Dolphin, the GameCube and Wii emulator - Forums

10-20% performance increase

dannzen Wrote:give some information about the expected lifetime of a cpu silicium is really resistent to heat and aging

Its not uncommon for OC'd CPUs to become unstable at their OC'd frequency despite working fine for years.

Quote:Its not uncommon for OC'd CPUs to become unstable at their OC'd frequency despite working fine for years.

My C2D e6750 has been running a 3.2ghz overclock (up from 2.66ghz, that's huge for a C2D) for about four years now. It's still going strong.

(04-25-2012, 02:24 AM)bret emerald Wrote: [ -> ]The 10degrees C thing is really just a very rough rule of thumb that a lot of overclockers have used for years.

jeah with the athlon xp... pentium4...
that cpus are still running in my mums pc since 2002
7years over the warranty...

(04-25-2012, 02:24 AM)bret emerald Wrote: [ -> ]If Intel specify a cpu with Tcase of 73degrees, and you run it at stock voltage with average core temps of 73degrees over the course of its lifecycle, statistically speaking you should just make it through the warranty period (some cpus might be 3yrs, others 2yrs).

the warranty says nothing about the lifecycle of a product
an cpu can easily run 10-12years and more...
previously the mainboard or psu capacitors will explode and damaging the cpu... they have a limited time

my last notebook with an amd Turion was 7years old...
and runned constantly at 80°C (bad cpu, bad airflow, used it on my bed)
do you know why he is broken?
i slammed my fist on the keyboard because half of the keys didn't responed while i was writing my coursework
the display was also defective (while cleaning to much pressure)

i guess the cpu is still working... the mainboard has maybe a little crack i never opened it


you are speaking about "statistically" show me these...
microcontroller are running for decades...

(04-25-2012, 11:20 AM)NaturalViolence Wrote: [ -> ]

Quote:The i7 3770k runs significantly hotter than the i7 2600k and should therefore be labelled with a higher TDP, somewhere in the region of 125W. This can be seen by looking at the graph below from 'Tweaktown', if you assume the 2600k is 95W and then normalise the graph.

http://cdn5.tweaktown.com/content/4/6/46...review.png

I also find it very disappointing that a lot of popular review sites didn't even publish thermal results( cough 'Anandtech'). I also see no evidence so far to suggest that Ivy bridge is any more resilient to heat than Sandy Bridge, and is something we won't know until Intel publish the Tcase and Tjunction for the cpus. It's important to remember that even a 10degree C thermal differential in average cpu usage can halve its life expectancy.

One last point, Ivy bridge is supposed to be much better for notebooks, but as the owner of 2 Sandy bridge ones (one was given to me in work), the biggest problem I have with them is excessive heat and irritating fan noise. Surely Ivy Bridge will just make the problem worse.

Sorry for the rant, but as a consumer I wish companies like Intel were more transparent and honest about product information.

Temperature != TDP != heat generated

A cpu can generate more heat and still have a lower TDP if the max safe temperature is higher. Intel is correct in their use of the term TDP.

I have no doubt that you know temperature and heat are 2 very different things. But heating is a cumulative process, a 125W TDP cpu being cooled by a 77W TDP fan will eventually overheat (disregarding external factors such as case cooling and environmental loses etc.).

Most importantly the max safe temperatures are not higher (at least it is negligible). Intel have specified a Tjunction of 99degrees C for Ivy Bridge desktop cpus whereas Sandy Bridge was 98degrees C.

---

(04-25-2012, 01:22 AM)ExtremeDude2 Wrote: [ -> ]

(04-25-2012, 12:17 AM)bret emerald Wrote: [ -> ]

(04-24-2012, 10:46 PM)ExtremeDude2 Wrote: [ -> ][quote='bret emerald' pid='209921' dateline='1335264845']
Thermal design power has everything to do with how much thermal energy a cooling solution for the cpu will have to dissipate and NOTHING to do with how much you will save in your electricity bill.

I lol'd it DOES effect how much power is used in some way or another
I also got some reading for you

http://en.wikipedia.org/wiki/Thermal_design_power

Haha I think you need to reread the article that you linked to, it completely backs up the point that I was trying to make. Sorry if I'm being offensive but you are certainly not alone in your lack of knowledge on the subject (TDP is often a very misinterpreted term).

You didn't read the part I wanted you to :p

Quote:While a processor with a TDP of 100 W will almost certainly use more power at full load than a processor with a 10 W TDP, it may or may not use more power than a processor from a different manufacturer that has a 90 W TDP.

I've read the part that you quoted and once again it only consolidates what I was saying. I don't mean to be rude but I have no idea what the hell you're trying to tell me. A few words followed by an emoticon doesn't help much, I'm not a mind reader....

I want to change mine E8400, but now I doubt, Ivy Bridge is better for OC than Sandy or it is better on Stock?

Which is better for OC to 4.5Ghz:

Core i5 2500K (Sandy)
Core i5 3570K (Ivy)

And which is better on Stock (no OC):

Core i5 2500K (Sandy)
Core i5 3570K (Ivy)

(04-25-2012, 06:25 PM)dannzen Wrote: [ -> ]

(04-25-2012, 02:24 AM)bret emerald Wrote: [ -> ]The 10degrees C thing is really just a very rough rule of thumb that a lot of overclockers have used for years.

jeah with the athlon xp... pentium4...
that cpus are still running in my mums pc since 2002
7years over the warranty...

(04-25-2012, 02:24 AM)bret emerald Wrote: [ -> ]If Intel specify a cpu with Tcase of 73degrees, and you run it at stock voltage with average core temps of 73degrees over the course of its lifecycle, statistically speaking you should just make it through the warranty period (some cpus might be 3yrs, others 2yrs).

the warranty says nothing about the lifecycle of a product
an cpu can easily run 10-12years and more...
previously the mainboard or psu capacitors will explode and damaging the cpu... they have a limited time

my last notebook with an amd Turion was 7years old...
and runned constantly at 80°C (bad cpu, bad airflow, used it on my bed)
do you know why he is broken?
i slammed my fist on the keyboard because half of the keys didn't responed while i was writing my coursework
the display was also defective (while cleaning to much pressure)

i guess the cpu is still working... the mainboard has maybe a little crack i never opened it


you are speaking about "statistically" show me these...
microcontroller are running for decades...

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.

@hyperspeed
The 3570K isn't THAT much faster than a 2500K anyway. If you have the money for it, by all means, go for the 3570K.
In both scenarios, the 3570K will be a bit faster, but it will probably have a slightly higher temperature than the Sandy Bridge.

(04-26-2012, 12:19 AM)Garteal Wrote: [ -> ]@hyperspeed
The 3570K isn't THAT much faster than a 2500K anyway. If you have the money for it, by all means, go for the 3570K.
In both scenarios, the 3570K will be a bit faster, but it will probably have a slightly higher temperature than the Sandy Bridge.

Thanks @Garteal.

I think I'll stay with Sandy since here 3570k costs almost R$200,00, the most than 2500k.

$200 more than a 2500K? That's a lot. But I guess that's how the prizes are in Brazil.
It's only €30,- more expensive here.

The 2500K will be a great upgrade for that E8400.
All you have to do is add a decent cooler to it, and overclock it.