Overclock.net › Forums › Industry News › Hardware News › [PCEvaluation] Intel i7 3770K Temperature Measured Without IHS
New Posts  All Forums:Forum Nav:

[PCEvaluation] Intel i7 3770K Temperature Measured Without IHS - Page 8

post #71 of 156
Quote:
Originally Posted by hazarada View Post

sure its a tick but a very odd one. For one this is the first tick in 8 years to increase IPC as in opposed to clock speed. Second the last 3 generations have increased the clock by at least 20% while the IPC increase on this yields performance benefits of around 10%.

You are too focused on the clock speed. Believe or not it is going to get even more difficult to squeeze extra performance out of silicon die, for both AMD and Intel. We have hit the point where Moore's law are about to be collapsed.
Misaka
(18 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5-3570K ASRock Z68 Extreme4 Gen3 Sapphire HD 7850 2 GB Samsung DDR3 16 GB (30 nm) 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 128 GB WD Caviar Blue 1 TB Lite-on DVD Burner Thermalright Venomous X 
OSOSMonitorKeyboard
Windows 7 Professional x64 (Host) Crunchbang Linux x64 (Guest) HP 2311x HP PS/2 Keyboard 
PowerCaseMouseAudio
Rosewill Capstone 450 W Rosewill Challenger Logitech M570 ASUS Xonar D1 
Other
Hauppauge HVR-1250 
  hide details  
Reply
Misaka
(18 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i5-3570K ASRock Z68 Extreme4 Gen3 Sapphire HD 7850 2 GB Samsung DDR3 16 GB (30 nm) 
Hard DriveHard DriveOptical DriveCooling
Crucial M4 128 GB WD Caviar Blue 1 TB Lite-on DVD Burner Thermalright Venomous X 
OSOSMonitorKeyboard
Windows 7 Professional x64 (Host) Crunchbang Linux x64 (Guest) HP 2311x HP PS/2 Keyboard 
PowerCaseMouseAudio
Rosewill Capstone 450 W Rosewill Challenger Logitech M570 ASUS Xonar D1 
Other
Hauppauge HVR-1250 
  hide details  
Reply
post #72 of 156
Quote:
Originally Posted by hazarada View Post

sure its a tick but a very odd one. For one this is the first tick in 8 years to increase IPC as in opposed to clock speed. Second the last 3 generations have increased the clock by at least 20% while the IPC increase on this yields performance benefits of around 10%.

The thing is, you'll never see large bumps in clock speed ever again. The only exceptions to this is a change to an architecture with a deeper pipeline, or the implementation of something that negates the high leakage at high clock speeds, like RCM.

Conquering leakage is going to take a massive breakthrough in material science, or a switch to an entirely different semiconductor, like graphene.
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom 9750 (stock) MSI MS-7548 (Aspen) HD 6950 @ 971/1387 1.25v 8GB DDR2 
Hard DriveOSMonitorPower
750GB Windows 7 64-bit ASUS VH238H 1920x1080 Seasonic X-650 Gold 
CaseMouseMouse Pad
Rosewill Smart One Razer Naga Razer Scarab 
  hide details  
Reply
My System
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom 9750 (stock) MSI MS-7548 (Aspen) HD 6950 @ 971/1387 1.25v 8GB DDR2 
Hard DriveOSMonitorPower
750GB Windows 7 64-bit ASUS VH238H 1920x1080 Seasonic X-650 Gold 
CaseMouseMouse Pad
Rosewill Smart One Razer Naga Razer Scarab 
  hide details  
Reply
post #73 of 156
Quote:
Originally Posted by trumpet-205 View Post

You are too focused on the clock speed. Believe or not it is going to get even more difficult to squeeze extra performance out of silicon die, for both AMD and Intel. We have hit the point where Moore's law are about to be collapsed.

Well we're still a while off before silicon is done but yes the end of the silicon era is on the horizon so to speak.
    
CPUMotherboardGraphicsRAM
6700k (EK Supremacy Evo) *4.7 @ 1.312 Asus Maximus VIII gene GTX 1080 FE (EK nickel/plexi)* 2075/5401 16gb Corsair 3200  
Hard DriveHard DriveHard DriveCooling
Samsung 950pro M.2 256gb Samsung 850 Evo - 512 - Steam Samsung 850 Evo - 512 - Other Apps EK Res 
CoolingCoolingCoolingOS
EK pump top XSPC EX 360 XSPC EX 240 10 
KeyboardPowerCaseMouse
Ducky Shine 3 TKL  EVGA 650 Fractal Arc Mini 2 G303 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
6700k (EK Supremacy Evo) *4.7 @ 1.312 Asus Maximus VIII gene GTX 1080 FE (EK nickel/plexi)* 2075/5401 16gb Corsair 3200  
Hard DriveHard DriveHard DriveCooling
Samsung 950pro M.2 256gb Samsung 850 Evo - 512 - Steam Samsung 850 Evo - 512 - Other Apps EK Res 
CoolingCoolingCoolingOS
EK pump top XSPC EX 360 XSPC EX 240 10 
KeyboardPowerCaseMouse
Ducky Shine 3 TKL  EVGA 650 Fractal Arc Mini 2 G303 
  hide details  
Reply
post #74 of 156
Sad thing is, enthusiasts have only themselves to blame, from what I remember, almost everyone of the big online forums/sites said "IF YOU HAVE A DECENT OC'ed SB THEN IVY IS NOT FOR YOU" the community heard that as "ZOMG IVY 5.5GHZ ON HYPER MASTER 212"


Quote:
Originally Posted by Homeles View Post

The thing is, you'll never see large bumps in clock speed ever again. The only exceptions to this is a change to an architecture with a deeper pipeline, or the implementation of something that negates the high leakage at high clock speeds, like RCM.
Conquering leakage is going to take a massive breakthrough in material science, or a switch to an entirely different semiconductor, like graphene.

The RCM has me intrigued. I know they are aiming for 4ghz+ with it, but the more I read on it the more it sounds like the enthusiast market won't like it. If it's constantly keeping energy "stored" in the chip like the hybrid car example some use then won't increases in voltage just create a perfect opportunity for something to go horribly wrong?
Edited by majin662 - 4/28/12 at 1:05pm
post #75 of 156
Quote:
Originally Posted by trumpet-205 View Post

We have hit the point where Moore's law are about to be collapsed.

Moore's "law" has nothing to do with it (being the doubling of inexpensive transistors on an IC every 18 months which is well on track and is predicted to stay there for at least another decade), its just intel (and to be fair amd too) has failed to make good use of it this year
post #76 of 156
Clock reached a point of diminishing returns years ago. Moore's Law lives on in increasing multiplicity of cores. It started out as transistor count increasing exponentially as a function of time, then processing power, which for single-cores meant Hz then word size and then IPC. All of those have hit the wall, and now it's all about the core count. Tacking on a GPU counts. And when silicon goes away it'll still be about the total zoot in the package doubling every 18 months. That they do that without increasing prices is phenomenal.

I'm not surprised IB has similar or higher temps than SB. It's using less power, but that power is in a smaller die, which means possibly more power density and definitely less surface to conduct it out to the case. The value in IB isn't that you can cool it lower and run it faster, it's that it doesn't take as much juice to run it.
post #77 of 156
Quote:
Originally Posted by trumpet-205 View Post


BD on the otherhand is a drastic changes, a "tock" by Intel standard.

BD is a Pentium 4 by Intel standards
Hex'd
(13 items)
 
  
Reply
Hex'd
(13 items)
 
  
Reply
post #78 of 156
i am pretty sure this can be calculated but i will over simplfy just for sake of comparison!
you got from 32 nm to 22 nm right so if you know the % shrink it means you will need this % more of cooling just to be on par with the previous gen!we are using techno of a 32nm solution on a 22 nm solution .it isnt hard to figure out that the heat removal capacity of a stock unit will do just fine since they compensated via other means i suspect like lower voltage, amp etc!but if it is even possible to put all setting on par with a 32 nm(wich is doubtfull cause the smaller you get the laws change yes intel adapted so there isnt to much bleed so technicly you could probably go to 32nm setting but then another factor jump in.cooling has to be upped to the same % as the shrink or so i suspect !i aint very good at mat conversion but the the 22 vs 32 is about 31.25% shrink!so in theory to have the same cooling effeciency or so ,you would need a cooler solution that is 31.25% more effecient at removing heat from the cpu.we arent talking to be better,but just to be = to a 32nm if everything is set as an sb!scale everything proportionally per oc application and you get the dilema we re in !
post #79 of 156
ivybridge is supposed to be launched with a price tag similar to it's predecessor, not being able to meet expectations, intel lowered it's price EVEN MORE before launch, that is an honest move IMHO, unlike amd launching faildozer with a huge price tag even though they know that it's a flop, and tried to blame it on the BIOS, windows etc.

ivybridge exceeds sandybridge's performance, however small the performance increase is, it's not close to being a flop. a flop is when it's predecessor is better in every aspect, including the pricing, like phenom ii > faildozer
post #80 of 156
Quote:
Originally Posted by drbaltazar View Post

i am pretty sure this can be calculated but i will over simplfy just for sake of comparison!
you got from 32 nm to 22 nm right so if you know the % shrink it means you will need this % more of cooling just to be on par with the previous gen!we are using techno of a 32nm solution on a 22 nm solution .it isnt hard to figure out that the heat removal capacity of a stock unit will do just fine since they compensated via other means i suspect like lower voltage, amp etc!but if it is even possible to put all setting on par with a 32 nm(wich is doubtfull cause the smaller you get the laws change yes intel adapted so there isnt to much bleed so technicly you could probably go to 32nm setting but then another factor jump in.cooling has to be upped to the same % as the shrink or so i suspect !i aint very good at mat conversion but the the 22 vs 32 is about 31.25% shrink!so in theory to have the same cooling effeciency or so ,you would need a coller solution that is 31.25% more effecient at removing heat from the cpu.we arent talking to be better,but just to be = to a 32nm if everything is set as an ib!scale everything proportionally per oc application and you get the dilema we re in !

it's not as simple as that bro.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Hardware News
Overclock.net › Forums › Industry News › Hardware News › [PCEvaluation] Intel i7 3770K Temperature Measured Without IHS