The saving is miniscule in comparison with the profit from inducing people to kill their chips via delidding attempts. Enthusiasts would gladly pay a few dollars to get efficient heat transfer so it wouldn't cost them anything other than the loss of the delidding profit.I suspect that Intel uses inferior TIM to save a little money
The saving is miniscule in comparison with the profit from inducing people to kill their chips via delidding attempts. Enthusiasts would gladly pay a few dollars to get efficient heat transfer so it wouldn't cost them anything other than the loss of the delidding profit.
The Broadwell and Skylake E chips that don't exist in the market. OK.Yet won't buy E-line chips which do have solder?
It's possible to look at a processor die separately from the heat bottleneck created by inefficient TIM.If they "are more than enough", then TIM must be sufficient, right.
16C stock and 20+ overclocked is huge to some, hence the commonness of the delidding practice, a practice that should not be nearly as common as it is.that really is not going to make a huge difference in the ultimate performance.
It doesn't seem like you did since you think the passage of time made the issue somehow no longer matter.I understood the furor when the switch was first made, but it is a 4 year old issue now
That was the same thing some people would have said to Agner about their compiler sabotaging AMD's processors:intel certainly is not going to change it.
Did the passage of time make the compiler issue no longer matter? Did Intel's refusal to change their compiler stop Agner from seeing it changed? No and no.agner said:When I started testing Intel's compiler several years ago, I soon found out that it had a biased CPU dispatcher. Back in January 2007 I complained to Intel about the unfair CPU dispatcher. I had a long correspondence with Intel engineers about the issue, where they kept denying the problem and I kept providing more evidence. I have later found out that others have made similar complaints to Intel and got similarly useless answers.
After Intel had flatly denied to change their CPU dispatcher, I decided that the most efficient way to make them change their minds was to create publicity about the problem. I contacted several IT magazines, but nobody wanted to write about it. Sad, but not very surprising, considering that they all depend on advertising money from Intel.
No, Intel continues to release new chips. It is only what consumers let them sell to them.So "it is what it is".
Time fallacy and hyperbole. Nothing is being "beaten to death". No chips are being harmed in this discussion nor is anyone forced to enter this thread. If anything, by beginning to once again use solder in quads for enthusiasts, chips would be helped rather than harmed.I dont really understand the point of beating the issue to death all over again.
Delidding a cheap APU is one thing but not only are i7s expensive the substrate has been thinned. This is why I voted for i7s having solder but i5s and lower having polymer TIM. If people want to risk an i5 then they can do that. Another option is the first poll one, where one SKU is soldered so people who have no interest in killing chips to get efficient thermal transfer don't have to deal with it. That option is less attractive because that SKU would doubtlessly be overpriced.AMD started using solder on Kaveri with the GV-A1 chips. I don't know if they soldered Carrizo, or if they will use solder on Bristol Ridge, but it's a nice step in the right direction for them. Though people that actually like going bare-die probably don't like the move, since it's harder to remove solder. It's all a matter of perspective.
I certainly didn't mind delid/relid on my 7700k, but avoiding that would save me a few hours of work (I lapped the IHS when I had it off).
Plus, those chips are red herring. We're talking about quad core chips. An i7 is more than enough for gaming and most other users' needs. The problem is the heat bottleneck.
It's not a problem at stock though. The TIM is more than sufficient for stock speeds.
A couple hundred mhz performance equivalent is usually what separates two generations of CPUs nowadays. (that and the USB upgrade we really buy it for ) However, people don't have to buy HEDT or competitor's products to teach Intel a lesson. All they have to do is... not buy: treat computing hardware like any other commodity (potatoes, furniture, deodorant).And even then, granted, they would probably run cooler, but only a couple hundred mhz more max overclock. Big whoop. -snip-
In any case, anyone who is offended by the use of TIM is certainly free to buy the HEDT line or buy those faster soldered chips from Intel's competitors.
PS: And I want a "free" Atom embedded board with my i5 - i7 purchase. If that Xiaomi guy gets one, why shouldn't I?
PSS: changed my mind, I think I'll buy an actual radio instead. It's vintage, and has better resale value.
PS: And I want a "free" Atom embedded board with my i5 - i7 purchase. If that Xiaomi guy gets one, why shouldn't I?
PSS: changed my mind, I think I'll buy an actual radio instead. It's vintage, and has better resale value.
Delidding a cheap APU is one thing but not only are i7s expensive the substrate has been thinned.
This is why I voted for i7s having solder but i5s and lower having polymer TIM. If people want to risk an i5 then they can do that.
Another option is the first poll one, where one SKU is soldered so people who have no interest in killing chips to get efficient thermal transfer don't have to deal with it. That option is less attractive because that SKU would doubtlessly be overpriced.
Well said my man. I'll gladly pay extra 5$ for soldered CPU than spend any minute voiding my warranty by delidding which in the end cost for me more than the precious metal inside the cpu by mass producing."Stop hating on Intel. Intel has some of the best engineers in the world when it comes to metallurgy. They know exactly what they are doing and the reason for conventional thermal paste in recent desktop CPUs is not as simple as it seems." - End of Quote
This fallacy is known as appeal to authority.
Intel folks of course know better than anyone what they are doing when it comes to Intel products, but the issue is that they have different interests than us. Intel's strategy is among other things to intentionally hamstring their desktop products to sell the more profitable and short lived mini/mobile PCs. TIM is also a cost cutting measure of course.
While people say that that the performance gap between low power processors and desktops is shrinking, the slowdown in node development means that the time gap, the time it would take for a mobile CPU to achieve performance parity to a desktop is actually drastically increasing.
Intel is also pushing Overclockers outside of mainstream onto the economically unattractive, crazy and soldered quad channel platform, something that can hardly be justified with typical PC use.
IIRC a few gen's back max was 15lb and 100PSI for the die.On cracking bare dies - Intel specifies 50 lbf for total pressure of cooler contact plate to IHS.
However, the pressure would be hard to police, so Intel would have to bundle their $100 insurance with every IHS-free chip they sold.
How about metallic glue?
http://techcrunch.com/2016/01/11/mesoglue-is-a-metallic-glue-that-replaces-hot-solder/
It's also a lot more than it would cost to get a chip with proper heat transfer, if Intel would deign to sell us what we actually want.Some guy came up with a Skylake delidding tool for just that reason. He charges something like $70 for it, but that's less than the cost of a dead 6700k.
16C is a problem for a chip that expensive.It's not a problem at stock though. The TIM is more than sufficient for stock speeds.
16C is a problem for a chip that expensive.
Interesting idea. I'd like to see some thermal conductivity data on it.
Metallic glue has a very high thermal conductivity (300-425 W/m*K ) as compared to common thermal greases (~1-10 W/m*K ) allowing much more effective heat transfer from the CPU to the heat sink
That sounds really great. It uses Indium and Gallium.
From their website:
Metallic glue has a very high thermal conductivity (300-425 W/m*K ) as compared to common thermal greases (~1-10 W/m*K ) allowing much more effective heat transfer from the CPU to the heat sink