Dolphin, the GameCube and Wii emulator - Forums

Quote:I have to say the antagonism in these forums is overwhelming at times. I was decent enough to explain what I meant and yet you still try to undermine what I've said by telling me your Mum's P4 is still running after 10 years.

Firstly you've told me nothing about the pc's usage....was your Mum running Prime95 for 12 hours a day, or perhaps intensive sessions of Maya 3D every day.....or perhaps she was just using it sporadically to check emails and web browse? What about dust levels or the average cpu operating temps? (I'm sure you haven't been monitoring these for 10 years). Also like I said if the temps were on average 10degrees below Tc, you should expect the cpu to last at least 6 years. I never said a cpu couldn't last 10 or 12 years, in fact quite the opposite. If a cpu is run at lower temps it is more likely to last longer.

There will always be outliers in statistics, data that doesn't follow the general trend, and can largely be ignored in the broad scheme of things. Your Turion laptop still works but my Sony Vaio SB with i5 2410m cpu which was constantly being run at temps around 80 C failed after just 6 months. I would still expect almost all of the other Vaio SBs to make it through the 2 year warranty period. One data point means nothing.

A warranty period can tell you a lot about a product's life expectancy. High end PSUs come with at least a 5 year warranty because they know that under normal usage most of them will make it through( probably 95% or so). You have to remember that Intel will design for the worst case scenario (in this case someone constantly using the cpu at thermals close to Tcase), and draw the line at how many cpu's they are willing to have fail the warranty period. It does not make economic sense for any company to over engineer a product, but there will always be a safety margin.

The 'statistics' you ask for do not exist. Statistical analysis is an integral part of any product design and is largely based on lab tests and theoretical studies. Intel do not wait around for 10 years to find out how many Pentium 4s have failed before they decide on a warranty period for a product that they launched 10 years previously.

Besides I feel that this is starting to veer off topic.

You seem like a nice guy so I'll let you in on a tip. If you feel like holding on to your sanity it's best to ignore dannzens posts.

Natural Violence, you are the sole reason I want a like button on this forum, as you say things I feel I wouldn't get away with (although I don't have that much against danzen (except his tone of voice) you are brilliant at dealing with people who are wrong, for example by thinking desktop uses are fools (when really we are just poor, so don't buy portability we don't want or want the most powerful hardware)).

Hmm. Just got a bit of info on why the ivy bridges may be running hot.

http://www.overclockers.com/ivy-bridge-temperatures

So, the review sample they received (which is technically an engineering sample) uses thermal paste instead of solder to connect the die to the heat spreader on top. Baaaad news. Thermal paste, while decent in connecting a heatsink to the heat spreader, is very very poor compared to something permanent like solder. And the more thermal paste connections you have, the less effective the heat transfer becomes. The reduction in heat transfer would result in a build up of heat while under load: more or less what we've seen in the reviews. This wouldn't be the first time Intel has done this, sadly, but for the they wised up and stopped it back in the Core 2 Duo era, when they invented a solder specifically for this role. Why they would do it now is beyond me; solder is not expensive, and die is not supposed to be user serviceable anyway.



Man I hope that isn't in the retail chips >_< . There is some hope though. For one, the the reviews have been inconsistent. But most importantly, two of the three Overclock.net guys with leaked chips reported alot better temperatures than the reviews. Specifically Punceh, who did thorough testing. I'll show you.

http://www.overclockers.com Wrote:Where Sandy Bridge would often be around the 60 °C range at a 4.5 GHz overclock, Ivy Bridge has been tested to be in the 80-90 °C range.

http://www.overclock.net/t/1242313/more-ivy-bridge-benchmarks-sandybridge-comparison-3770k-vs-2600k-performance-temps-etc-couple-of-ln2-scores-are-up/50#post_16963132

Punceh Wrote:ive been testing with mine for a bit and i go up to about 60-65C @ 1.18V @ 4.5ghz.

Punceh Wrote:it went 69 peak on one core, 65-66 on the other cores peak on prime for an hour. no glitches..

So, it may very well be that they used paste on the engineering samples sent out to some reviewers, and solder on the retail chips. Punceh's chip is quite awesome, and he's able to hit 5ghz on air, something a Sandy bridge could never do. This could explain some of the contradictions I've been seeing...

Quote:Natural Violence, you are the sole reason I want a like button on this forum,

Please don't say things like that. Everyone with a brain knows that like buttons destroy any chance of intellectual discussion in a community. It encourages people to take on the opinions of the community and maintain the established order. People do whatever everyone else is doing, think whatever everyone else is thinking, and to top it off the upvoted comments/posts are usually read by more people as a result so they carry a higher weight and make the situation even worse. It turns the entire community into one big hivemind circle jerk. Anyone who doesn't support these opinions or attempts to debate someone who does is shot down immediately. Need evidence? Look what happened to youtube! When a new meme starts up watch how fast it becomes the top comment on every single video (like the arrow in the knee meme a few months back). Every wonder why that is? Well now you do.

People shouldn't need validation from the community to voice an opinion and be heard.

Quote:as you say things I feel I wouldn't get away with (although I don't have that much against danzen (except his tone of voice)

To be honest I'm amazing SS still hasn't banned me or at least warned me (except that one time I made the mistake of posting in one of squalls derailed threads, it was too tempting to pass up). I know he must be tempted from time to time.



(For the love of god don't remove this image, it's hilarious in context and you know it)

Quote:you are brilliant at dealing with people who are wrong, for example by thinking desktop uses are fools (when really we are just poor, so don't buy portability we don't want or want the most powerful hardware)).

If you say so *shrugs*. I'm not always right, I'm just good at making people defend their position and elaborate. Part of that has to do with my family upbringing. You people probably don't think I argue this way in real life because you don't see it a lot outside of the internet, but that's only because people rarely accuse me of being wrong or make me defend my position outside of the internet. If they did then this exact same style of debate would occur in verbal form.

Quote:Please don't say things like that.

It wasn't meant as a serious suggestion, because as you said, it would destroy any intellectual discussion.

I think all we could ever justify would be extremedude2's:

Quote:This^^

(04-26-2012, 04:47 AM)NaturalViolence Wrote: [ -> ]You seem like a nice guy so I'll let you in on a tip. If you feel like holding on to your sanity it's best to ignore dannzens posts.

Thanks for the tip, I would like to hold on to my sanity if possible I'm pretty new to these forums so it'll probably be a while before I can read people properly. I'm sure most comments are harmless, even if at first glance they appear not to be.

http://www.overclockers.com/ivy-bridge-temperatures

Quote:So why is Ivy Bridge hot?

Intel is using TIM paste between the Integrated Heat Spreader (IHS) and the CPU die on Ivy Bridge chips, instead of fluxless solder.

... that says everything...
cheap medium to prevent high clocks... and to sell later the "Extreme Editions" with higher clocks and soldered heat spreader

the same was done with
Core 2 Duos der E6000-Serie and E4000
the E4000 was pasted
the E6000 soldered

not really... this is not true...

you know the aging process of thermal paste?
2-3years and that cpu is dead (overclocked)

EATS SANDYBRIDGE FOR BREAKFAST

keep in mind, Dirt 2 is running with Tesselation enabled hence the higher score on SB's DX10.1 gpu.