• Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

Did Intel Make A Big Mistake With sandy Bridge?

Andy7

Andy7

Andy7

Associate
Joined
3 Jul 2014
Posts
626
Location
Cheshire
We all know that the Sandy Bridge chips were amazing when it came to overclocking. Some chips would hit 5Ghz on air.

Then Intel released Ivy Bridge and Haswell. Both were rather lame and disappointing as you didn't get much of a performance increase at stock and both would only overclock to around 4.2 to 4.4ghz at most.

The Devil's canyon chips promised 5ghz by Intel but failed big time 4.6ghz seems about tops for these chips.

Intel has gone down hill since Sandy Bridge in my opinion as they can't surpass it. Thoughts?
 
Ive never owned a sandybridge cpu, but have owned both ivybridge i5/i7 and several haswell i7's. All of theese have been a bit dissapointing clocking wise as theyre temp limited. The potential for high overclocks are there but you have to delid them to get there. They said that devils canyon would run cooler than previous haswell chips. Well that turned out to be a blatant lie, theyre worse than the chips they replaced.
 
in our minds more Ghz = better

& for some of us thats how we think when it comes to chips
but what intel has done is improve the IPC & improved the efficiency of the chips which has caused other factors like heat and design to get in the way, we would all love a current chip to hit 5 ghz they're only a few and haswell-e has gone back a few steps when it comes to Ghz but its a whole different ball game when it comes to cpus of current gen as I bet a 4.2ghz haswell is just as fast or on par as a 5ghz sany bridge

btw my devils canyon can do 4.9ghz @ 1.38v (non delided) with 2600mhz ram so theres still some out there
 
I sent an E-mail to Intel about the poor thermal paste they were using on the Ivy Bridge platform and that people were taking them apart and putting better stuff on.

I never got a reply strangely enough. As a huge company they are really lame.
 
SB crushed AMD now intel dont need to make big jumps in power because all the geek boy will but ever thing they put out for the 5% they promise.
also SB heat sink was soldered , newer chips have gone back to past and it cooks the chips at high OC's, the ppl that delid do ok on overclocks

i was told last night when asking for CPU spec to go out and buy a used 2500k
 
Minstadave said:
IPC increase has been pretty underwhelming tbh.

I don't think in 4.2ghz Haswell is as fast as 5ghz Sandy.
Click to expand...
Not totally sure about sandy, but my old 3770k at 4.5ghz was matched in benches by my 4770k @4.2ghz. Devils canyon is faster out of the box and does oc a bit better. 4.7ghz seems to be about average where later 4770k's were struggling to get over 4.3ghz.

Andy7 said:
I sent an E-mail to Intel about the poor thermal paste they were using on the Ivy Bridge platform and that people were taking them apart and putting better stuff on.

I never got a reply strangely enough. As a huge company they are really lame.
Click to expand...
It was indeed poor on ivy, but on haswell the addition of on die vrm's increased temps quite a lot more. Devils canyon was supposed to feature improved TIM and better power delivery to lower temps, this is a complete myth from what ive found. The two 4790k's ive owned ran hotter than my 4770k. I can remember Linus interviewing an intel rep on twitch tv and asking about the thermal issues on hw/ib, he basically skirted round the question. In regards to what new paste was being used the guy said, im afraid we cant go into details, it's like intels special sauce. Too right, must be hot sauce they use.:D
 
Not really it's just that Intel are more focussed on power efficiency since SB, the reason why Ivy Bridge onwards overclock poorly is probably because of the Tri-gate transistors.
 
setter said:
Yep, i remember the hype about how well they would supposedly clock on air. Tbh ive not seen that many hitting 5ghz for 24/7 use on reasonable voltage. Even on custom watercooling you will still be temp limited as water makes virtually zero difference over good air cooling on hw.
Click to expand...

:eek: custom water as the be better the high end air?
 
P4Clock said:
:eek: custom water as the be better the high end air?
Click to expand...
Due to the thick layer of paste used, and the adhesive that holds the ihs in place, the heat cant be transferred fast enough from the die. Hence why delidding works wonders. Much thinner layer of paste, (use coollabs pro/ultra) no thick layer of glue so the ihs sits closer to the die.
 
setter said:
Due to the thick layer of paste used, and the adhesive that holds the ihs in place, the heat cant be transferred fast enough from the die. Hence why delidding works wonders. Much thinner layer of paste, (use coollabs pro/ultra) no thick layer of glue so the ihs sits closer to the die.
Click to expand...

on the SB where the ihs in soldered it must be better tho
 
setter said:
Yep, i remember the hype about how well they would supposedly clock on air. Tbh ive not seen that many hitting 5ghz for 24/7 use on reasonable voltage. Even on custom watercooling you will still be temp limited as water makes virtually zero difference over good air cooling on hw.
Click to expand...

I was never temp limited with haswell under a custom loop even with more volts than i should have been throwing at it, just had an average chip that didn't scale very well :(
 
Yep, sb and all previous chips are soldered. As are the skt 2011 enthusiast range.

@nick, was your 4770k not delidded? From reading the massive DC owners thread on ocn, temps on the 4790k's are still extremely hot even on custom water. Certain stress programs are the worst culprits for high temps. Anything that uses AVX instructions will send temps skyrocketing at stock. Both my 4790k's have hit the mid 90's at stock under p95 ver 28.5. 1.264v on the first chip, 1.200v on the current one in sig.
 
Last edited:
The problem was SB was too good. Now they're just sitting back and raking it in until someone challenges them again. Not surprising, but still disappointing.

I think what everyone would love to see would be two (or more) closely-matched companies trading back and forth with every release. Shame it's just a dream.
 
You must log in or register to reply here.
Back
Top Bottom