You are funny, really. You put an argument, that AMD TDP definition is ambiguous, and then you start to fight against your own argument as if we have put it here.
Stop the BS, nobody is saying that AMD TDP definition is ambiguous, quite the opposite, AMD TDP definition is very clear and straightforward:
Read #53. Once again TDP != power consumption for both AMD and Intel.
Read #53. Once again TDP != power consumption for both AMD and Intel.
I have given you two specific examples of ambiguity found in the definition used by Intel.
Analysis indicates that real applications are unlikely to cause the processor to consume maximum power dissipation for sustained periods of time
Intel is listing TDP numbers that are significantly lower than the actual maximum power draw of their CPUs. They are then relying on the fact that most applications barely use the CPU, assuming that it will remain idle most of the time. In the case that an application does max out the CPU for any period of time, Intel relies on their Thermal Monitor to automatically slow down the CPU when it becomes too hot to protect it from overheating.
You can all now discuss what "real applications" means and keep this pathetic thread going a bit longer.AMD, on the other hand, lists TDP numbers that are significantly higher than the maximum power draw of their CPUs.
That article is nearly a decade old, and thus from a timeframe when AMD made the more efficient CPUs and Intel the less efficient ones. And now it's the opposite, and we're seeing the same phenomenon in reverse.
http://forums.anandtech.com/showpost.php?p=35098358&postcount=23Our test of the new "Haswell" processors has revealed that under maximum load, extremely high temperatures are reached . While this glaring weakness in the desktop user can still easily conceal, the options are limited for micro PC quickly, like a visit to Gigabyte Computex show.
At an event the day before Computex 2013 on the 36th Floor of Taipei 101 were members of the press not only enjoy the beautiful view of the city, but also examine gigabytes of new products. In addition to countless Haswell motherboards also a micro-PC was issued, in which the latest technology was installed. When trying to take a closer look at the computer, we inevitably came into contact with the model and have reflexively directly let it fall again, so hot was the device. The product manager confirmed thermal problems with the model, but did not want to get carried away with the statement that these are also partly due to the partner Intel.
And has been explained numerous times, the issue isn't that TDP doesn't equal maximum power consumption, something nobody has debated. The issue is that TDP only has meaning within a specific temperature context, which Intel provides and AMD does not, and when a company refuses to provide an important metric, it's often because they are hiding something.
I've reviewed your posts in this thread and I'm not seeing it. If you really have identified an ambiguity, it would be more productive to summarize it here and/or point to a specific post, rather than just claiming you have already made the point.
From where I sit, there's nothing unambiguous in what Intel has said: TDP is the power level for which a cooling system must be designed to keep temperature within specific limits while running typical loads. This is not rocket surgery.
And even if "typical loads" is subject to some degree of interpretation, that doesn't change the fact that Intel is using TDP properly in correlating power to temperature, and AMD is not.
Overheating doesn't necessarily have anything to do with TDP. You can stay inside your TDP, but if the heat is not being transferred to the cooler quickly enough, the chip will throttle if it reaches the predefined threshold for safe operation. Just stop the fan on your CPU heat sink under medium load and you will see throttling although the CPU still operates well below its TDP value.
Overheating doesn't necessarily have anything to do with TDP. You can stay inside your TDP, but if the heat is not being transferred to the cooler quickly enough, the chip will throttle if it reaches the predefined threshold for safe operation. Just stop the fan on your CPU heat sink under medium load and you will see throttling although the CPU still operates well below its TDP value.
If the CPU starts to throttle then you have overcome the TDP.
If the CPU starts to throttle then you have overcome the TDP.
Overclocking, you turn off those protections.Why don't overclocked CPUs throttle, then? They may use 2-3x the power than what their TDP states - but that is no problem as long as temperatures are ok.
The article shows that Intel TDPs were unrealistic (underestimated) but I don't see Intel doing it better now.
In a previous reply to you, I identified two ambiguous words used by Intel in the definition of TDP. I do not think that repeating the words again will change anything.
Overclocking, you turn off those protections.
I never did. Just up voltage&frequency, done.
These relate to overclocking. Not relevant to a TDP discussion.
No, as I have explained. Heat can accumulate and make the chip hotter, leading to throttling.
And that happens because you have overcome the TDP limit. The Heat-Sink is not able to dissipate the excessive Thermal Power and your Temperature is starting to rise until it reaches TJunction Max and that will trigger the Throttling Mechanism of the CPU.
Not necessarily, imagine a 20W chip at idle with a huge passive heatsink.
Now remove the heatsink, the chip will remain at idle, there is zero chance to ever reach that 20W TDP.
However it will throttle or even shut down... because there is no heat transfer.
If you consider TDP as "maximum allowed power consumption" then the only possible way to exceed this is via overclocking.
If you consider TDP as "usual power consumption"... which is somewhat closer to what we see now with Intel's new chips and AMD's TDP...
Then yes, there are several ways to exceed that TDP and have the CPU throttle.
Sadly "New TDP" is linked to cooling, so a 45W chip that will not throttle under usual conditions, will throttle with the 38-41° you have in Athens.
So you should expect quite often to see something like "Intel 35W Chip".
The laptop designers decide to opt for a 30W cooling system, thinking "99% of our customers will never charge all 4 cores at 100%"
Now try running Prime95 or that same laptop in June in Athens....
It's even possible that the chip never draws more than 35W but shuts down at 33W used due to overheating.
You end up with an "under-designed laptop" + "uncommon load" + "athens summer".... well played !
So TDP discussions are always tricky