Wednesday, October 26th 2011

Intel Sandy Bridge-E Can Reach Close to 5 GHz on Air-Cooling

As Intel's Core i7 "Sandy Bridge-E" processors in the LGA2011 package inch closer to their mid-November launch, there is already hectic activity among manufacturers of related components such as motherboards, memory, and coolers. By now, a large section of the industry has engineering samples to help design and test their components. OCWorkbench was witness to one such pre-release setup on which a Core i7 "Sandy Bridge-E" (unknown model, could even be quad-core for all we know), overclocked to 4.92 GHz with a "regular" air-cooler. The chip was idling at 45°C.

Sandy Bridge-E, as we know, can be effectively overclocked by increasing its base clock (BClk). On this particular setup, the BClk was set at 120 MHz, with a multiplier value of 41X, and core voltage of 1.51V. The memory used was DDR3-2400 MHz with CAS latency of 10T. This is particularly encouraging, not just to enthusiasts on a tight budget, but also the cooling products industry in general. Core i7 "Sandy Bridge-E" retail boxes don't contain a cooling solution, and Intel has been showing off its branded closed-loop water-cooling solution (to be purchased separately) as something that's "recommended" for Core i7 "Sandy Bridge-E". This gave many an impression that you need at least closed-loop water coolers for any hope of achieving decent overclocked speeds with these chips, and that perhaps these chips are bad overclockers in general. The likes of Xigmatek, Thermalright, Noctua, and Scythe can breathe a huge sigh of relief.
Source: OCWorkbench
Add your own comment

34 Comments on Intel Sandy Bridge-E Can Reach Close to 5 GHz on Air-Cooling

#26
cadaveca
My name is Dave
FreedomEclipsehmmm, come to think of it....wouldnt 1.5v be pretty dangerious to a 22nm CPU?? If anything about 1.4v hurts standard SB chips. 1.5v through a 22nm CPU cant be healthy
Personally, I don't think CPU Voltage is what has been killing chips. I'm more inclined to think IMC volts or PLL volts.

Standard SB chips are rated for 1.525v max VID under stock conditions, so it's most definitely NOT CPU voltage. Perhaps when combined with other things, sure.
Posted on Reply
#27
FreedomEclipse
~Technological Technocrat~
cadavecaPersonally, I don't think CPU Voltage is what has been killing chips. I'm more inclined to think IMC volts or PLL volts.

Standard SB chips are rated for 1.525v max VID under stock conditions, so it's most definitely NOT CPU voltage. Perhaps when combined with other things, sure.
But ive read threads about it on Xtreme forums about people runing 1.5v through their 2x00k's and the CPU ending up either dead of degraded within 2weeks despite having good cooling.
Posted on Reply
#28
cadaveca
My name is Dave
FreedomEclipseBut ive read threads about it on Xtreme forums about people runing 1.5v through their 2x00k's and the CPU ending up either dead of degraded within 2weeks despite having good cooling.
Yes, I have read those posts too. And nearly every single one was ALSO running PLL overvoltage, or high IMC volts, or were ALSO running high clocks(ie, more current) than what Intel has spec'd. Or they were NOT using Turbo to OC. The actual voltage REALLY has NOTHING to do with it. Current kills, not voltage.

There's a user here that ran 1.7v and killed his chip, too. Still doesn't mean it was the volts. ;)




Posted on Reply
#29
deleted
This is a clearly a suicide run. We heard the same story just before SB was released, and no 2500K or 2600K will ever be stable for 3 years+ at 5 GHz on air, regardless of the settings used to reach that OC. We're looking at roughly the same overclocking ability here that the 2600K has. The real point is despite the ability to fit in 50% more cores, there's no real decrease in maximum clock speeds and thermals aren't inordinately higher.
Posted on Reply
#30
[H]@RD5TUFF
Good news to me, I can't wait to get my hands on these!
Posted on Reply
#31
MikeMurphy
n-sterHow so?

AFAIK no official price are out, and even if the non-official ones were true, the 4 core is well priced, and the 6-core @ 500~600$ is understandable. A 4 core SB-E setup is getting dangerously close priced as a 2700K setup, and with many advantages that X79 offers
I apologize I didn't realize the prices I were reading were not yet confirmed.

But, 500-$600 is not understandable for adding 2 cores and cache. Price is almost double a 4-core 2600k. Sure you pay a premium for high end, but a premium was already factored in to the 2600k price and further increased on the chipset price which will be $$$. Total cost (which goes to Intel) for CPU and Mobo will be sky high.

It will be interesting this time around though as last time 1366 was released much in advance of 1156. It's reversed this time around. I wonder if many will see the value play in 2011.
Posted on Reply
#32
n-ster
deletedThis is a clearly a suicide run. We heard the same story just before SB was released, and no 2500K or 2600K will ever be stable for 3 years+ at 5 GHz on air, regardless of the settings used to reach that OC. We're looking at roughly the same overclocking ability here that the 2600K has. The real point is despite the ability to fit in 50% more cores, there's no real decrease in maximum clock speeds and thermals aren't inordinately higher.
For all we know, theses results were done with the 4-core
MikeMurphyI apologize I didn't realize the prices I were reading were not yet confirmed.

But, 500-$600 is not understandable for adding 2 cores and cache. Price is almost double a 4-core 2600k. Sure you pay a premium for high end, but a premium was already factored in to the 2600k price and further increased on the chipset price which will be $$$. Total cost (which goes to Intel) for CPU and Mobo will be sky high.

It will be interesting this time around though as last time 1366 was released much in advance of 1156. It's reversed this time around. I wonder if many will see the value play in 2011.
I think the prices of the 6-core will come down around the time Ivy comes out. Perhaps a non-K version might be significantly cheaper. For now, the value is in the 4-core LGA 2011 if the speculated prices are true. 450$, I think, would be a fair price for a 6-core

P.S: OMFGBBQ 7000 posts :eek:
Posted on Reply
#33
SteelSix
Delta6326But 1.51Volts is crazy I would never put that through my chip. I don't like to go over 1.36V... I'm at 1.25
Yea that's my thinking with both CPU and GPU. I have voltage thresholds, leaving anything higher for the bleeding edge folks out there..
Posted on Reply
#34
nt300
Jarmani doubt they would struggle with 1.51 volts coursing through their veins....i doubt they would last very long either though!!
I was thinking the same thing. Core voltage at 1.51V will kill this CPU.
Posted on Reply
Add your own comment
Apr 19th, 2024 20:29 EDT change timezone

New Forum Posts

Popular Reviews

Controversial News Posts