Overclock.net › Forums › Intel › Intel CPUs › Haswell Overclocking Guide [With Statistics]
New Posts  All Forums:Forum Nav:

Haswell Overclocking Guide [With Statistics] - Page 1159

post #11581 of 19540
Quote:
Originally Posted by pruik6 View Post

Hello people,
Anyone tried with his 4770k with hyperthreading off? or is that a stupid question and is hyperthreading that good rolleyes.gif
I know from my corei7 920 it decreased 12 to 15 degrees celcius.(whenHT was off)
I have 4770k 4.5 core and uncore 4.5 ghz ,tried 4.6 core and 4.6 uncore and 4.7 core and 3.2 ghz uncore. 4.7ghz did felt insane smoothly in everything but the temps are to high on full load.
Around 83 celcius bit scared for that biggrin.gif

Score wise on Cinebench 4.6hgz and 4.6ghz uncore has better score then 4.7ghz and 3.2 and 3.5ghz uncore.
And my 4.5ghz core and 4.5ghz uncore little bit lessor on cinebench then the 4.7ghz.
I want to try hyperthreading off is it worth it to get the 4.7gigahertz without HT?

Sorry for my bad english, but this topic is very usefull and fun to watch other people speeds and settings and the start information is very nice written.
Thanks for that:thumb:

Some do disable HT, quite a few games & apps only use 4 threads or less. But having a processor with HT you will probably want to make a stable profile to save that has HT enabled for running things that can take advantage of HT, & a profile with HT off.
Depending on what you are running that you are trying to get maximum performance from, just have to reboot & load the profile to have the best of both worlds.
    
CPUMotherboardGraphicsRAM
i7 990x/3930k/3770k/4770k GB x58a-OC/Asus RIVE/x79 gd65/z77 mpower/z77 MV... gtx480s/gtx580s/7970s/680s/Titan Hyper/PSC/BBSE/Hynix/Samsung 
Hard DriveOptical DriveCoolingOS
HDDs/SSDs DVD Air/water/phase/copper pot Windowz 
MonitorKeyboardPowerCase
samsung 2320 Razer Lycosa ax1200 carboard box 
Mouse
Trackball 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
i7 990x/3930k/3770k/4770k GB x58a-OC/Asus RIVE/x79 gd65/z77 mpower/z77 MV... gtx480s/gtx580s/7970s/680s/Titan Hyper/PSC/BBSE/Hynix/Samsung 
Hard DriveOptical DriveCoolingOS
HDDs/SSDs DVD Air/water/phase/copper pot Windowz 
MonitorKeyboardPowerCase
samsung 2320 Razer Lycosa ax1200 carboard box 
Mouse
Trackball 
  hide details  
Reply
post #11582 of 19540
If cinebench is how you're judging performance. You will take a big loss approx 20% when you turn ht off.

In gaming or apps that cannot use ht the higher clocks will give you better performance. A few games can use it bf3,bf4 and crytek games come to mind.
Cable box
(24 items)
 
Phat Bits
(16 items)
 
Lament Box
(16 items)
 
CPUMotherboardGraphicsRAM
i7-4790K 4.8ghz @ 1.245v Asus Maximus Hero VI 4GB Asus Strix GTX 980 Gskil Trident 2400 CL10 
Hard DriveHard DriveHard DriveOptical Drive
128GB sandisk extreme SSD 2 tb Western digital media intel 530 128 gb SSD External LG BD RW 
CoolingCoolingCoolingCooling
EK Coolstream 360mm PE Rad  Gentle typhoon AP-15 X 3 Fractal Design 140mm 1000RPM X3 Fractal Design Venturi High Flow Series 140mm F... 
CoolingCoolingCoolingCooling
Bitspower DDC Heatsink Black EK Coolstream XTC 140mm Rad EK Supremecy MX  EK FC980 Strix Water Block 
CoolingCoolingOSKeyboard
swiftech MCP-35X EK XRES 140 DDC Top WIN 7 ULTIMATE smk link 
PowerCaseMouseOther
Seasonic G-750  Fractal Design R5 G700s HDHOMERUN 3 TUNER CABLE CARD 
CPUMotherboardGraphicsRAM
4770k 4.5ghz @1.29v asus z87-A EVGA 4gb GTX 970 FTW 1510mhz 8000mem 1.23v Adata XPG 8GB @ 2400mhz 11-13-13-35-1 
Hard DriveHard DriveCoolingCooling
OCZ VERTEX 3 90GB 1tb 7200 Seagate hybrid  D5 vario Primo chill 120mm D5 Compression Res 
CoolingCoolingCoolingCooling
Swiftech MCR-320 XP 360mm rad Swiftech MCR-320 XP 120mm rad EK Supreme LTX Cpu block Biitspower ACX GTX 970 Water block 
OSKeyboardPowerCase
win7 64bit ultimate G13 Seasonic M211 620 watt Nzxt Guardian 
  hide details  
Reply
Cable box
(24 items)
 
Phat Bits
(16 items)
 
Lament Box
(16 items)
 
CPUMotherboardGraphicsRAM
i7-4790K 4.8ghz @ 1.245v Asus Maximus Hero VI 4GB Asus Strix GTX 980 Gskil Trident 2400 CL10 
Hard DriveHard DriveHard DriveOptical Drive
128GB sandisk extreme SSD 2 tb Western digital media intel 530 128 gb SSD External LG BD RW 
CoolingCoolingCoolingCooling
EK Coolstream 360mm PE Rad  Gentle typhoon AP-15 X 3 Fractal Design 140mm 1000RPM X3 Fractal Design Venturi High Flow Series 140mm F... 
CoolingCoolingCoolingCooling
Bitspower DDC Heatsink Black EK Coolstream XTC 140mm Rad EK Supremecy MX  EK FC980 Strix Water Block 
CoolingCoolingOSKeyboard
swiftech MCP-35X EK XRES 140 DDC Top WIN 7 ULTIMATE smk link 
PowerCaseMouseOther
Seasonic G-750  Fractal Design R5 G700s HDHOMERUN 3 TUNER CABLE CARD 
CPUMotherboardGraphicsRAM
4770k 4.5ghz @1.29v asus z87-A EVGA 4gb GTX 970 FTW 1510mhz 8000mem 1.23v Adata XPG 8GB @ 2400mhz 11-13-13-35-1 
Hard DriveHard DriveCoolingCooling
OCZ VERTEX 3 90GB 1tb 7200 Seagate hybrid  D5 vario Primo chill 120mm D5 Compression Res 
CoolingCoolingCoolingCooling
Swiftech MCR-320 XP 360mm rad Swiftech MCR-320 XP 120mm rad EK Supreme LTX Cpu block Biitspower ACX GTX 970 Water block 
OSKeyboardPowerCase
win7 64bit ultimate G13 Seasonic M211 620 watt Nzxt Guardian 
  hide details  
Reply
post #11583 of 19540
Quote:
If cinebench is how you're judging performance. You will take a big loss approx 20% when you turn ht off.

Cine has mad ht scaling, if you single core score 200, then it's like 800 without ht, 1025 with it lol but a lot of stuff does not use 7-8 threads well enough to justify losing 200mhz clock speed, +5% for benching or even playing some cpu games

20% is easily blind testable though and 5% is not, so i use ht on most of the time
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
post #11584 of 19540
Quote:
Originally Posted by Cyro999 View Post

Cine has mad ht scaling, if you single core score 200, then it's like 800 without ht, 1025 with it lol but a lot of stuff does not use 7-8 threads well enough to justify losing 200mhz clock speed, +5% for benching or even playing some cpu games

20% is easily blind testable though and 5% is not, so i use ht on most of the time
each 100mhz is 3% (I got that number from the asus aisuite software)4.7ghz = 30% increase. So 200mhz vrs ht. I think ht wins unless you know majority of your work is lightly threaded.

Why not keep ht and increase cooling or delid? Thats what I would want.
Cable box
(24 items)
 
Phat Bits
(16 items)
 
Lament Box
(16 items)
 
CPUMotherboardGraphicsRAM
i7-4790K 4.8ghz @ 1.245v Asus Maximus Hero VI 4GB Asus Strix GTX 980 Gskil Trident 2400 CL10 
Hard DriveHard DriveHard DriveOptical Drive
128GB sandisk extreme SSD 2 tb Western digital media intel 530 128 gb SSD External LG BD RW 
CoolingCoolingCoolingCooling
EK Coolstream 360mm PE Rad  Gentle typhoon AP-15 X 3 Fractal Design 140mm 1000RPM X3 Fractal Design Venturi High Flow Series 140mm F... 
CoolingCoolingCoolingCooling
Bitspower DDC Heatsink Black EK Coolstream XTC 140mm Rad EK Supremecy MX  EK FC980 Strix Water Block 
CoolingCoolingOSKeyboard
swiftech MCP-35X EK XRES 140 DDC Top WIN 7 ULTIMATE smk link 
PowerCaseMouseOther
Seasonic G-750  Fractal Design R5 G700s HDHOMERUN 3 TUNER CABLE CARD 
CPUMotherboardGraphicsRAM
4770k 4.5ghz @1.29v asus z87-A EVGA 4gb GTX 970 FTW 1510mhz 8000mem 1.23v Adata XPG 8GB @ 2400mhz 11-13-13-35-1 
Hard DriveHard DriveCoolingCooling
OCZ VERTEX 3 90GB 1tb 7200 Seagate hybrid  D5 vario Primo chill 120mm D5 Compression Res 
CoolingCoolingCoolingCooling
Swiftech MCR-320 XP 360mm rad Swiftech MCR-320 XP 120mm rad EK Supreme LTX Cpu block Biitspower ACX GTX 970 Water block 
OSKeyboardPowerCase
win7 64bit ultimate G13 Seasonic M211 620 watt Nzxt Guardian 
  hide details  
Reply
Cable box
(24 items)
 
Phat Bits
(16 items)
 
Lament Box
(16 items)
 
CPUMotherboardGraphicsRAM
i7-4790K 4.8ghz @ 1.245v Asus Maximus Hero VI 4GB Asus Strix GTX 980 Gskil Trident 2400 CL10 
Hard DriveHard DriveHard DriveOptical Drive
128GB sandisk extreme SSD 2 tb Western digital media intel 530 128 gb SSD External LG BD RW 
CoolingCoolingCoolingCooling
EK Coolstream 360mm PE Rad  Gentle typhoon AP-15 X 3 Fractal Design 140mm 1000RPM X3 Fractal Design Venturi High Flow Series 140mm F... 
CoolingCoolingCoolingCooling
Bitspower DDC Heatsink Black EK Coolstream XTC 140mm Rad EK Supremecy MX  EK FC980 Strix Water Block 
CoolingCoolingOSKeyboard
swiftech MCP-35X EK XRES 140 DDC Top WIN 7 ULTIMATE smk link 
PowerCaseMouseOther
Seasonic G-750  Fractal Design R5 G700s HDHOMERUN 3 TUNER CABLE CARD 
CPUMotherboardGraphicsRAM
4770k 4.5ghz @1.29v asus z87-A EVGA 4gb GTX 970 FTW 1510mhz 8000mem 1.23v Adata XPG 8GB @ 2400mhz 11-13-13-35-1 
Hard DriveHard DriveCoolingCooling
OCZ VERTEX 3 90GB 1tb 7200 Seagate hybrid  D5 vario Primo chill 120mm D5 Compression Res 
CoolingCoolingCoolingCooling
Swiftech MCR-320 XP 360mm rad Swiftech MCR-320 XP 120mm rad EK Supreme LTX Cpu block Biitspower ACX GTX 970 Water block 
OSKeyboardPowerCase
win7 64bit ultimate G13 Seasonic M211 620 watt Nzxt Guardian 
  hide details  
Reply
post #11585 of 19540
Quote:
Originally Posted by Wirerat View Post

each 100mhz is 3% (I got that number from the asus aisuite software)4.7ghz = 30% increase. So 200mhz vrs ht. I think ht wins unless you know majority of your work is lightly threaded.

Why not keep ht and increase cooling or delid? Thats what I would want.

Ideally you would just delid but if it's $400 to replace (UK) then you have to be quite confident with not breaking stuff. Until then, can't add the 0.13v without dropping ht as 1.4v + ht on air/clc without delid is too much to ask with comfortable temps

47/45 = 1.04444% (repeating of course biggrin.gif)

Considering my chip uses like 1.08v @3700mhz.. it's nowhere near tripping power limits, not even close, so just runs 3.7ghz - 3.9ghz depending on how heavy load is on all cores, so i'd consider 4.7 to be a ~27% frequency gain (over 3.7)
Edited by Cyro999 - 4/2/14 at 1:04pm
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
post #11586 of 19540
Exactly what i needed as answers,thanks for it you guys. That 2 profiles is a smart idea, i gonna do that.
Is there a game streaming setting benchmark or other good benchmark to see the difference between ht on and off as i did wrote above? (other then cinebench)

Wirerat: i have h100i and custom watercooling i to dumb for ,thats so diffecult to make wink.gif
post #11587 of 19540
A little update on my previous post... Warning: Spoiler! (Click to show)
Quote:
Originally Posted by angelotti View Post


It's true that it's about frequency and not volts, but under full frequency (above stock 3.4/3.5GHz) the cores receive full voltage (as per the instruction: manual input V or offset V or adaptive V)..., unless otherwise implemented by the manufacturer. In my case, all i need for the voltage to drop (under adaptive) is EIST and balanced 'power plan' (or as you said, for high, drop the min state to 5%)
The 'c-states' only drop the voltage further but should not be required to drop the voltage 'at all'. I for one, also have a "cpu overclock fixed mode" setting in the bios, and if that is enabled, there will be no VID drop.

If your concern (iamkraine) is power saving, then you should not worry to much about this, since the difference between idle at full voltage and idle at 0.7V will be only 2-5 watts and further 2-4 watts if you have all 'c-states' on. The power drow from the mains increases with load (at a given freq and voltage)
You can check this yourself with a 'wattmeter', or a multimetter (with amp-metter clamp) using W = A × V (watt = amp × volt).

If the concern is over degradation, again, it will be insignificant at idle+full voltage versus idle+lower voltage. You need lots of amps (caused by load) for max degradation.
The idea is, for max degradation you need high amps (caused by heavy load) which increases with frequency (and obviously you also need higher voltage for higher freq in order not to crash) and all this also generates allot of heat.

degradation = high freq at heavy load + high voltage + heat
the lower any of the three, the less degradation will occur.

ex1: 4.2GHz@1.3V@80°C is less degrading than 4.6GHz@1.3V@80°C
ex2: 4.2GHz@1.2V@80°C is less degrading than 4.2GHz@1.3V@80°C (but the difference is not as big as in the previous example)
ex3: 4.6GHz@1.3V@70°C is less degrading than 4.6GHz@1.3V@90°C (this is pretty obvious)
ALL OF THE ABOVE AT LOAD

If we were to talk about killing the chip and not degradation, then it's as intel said: "voltage kills the chip not the heat"

I took the trouble to check again since on my last post i spoke from memory , and here are the results (all at idle):

C-States OFF (ALL of them)
~77W - EIST/off + Adaptive
~79W - EIST/off + Manual
~57W - EIST/on + Adaptive (very close to C-States ON)
~61W - EIST/on + Manual

C-States ON (ALL of them)
~55W - EIST/off + Adaptive
~55W - EIST/off + Manual
~55W - EIST/on + Adaptive
~55W - EIST/on + Manual
(there were spikes to 57 every 2-3 seconds)

As you can see, with 'C-States' ON, there is no appreciable difference between adaptive vs manual or speedstep on/off.
Edited by angelotti - 4/3/14 at 1:59am
post #11588 of 19540
Makes sense, since the chip is mainly power gated at idle anyway with the C states enabled. I had mine off the other day troubleshooting and it was using 30W or so more than normal at idle, if I remember correctly.
post #11589 of 19540
Finally found the culprit of my overclocking. XMP enabled has been causing frequent BSOD/screen freezes during stability testing.
On a whim- I decided to change the profile to manual with DRAM Frequency at 1600mhz. I'm seeing voltage differences of up to 0.700v and about 10 deg celsius temp drop...
I was only getting 4.4ghz @ 1.320v (63celsius max core temp) stable. Now I'm on 4.4ghz @ 1.250v (53celsius max core temp) stable going 12hours on prime95.

Now my question is- could it be the ram? I have 2x4gb corsair vengeance 1600mhz that I purchased used off eBay.
I'm going to conduct a memtest (which I should have already done) this weekend.
post #11590 of 19540
Quote:
Originally Posted by pruik6 View Post

Exactly what i needed as answers,thanks for it you guys. That 2 profiles is a smart idea, i gonna do that.
Is there a game streaming setting benchmark or other good benchmark to see the difference between ht on and off as i did wrote above? (other then cinebench)

Wirerat: i have h100i and custom watercooling i to dumb for ,thats so diffecult to make wink.gif

x264 is the encoder that you would be using for CPU encoding for livestreaming, perf difference is approx 15-20% now. You need an up to date benchmark as the 5.0.1 package is out of date - the thread OP should have something, though it might be stability test settings instead of reproduce-livestream-benchmark-settings
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
Insert Name Here
(14 items)
 
  
CPUMotherboardGraphicsRAM
6700k Asus Maximus VIII Hero Gigabyte Aorus Xtreme 1080ti Corsair LPX 2x8GB 3200c16 
Hard DriveHard DriveCoolingOS
Old Seagate HDD Samsung 850 EVO Thermalright Silver Arrow SB-E SE Windows 7 Home Premium 64 bit 
MonitorKeyboardPowerCase
Asus PG258Q (240hz + Gsync) WASDKeyboards.com v1 semi custom w/ mx browns, ... Superflower Golden Green HX550 Air540 
MouseMouse Pad
Logitech G Pro Qck+ 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › Haswell Overclocking Guide [With Statistics]