I predict physics will win. Until we get a fundamentally different process, I don't know if we will see more than another decade or two of semi-crippled Moore's law (more cores instead of faster core is kind of cheating).
People jsut need to stop being skittish about popping a simple heatspreader off a chip. Its not like you're performing surgery. You take the thing off, grind off a few microns, then put is back on. No big deal. Temps drop so much you can run a 4670k at 4.4GHz on stock cooling. Thats what I expect anyway, I havent seen any actual data yet for haswell, there is too much focus on performance using $50+ coolers, which is kind of retarded. But I've seen some ivb results and that's about what you can expect.
I predict me buying whatever future top tier CPU is released.
Except...you are. IDC needed to use a tack hammer to drive a razor blade between the IHS and the PCB. He then dremeled away the excess glue. Both of those steps can cripple the CPU if you just slip an inch.
There is a alternate method which uses a vise grip and a block of wood. How does that work?
wrong.
because de-lidded Ivy does NOT "overclock like crazy", just the temps are greatly improved.
Yes, in percentage terms IVB overclockability is not as good as its predecessor SB.
i5-2500K was and is an absolute legend. What a tremendous value it looks in hindsight.
You really don't need a hammer. It takes just no more than 5. Minutes to open it w a razor. You would have to be really careless if you hit the die .Except...you are. IDC needed to use a tack hammer to drive a razor blade between the IHS and the PCB. He then dremeled away the excess glue. Both of those steps can cripple the CPU if you just slip an inch.
You really don't need a hammer. It takes just no more than 5. Minutes to open it w a razor. You would have to be really careless if you hit the die .
-delidding is great to run a $ 20.00 cooler on a $300+ cpu + mb lol.
-I'd like to have a dollar for every post -"my ib runs hot even after buying a hyper 212"
-I run @ 4.6 , just did a P95 blend test and no cores went over 68c. [batch 3219B389]
and the voltage is not really tweaked
-newer batches are even a lot better imo , as will haswell's batches they should become better.
With regards to IB-E, I think it's pretty much a guarantee that it will have solder instead of TIM, being a hexa core pretty much makes that a necessity. What type of overclocks people may achieve with IB-E, it's really hard to say at this point. I'd like to say it *may* be better than IVB, but then again you're running into the same problem that IB had - you're dissipating heat from a smaller area so the temps will definitely be higher. There's no way around that.
In reality
SB-E is a 8 core CPU 20mb L2
IVY-E is a 12 core CPU 30mb L3
They are going to disable half the cores. blah
I hope they will leave more then 15mb l3, if they left all of that cache intact on Extreme Edition, it would even have the same IPC as HW if not better. 6cores 30mb L3 would be awesome.
There are 6, 10 and 15 cores IB-E/EP/EX dies.
How much cache will the 6 core die have? Will IB-E have 2.5mb per core while the EX variant only 2mb?There are 6, 10 and 15 cores IB-E/EP/EX dies.
I really wish that they would use solder on IVY-E and that it could be run near 5GHz with decent temps. Then maybe I'd switch platforms. HW isn't compelling enough for me, 400$ for the CPU and 350$ for the mobo, nah, I'll pass.(local prices)
In reality
SB-E is a 8 core CPU 20mb L2
IVY-E is a 12 core CPU 30mb L3
They are going to disable half the cores. blah
I hope they will leave more then 15mb l3, if they left all of that cache intact on Extreme Edition, it would even have the same IPC as HW if not better. 6cores 30mb L3 would be awesome.
TDP is 130W on IVY E, so you can pretty much bank on solder being used. No idea about Ocing though. Intel using paste on low TDP cpus like haswell and ivy is the historic norm, intel chasing power efficiency for mobiles just allowed specs of haswell/ivy to qualify for cheaper paste tim. Multiple white papers in past, that describe need for intels indium 87 w/mk solder was based on higher TDP + high power density.
And Im in same boat, if IvyE overclocks well, I will likely change to enthusiast platform though get the $500 dollar version. If doesnt OC well, I will just go back to playing games...was excited to play with benches via new mobo/cpu, until haswell spoiled that mood.
My ivy benches 5 with 1.5 easy, and 4.7 prime stable with 1.3v, I dont mind playing silicone lottery but not when likely outcome is similar performance with worse oc.
Just FYI, since there are cringe-worthy conspiracy theories about TIM vs solder being perpetuated, it is most certainly not a cost issue. There really is not a difference between the two worth even noting. Additionally, it is not the historic norm to use TIM on mainstream processors - as many prior mainstream processors did not. However, the difference with the IVB is that it is a small and fragile part, and using solder on that type of process can potentially cause significant mechanical stress. The Haswell is even more delicate in this respect due to the integrated VR being on die - Being able to guarantee quality without issues from mechanical stress far outweigh the needs of extreme overclockers to intel, as people expect quality, stability, and longevity - especially their OEMs which sell processors/PCs in bulk.
My E4400 has TIM instead of solder.
That's because it has fragile 22nm die obviously.
Get a clue.
And sandy bridge was soldered. The hell is your problem? I could easily turn the same "get a clue" statement on you because you really don't know what you're talking about. But that won't stop you from making inflammatory attacks on me. Whatever, go for it, I could care less.