Overclock.net › Forums › Intel › Intel CPUs › More Ivy Bridge benchmarks - Sandybridge comparison - 3770K vs 2600K (Performance, temps etc) "Couple Of LN2 scores are up"
New Posts  All Forums:Forum Nav:

More Ivy Bridge benchmarks - Sandybridge comparison - 3770K vs 2600K (Performance, temps etc) "Couple Of LN2 scores are up" - Page 60

post #591 of 690
Quote:
Originally Posted by alancsalt View Post

What do you use to read msr?
More specifically, is there something written for Win7?

MSR editor. Cystalcpuid under function has msr editor. Then you need to know which bit to read (info is in intel white papers), then convert to decimal from hex. I played with that msr editor and one written by Unclewebb (RT developer), but havent in a while, so dont know which bit for sandy/ivy. But info is in white papers on intels site.

But basically, all realtemp and coretemp are doing is reading it straight from cpu.

Hwmonitor accesses temps via PECI interface, which gives a rolling average of the highest core temp (per intel white papers) which is why HWmonitor gives temps 2-3C lower at load, and sometimes 1-2C higher at idle than realtemp and coretemp which read directly from MSR on cpu, ie real time measurements instead of hwm rolling avg.

here is directly from ivy bridge white paper, page 55:
6.2.3 Digital Thermal Sensor
Quote:
Each processor execution core has an on-die Digital Thermal Sensor (DTS) which
detects the core’s instantaneous temperature. The DTS is the preferred method of
monitoring processor die temperature because
•It is located near the hottest portions of the die.
•It can accurately track the die temperature and ensure that the Adaptive Thermal Monitor is not excessively activated.
Temperature values from the DTS can be retrieved through
•A software interface using processor Model Specific Register (MSR).
•A processor hardware interface as described in Chapter 7, “PECI Interface”.
Note:When temperature is retrieved by processor MSR, it is the instantaneous temperature of the given core. When temperature is retrieved using PECI, it is the average of the highest DTS temperature in the package over a 256 ms time window.

Edited by opt33 - 4/23/12 at 6:12pm
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
My System
(15 items)
 
  
CPUMotherboardGraphicsRAM
6900K 4.4ghz  MSI X99A Krait Titan XP Gskill 32GB 3200 C14 
Hard DriveHard DriveCoolingOS
Samsung 960 Pro 1TB OS Samsung 850 pro 1TB games D5, XT45 280 rad, Nemesis 280 rad,  Windows 10 64bit 
MonitorPowerCase
Acer Predator XB271HU 144hz Seasonic Prime Titanium 850 Fractal Design R5 
  hide details  
Reply
post #592 of 690
Could anybody say what's going on here, internet is full of reviews like this:

http://www.neoseeker.com/Articles/Hardware/Reviews/Intel_Core_i7_3770K/4.html

where they say, IB is a great overclocker, even more than SB, ??¿?

but tweaktown has this:

http://www.tweaktown.com/reviews/4673/intel_core_i7_3770k_lga_1155_ivy_bridge_cpu_review/index13.html

somebody is lying to the point i'm confused here. this is goin too far from my understanding and simple English knowledges.
post #593 of 690
Quote:
Originally Posted by Sujeto 1 View Post

Could anybody say what's going on here, internet is full of reviews like this:
http://www.neoseeker.com/Articles/Hardware/Reviews/Intel_Core_i7_3770K/4.html
where they say, IB is a great overclocker, even more than SB, ??¿?
but tweaktown has this:
http://www.tweaktown.com/reviews/4673/intel_core_i7_3770k_lga_1155_ivy_bridge_cpu_review/index13.html
somebody is lying to the point i'm confused here. this is goin too far from my understanding and simple English knowledges.

From what I have seen, results with IB are far more of a crapshoot than they were with SB. So far it looks that, if you get lucky, then you could get one that clocks well, but if you're unlucky you could get one that fails horribly.

http://www.tomshardware.com/reviews/ivy-bridge-benchmark-core-i7-3770k,3181-9.html

Tom's had 3 individual samples and there was quite a lot of difference between those 3 alone. I think that overall IB won't be as good a clocker, but if you get lucky you could find a few golden Ivy chips, but we won't see the same situation we did with Sandy, where every chip will do 4.5 at a reasonable temp. It seems to be that there are a lot that are performing worse than Sandy, some that are performing better and a lot which are just in the middle; with a lot more variance either side than we have seen currently.

This is likely due to the fact that this is the first use of 22nm and the first use of Tri-Gate technology and we will probably see a lower variance as the production continues and Intel manages to achieve greater consistency with their chips. That said, if they do release a six-core Ivy, as some are speculating on due to the 95W TDP rating despite the actual 77W consumption; Ivy could turn out not half bad after all!

EDIT: I misread that page in Tom's, I didn't realise the last screenie was a 2700K redface.gif But I still think that all the benchmarks etc. are being quite honest, it just seems that Ivy is a very unpredictable beast at this point in time and it's probably due to the new manufacturing process and I think we will see it improve on the whole as time goes on, the average quality will probably rise, but I imagine that the top-end won't change much as far as OC goes.
Edited by DirektEffekt - 4/23/12 at 9:13pm
My System
(21 items)
 
  
CPUMotherboardGraphicsGraphics
i7 4790K 4.8GHz @ 1.306v (Actual) ASUS Maximus VII Gene Gigabyte GTX 780 Galaxy GTX 780 
RAMHard DriveHard DriveCooling
16GB G.Skill Ares (4x4GB @ 1866MHz OC'd to 2133) Samsung 840 Evo 500GB WD Black 2.5" 750GB EK Supremacy EVO Acetal 
CoolingCoolingCoolingCooling
Aquacool NeXXos UT60 240mm Radiator EK XE 360 Radiator x 2 EK-FC780 GTX Ti Acetal /w Backplate and Reinfor... 8 x EK Vardar F4 
CoolingOSMonitorMonitor
EK MOSFET ASUS M7G Acetal Nickel Windows 8 x64 Pro Dell U2711 (2560*1440) Dell P2715q 
KeyboardPowerCase
Ducky Shine 3 Year of the Snake Edition Corsair AX1200 Thermaltake Cube X2 
  hide details  
Reply
My System
(21 items)
 
  
CPUMotherboardGraphicsGraphics
i7 4790K 4.8GHz @ 1.306v (Actual) ASUS Maximus VII Gene Gigabyte GTX 780 Galaxy GTX 780 
RAMHard DriveHard DriveCooling
16GB G.Skill Ares (4x4GB @ 1866MHz OC'd to 2133) Samsung 840 Evo 500GB WD Black 2.5" 750GB EK Supremacy EVO Acetal 
CoolingCoolingCoolingCooling
Aquacool NeXXos UT60 240mm Radiator EK XE 360 Radiator x 2 EK-FC780 GTX Ti Acetal /w Backplate and Reinfor... 8 x EK Vardar F4 
CoolingOSMonitorMonitor
EK MOSFET ASUS M7G Acetal Nickel Windows 8 x64 Pro Dell U2711 (2560*1440) Dell P2715q 
KeyboardPowerCase
Ducky Shine 3 Year of the Snake Edition Corsair AX1200 Thermaltake Cube X2 
  hide details  
Reply
post #594 of 690
Quote:
Originally Posted by -XOR- View Post

Ya, little bit of a misquote from Psyco.
Can only be told with time. It's the metrics and someone actually doing it. I agree 90°C 24/7 sounds nasty bad and I don't think it's plausible, but if the architecture can stand it for several years without degradation then it's a moot point. Should warm up your room/office to boot in winter wink.gif. Even if it does degrade after 2 years, pretty much everyone on these (enthusiast/overclocker) forums will be likely to swap out processors quicker than on a 2 year cycle, so might as well cane the CPU whilst you have it in your rig if it lets you - just hypothetically you understand and not saying IB will -
XOR

Did Intel make some advance in the conductive interconnects between the transistors? That is where electromigration kills the chip - eroding away the conductors. If not then it may not last 2 years at 90c I would think.

Since the die is smaller, the conducting paths in the chip are narrower, so I would think they would be damaged by electromigration sooner than they would be on the 35nm die given the same temperature. Unless there is some improvement here by Intel.
Edited by GeneO - 4/24/12 at 2:45am
post #595 of 690
I've seen another review on youtube and it's along the same lines as Tweaktown. As soon as bump up the voltage the heat goes through the roof. I think I'll be sticking with sandy for a while considering the performance increase is not that great but bloody hell the heat is.
Edited by K62-RIG - 4/23/12 at 9:45pm
post #596 of 690
Thread Starter 
I have realized that the thing with these chips is that temperature tests need to be done over an extended time. After 5min i saw 20c lower max temps then after 30min, even tho the water temp did not change much. The 6x120 GTX is able to keep it to ~4c air vs water, and the water temp only rises with ~2c during load. But it looks like the chip is slowly building up heat. With the 3770K's and HT @ 1.35v i got temps in the mid 50's after 5min, 10min was high 60's, 20min low 70's, and @ 25min it had jumped up to ~77c before i shut it down.

The tests that are showing prime runs on air/water after 5-6 min does not mean a thing. If they have air runs @ 5min with 80C it will for sure hit TjMax after 30min.
post #597 of 690
I'm wondering whether any of the better 3570's are lasered 3770's. not because of binning, but because they would have a greater surface area under the lid.

I'm surprised at the temp drop over time idea as the chip should be thermally soaked quite quickly given the wattage and the mass that's being spoken about unless silicon is an insulator, and a good one at that. sorry read eggy's the wrong way round
Edited by 13thmonkey - 4/24/12 at 1:05am
 
WHS
(6 items)
 
 
CPUMotherboardGraphicsRAM
3570k Asus P5B DL Wifi AP GTX470 Stock Crucial Ballistix 
Hard DriveHard DriveHard DriveCooling
Crucial M4 128GB WD 500GB RE4 OCZ Solid 30GB x2 Xigmatek Achillies 
OSMonitorMonitorPower
win 7 64 dell 2048x1152 HP 1050x1680 - portrait Seasonic S12 II 600W 
CaseMouseAudio
P180B  G500 Creative Xi-Fi Gamer 
CPUMotherboardGraphicsRAM
AMD Athlon II X2 255 M4A78LT-M-LE Onboard 2GB of DDR3 
Hard DriveCase
5TB of WHS storage (5TB before duplication) Fractal Define Mini 
  hide details  
Reply
 
WHS
(6 items)
 
 
CPUMotherboardGraphicsRAM
3570k Asus P5B DL Wifi AP GTX470 Stock Crucial Ballistix 
Hard DriveHard DriveHard DriveCooling
Crucial M4 128GB WD 500GB RE4 OCZ Solid 30GB x2 Xigmatek Achillies 
OSMonitorMonitorPower
win 7 64 dell 2048x1152 HP 1050x1680 - portrait Seasonic S12 II 600W 
CaseMouseAudio
P180B  G500 Creative Xi-Fi Gamer 
CPUMotherboardGraphicsRAM
AMD Athlon II X2 255 M4A78LT-M-LE Onboard 2GB of DDR3 
Hard DriveCase
5TB of WHS storage (5TB before duplication) Fractal Define Mini 
  hide details  
Reply
post #598 of 690
Quote:
Originally Posted by 13thmonkey View Post

I'm wondering whether any of the better 3570's are lasered 3770's. not because of binning, but because they would have a greater surface area under the lid.
I'm surprised at the temp drop over time idea as the chip should be thermally soaked quite quickly given the wattage and the mass that's being spoken about unless silicon is an insulator, and a good one at that. sorry read eggy's the wrong way round

3570s and 3770s are the exact same chip with the exact same surface area. HT is an option which is enabled or disabled from the factory depending on the war the chip is set up. I don't know exactly what they do to lock it out on a 3570, but the two chips are identical, the HT hardware is and always has been in the HT and non-HT chips and just disabled depending on which they are to be sold as. No physical modification is made, it is simply put into the package in such a way that it can't be used on the non-HT models.
My System
(21 items)
 
  
CPUMotherboardGraphicsGraphics
i7 4790K 4.8GHz @ 1.306v (Actual) ASUS Maximus VII Gene Gigabyte GTX 780 Galaxy GTX 780 
RAMHard DriveHard DriveCooling
16GB G.Skill Ares (4x4GB @ 1866MHz OC'd to 2133) Samsung 840 Evo 500GB WD Black 2.5" 750GB EK Supremacy EVO Acetal 
CoolingCoolingCoolingCooling
Aquacool NeXXos UT60 240mm Radiator EK XE 360 Radiator x 2 EK-FC780 GTX Ti Acetal /w Backplate and Reinfor... 8 x EK Vardar F4 
CoolingOSMonitorMonitor
EK MOSFET ASUS M7G Acetal Nickel Windows 8 x64 Pro Dell U2711 (2560*1440) Dell P2715q 
KeyboardPowerCase
Ducky Shine 3 Year of the Snake Edition Corsair AX1200 Thermaltake Cube X2 
  hide details  
Reply
My System
(21 items)
 
  
CPUMotherboardGraphicsGraphics
i7 4790K 4.8GHz @ 1.306v (Actual) ASUS Maximus VII Gene Gigabyte GTX 780 Galaxy GTX 780 
RAMHard DriveHard DriveCooling
16GB G.Skill Ares (4x4GB @ 1866MHz OC'd to 2133) Samsung 840 Evo 500GB WD Black 2.5" 750GB EK Supremacy EVO Acetal 
CoolingCoolingCoolingCooling
Aquacool NeXXos UT60 240mm Radiator EK XE 360 Radiator x 2 EK-FC780 GTX Ti Acetal /w Backplate and Reinfor... 8 x EK Vardar F4 
CoolingOSMonitorMonitor
EK MOSFET ASUS M7G Acetal Nickel Windows 8 x64 Pro Dell U2711 (2560*1440) Dell P2715q 
KeyboardPowerCase
Ducky Shine 3 Year of the Snake Edition Corsair AX1200 Thermaltake Cube X2 
  hide details  
Reply
post #599 of 690
Quote:
Originally Posted by Eggy88 View Post

I have realized that the thing with these chips is that temperature tests need to be done over an extended time. After 5min i saw 20c lower max temps then after 30min, even tho the water temp did not change much. The 6x120 GTX is able to keep it to ~4c air vs water, and the water temp only rises with ~2c during load. But it looks like the chip is slowly building up heat. With the 3770K's and HT @ 1.35v i got temps in the mid 50's after 5min, 10min was high 60's, 20min low 70's, and @ 25min it had jumped up to ~77c before i shut it down.
The tests that are showing prime runs on air/water after 5-6 min does not mean a thing. If they have air runs @ 5min with 80C it will for sure hit TjMax after 30min.

Check your PM inbox. wink.gif
    
CPUMotherboardGraphicsGraphics
Intel Core i7 7700k ASUS Maximus IX Formula EVGA GTX1080 Ti *soon* EVGA GTX1080 Ti *soon* 
RAMHard DriveHard DriveHard Drive
16GB Corsair Dominator Platinum 2666MHz CL16 Samsung 950 PRO M.2 512GB WD Red 4TB WD Green 1TB 
CoolingOSMonitorKeyboard
Corsair H110i GT Windows 10 PRO ASUS ROG Swift PG279Q Corsair K70 RGB MX Brown 
PowerCaseMouseMouse Pad
EVGA Supernova T2 850W Phanteks Enthoo Luxe Logitech G900 Chaos Spectrum Razer Destructor 2 
AudioAudioAudio
Beyerdynamic DT-770 PRO 80ohm JDS O2+ODAC Zalman MIC-M1 
  hide details  
Reply
    
CPUMotherboardGraphicsGraphics
Intel Core i7 7700k ASUS Maximus IX Formula EVGA GTX1080 Ti *soon* EVGA GTX1080 Ti *soon* 
RAMHard DriveHard DriveHard Drive
16GB Corsair Dominator Platinum 2666MHz CL16 Samsung 950 PRO M.2 512GB WD Red 4TB WD Green 1TB 
CoolingOSMonitorKeyboard
Corsair H110i GT Windows 10 PRO ASUS ROG Swift PG279Q Corsair K70 RGB MX Brown 
PowerCaseMouseMouse Pad
EVGA Supernova T2 850W Phanteks Enthoo Luxe Logitech G900 Chaos Spectrum Razer Destructor 2 
AudioAudioAudio
Beyerdynamic DT-770 PRO 80ohm JDS O2+ODAC Zalman MIC-M1 
  hide details  
Reply
post #600 of 690
Quote:
Originally Posted by DirektEffekt View Post

3570s and 3770s are the exact same chip with the exact same surface area. HT is an option which is enabled or disabled from the factory depending on the war the chip is set up. I don't know exactly what they do to lock it out on a 3570, but the two chips are identical, the HT hardware is and always has been in the HT and non-HT chips and just disabled depending on which they are to be sold as. No physical modification is made, it is simply put into the package in such a way that it can't be used on the non-HT models.

They are not the exact same... The i5 has 6mb L3 and the i7 has 8mb...
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
  • More Ivy Bridge benchmarks - Sandybridge comparison - 3770K vs 2600K (Performance, temps etc) "Couple Of LN2 scores are up"
Overclock.net › Forums › Intel › Intel CPUs › More Ivy Bridge benchmarks - Sandybridge comparison - 3770K vs 2600K (Performance, temps etc) "Couple Of LN2 scores are up"