see that little heatsink doing anything. See my first post why i dont agree that its a single chip overheating.
 
Back
Top