Overclock.net › Forums › Intel › Intel CPUs › The Intel Devil's Canyon Owners Club
New Posts  All Forums:Forum Nav:

The Intel Devil's Canyon Owners Club - Page 2075

post #20741 of 22603
Quote:
Originally Posted by dmfree88 View Post

Feeling like a boss. 5.0ghz at 1.4v no issues (except the peak of my cooling capabilities I reached 91c stress testing).

Nice thumb.gif .
Quote:
Originally Posted by dmfree88 View Post

Any real benefits in graphic rendering speed or gaming FPS from increasing uncore clock further?

Practically none IMO, see heading Ring Bus Doesn't Matter [Evidence] in OP of this.
The Stilt Machine
(14 items)
 
XPS - i5 4690K
(12 items)
 
XPS - Q6600
(8 items)
 
CPUMotherboardGraphicsRAM
Ryzen 7 1700 Asus Crosshair VI Hero Sapphire Fury X@1145/545 G.Skill Trident Z 2x 8GB 3200MHz C14 
Hard DriveCoolingCoolingCooling
Samsung Evo 840 ThermalRight Archon IB-E X2 + 2x TY143 ThermalRight TY-143 2x front case intake Arctic Cooling F12 + 2x F9 as rear case exhaust 
OSMonitorKeyboardPower
Win 7 Pro x64 / Win 10 Pro x64 Asus MG279Q Cherry MX-Board 3.0 Cooler Master V850 
CaseMouse
SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
i5 4690K@4.9GHz 1.255V Asus Maximus VII Ranger Sapphire Fury X@1145/545 HyperX Savage 2x8GB 2400MHz C11 
Hard DriveCoolingOSMonitor
Samsung Evo 840 TR Archon SB-E X2 Win 7 Pro x64 / Win 10 x64 Asus MG279Q 
KeyboardPowerCaseMouse
Cherry MX-Board 3.0 Cooler Master V850 SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
Intel Q6600 G0 Asus P5K Premium Black Pearl Sapphire Toxic HD5850 Corsair Dominator 4GB 
Hard DriveCoolingOSPower
Crucial MX 100 256GB TR TRUE Rev.A Win 7 Pro x64 Cooler Master V650 
  hide details  
Reply
The Stilt Machine
(14 items)
 
XPS - i5 4690K
(12 items)
 
XPS - Q6600
(8 items)
 
CPUMotherboardGraphicsRAM
Ryzen 7 1700 Asus Crosshair VI Hero Sapphire Fury X@1145/545 G.Skill Trident Z 2x 8GB 3200MHz C14 
Hard DriveCoolingCoolingCooling
Samsung Evo 840 ThermalRight Archon IB-E X2 + 2x TY143 ThermalRight TY-143 2x front case intake Arctic Cooling F12 + 2x F9 as rear case exhaust 
OSMonitorKeyboardPower
Win 7 Pro x64 / Win 10 Pro x64 Asus MG279Q Cherry MX-Board 3.0 Cooler Master V850 
CaseMouse
SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
i5 4690K@4.9GHz 1.255V Asus Maximus VII Ranger Sapphire Fury X@1145/545 HyperX Savage 2x8GB 2400MHz C11 
Hard DriveCoolingOSMonitor
Samsung Evo 840 TR Archon SB-E X2 Win 7 Pro x64 / Win 10 x64 Asus MG279Q 
KeyboardPowerCaseMouse
Cherry MX-Board 3.0 Cooler Master V850 SilverStone TJ06 Logitech G700S 
CPUMotherboardGraphicsRAM
Intel Q6600 G0 Asus P5K Premium Black Pearl Sapphire Toxic HD5850 Corsair Dominator 4GB 
Hard DriveCoolingOSPower
Crucial MX 100 256GB TR TRUE Rev.A Win 7 Pro x64 Cooler Master V650 
  hide details  
Reply
post #20742 of 22603
I would say it goes like this:
#1 cores that do the hardwork
#2 memory that stores and transfers the data around, bandwidth and latency
#3 uncore/ring bus what ever the hell it is and sits around doing nothing biggrin.gif it is fast enough to transfer/communicate between the cores and memory and so on, there is almost no real benefit of increasing it's speed, I guess on older CPUs it was advantageous to increase this but with Haswell+ it's fast enough to not be a limit
post #20743 of 22603
Quote:
Originally Posted by LostParticle View Post


...

PS: @JackCY, mind sharing how you test for stability on your per-core OC? Per-core is the BEST overclock a person can use! (according to my personal & subjective opinion)


? smile.gif
    
CPUMotherboardGraphicsRAM
Intel Core i7-4790K ASRock Z97 OC Formula; ASUS MAXIMUS VII HERO; G...  EVGA GTX 1060 SC GAMING 6 GB G.SKILL F3-1866C8D-16GTX 
Hard DriveHard DriveHard DriveHard Drive
Samsung 850 EVO 250 GB SanDisk Extreme Pro 240GB Samsung 840 PRO 128GB Plextor PX-128M5S 128GB 
Hard DriveCoolingCoolingCooling
WD 1002FAEX 1TB Caviar BLACK (in enclosure) Noctua NH-D15S Corsair H110 2 x Noctua NF-A14 industrialPPC-2000/3000 PWM ... 
CoolingCoolingCoolingOS
Case Fan: 1 x Noctua NF-A14 PWM (rear, exhaust) Case Fan: Noctua NF-S12A PWM (front, intake) Case Fans: 2 x Noctua NF-S12A FLX (front, intake) Windows 10 Pro 64 bit 
OSOSMonitorMonitor
Windows 10 TP KDE Neon Developer Stable HP ZR2440w Dell U2311H 
KeyboardPowerCaseAudio
Microsoft Ergonomic keyboard 4000 Corsair AX760 (Platinum); EVGA SuperNova 750 G2 Corsair Air 540 Asus Xonar Essence STX II 
AudioOtherOther
Sennheiser HD 600 Aquaero 5LT  THRUSTMASTER GPX LIGHTBACK FERRARI F1 GAMEPAD 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel Core i7-4790K ASRock Z97 OC Formula; ASUS MAXIMUS VII HERO; G...  EVGA GTX 1060 SC GAMING 6 GB G.SKILL F3-1866C8D-16GTX 
Hard DriveHard DriveHard DriveHard Drive
Samsung 850 EVO 250 GB SanDisk Extreme Pro 240GB Samsung 840 PRO 128GB Plextor PX-128M5S 128GB 
Hard DriveCoolingCoolingCooling
WD 1002FAEX 1TB Caviar BLACK (in enclosure) Noctua NH-D15S Corsair H110 2 x Noctua NF-A14 industrialPPC-2000/3000 PWM ... 
CoolingCoolingCoolingOS
Case Fan: 1 x Noctua NF-A14 PWM (rear, exhaust) Case Fan: Noctua NF-S12A PWM (front, intake) Case Fans: 2 x Noctua NF-S12A FLX (front, intake) Windows 10 Pro 64 bit 
OSOSMonitorMonitor
Windows 10 TP KDE Neon Developer Stable HP ZR2440w Dell U2311H 
KeyboardPowerCaseAudio
Microsoft Ergonomic keyboard 4000 Corsair AX760 (Platinum); EVGA SuperNova 750 G2 Corsair Air 540 Asus Xonar Essence STX II 
AudioOtherOther
Sennheiser HD 600 Aquaero 5LT  THRUSTMASTER GPX LIGHTBACK FERRARI F1 GAMEPAD 
  hide details  
Reply
post #20744 of 22603
I've just ballparked it with x264 setting it to 1 or 2 threads and then force x264 to run on a specific core or two. Switch it around to test all the HW cores. Of course you can't run or do much else as the turbo limit can easily force to 2-3 cores state instead of your desired 1-2 state you are testing. Definitely not easy to say oh that is 100% stable without some automated tools. Even try using 1 less thread in x264 than what you want to test to avoid the turbo to falling down a speed.

Say I want to test 1 core = max speed, I will run 1 thread x264 and manually switch it between the HW cores.
2 cores, again 1 thread to avoid it slowing down, or 2 threads can work too if it doesn't slow down often.

1 core test is hardest as it often slows down, but with 2 core speed set the same as 1 core say 47-47-46-46 it doesn't slow down from 47 to 46 often at all on light usage or 1-2 thread high load.

So the trick is to load at max the number of cores you want to test and then assign the load to specific cores in process explorer or something else that can do it.

You could assign 16 thread x264 to just 1 core but I don't see the need for 16 threads switching on 1 core over and over since it's easy to limit x264 to 1 thread.
post #20745 of 22603
Quote:
Originally Posted by JackCY View Post

I've just ballparked it with x264 setting it to 1 or 2 threads and then force x264 to run on a specific core or two. Switch it around to test all the HW cores. Of course you can't run or do much else as the turbo limit can easily force to 2-3 cores state instead of your desired 1-2 state you are testing. Definitely not easy to say oh that is 100% stable without some automated tools. Even try using 1 less thread in x264 than what you want to test to avoid the turbo to falling down a speed.

Say I want to test 1 core = max speed, I will run 1 thread x264 and manually switch it between the HW cores.
2 cores, again 1 thread to avoid it slowing down, or 2 threads can work too if it doesn't slow down often.

1 core test is hardest as it often slows down, but with 2 core speed set the same as 1 core say 47-47-46-46 it doesn't slow down from 47 to 46 often at all on light usage or 1-2 thread high load.

So the trick is to load at max the number of cores you want to test and then assign the load to specific cores in process explorer or something else that can do it.

You could assign 16 thread x264 to just 1 core but I don't see the need for 16 threads switching on 1 core over and over since it's easy to limit x264 to 1 thread.


Thanks! So, you mean something like this?




I will read again your post because I do not fully understand it. You mean, you switch the Affinity during stress testing?
    
CPUMotherboardGraphicsRAM
Intel Core i7-4790K ASRock Z97 OC Formula; ASUS MAXIMUS VII HERO; G...  EVGA GTX 1060 SC GAMING 6 GB G.SKILL F3-1866C8D-16GTX 
Hard DriveHard DriveHard DriveHard Drive
Samsung 850 EVO 250 GB SanDisk Extreme Pro 240GB Samsung 840 PRO 128GB Plextor PX-128M5S 128GB 
Hard DriveCoolingCoolingCooling
WD 1002FAEX 1TB Caviar BLACK (in enclosure) Noctua NH-D15S Corsair H110 2 x Noctua NF-A14 industrialPPC-2000/3000 PWM ... 
CoolingCoolingCoolingOS
Case Fan: 1 x Noctua NF-A14 PWM (rear, exhaust) Case Fan: Noctua NF-S12A PWM (front, intake) Case Fans: 2 x Noctua NF-S12A FLX (front, intake) Windows 10 Pro 64 bit 
OSOSMonitorMonitor
Windows 10 TP KDE Neon Developer Stable HP ZR2440w Dell U2311H 
KeyboardPowerCaseAudio
Microsoft Ergonomic keyboard 4000 Corsair AX760 (Platinum); EVGA SuperNova 750 G2 Corsair Air 540 Asus Xonar Essence STX II 
AudioOtherOther
Sennheiser HD 600 Aquaero 5LT  THRUSTMASTER GPX LIGHTBACK FERRARI F1 GAMEPAD 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel Core i7-4790K ASRock Z97 OC Formula; ASUS MAXIMUS VII HERO; G...  EVGA GTX 1060 SC GAMING 6 GB G.SKILL F3-1866C8D-16GTX 
Hard DriveHard DriveHard DriveHard Drive
Samsung 850 EVO 250 GB SanDisk Extreme Pro 240GB Samsung 840 PRO 128GB Plextor PX-128M5S 128GB 
Hard DriveCoolingCoolingCooling
WD 1002FAEX 1TB Caviar BLACK (in enclosure) Noctua NH-D15S Corsair H110 2 x Noctua NF-A14 industrialPPC-2000/3000 PWM ... 
CoolingCoolingCoolingOS
Case Fan: 1 x Noctua NF-A14 PWM (rear, exhaust) Case Fan: Noctua NF-S12A PWM (front, intake) Case Fans: 2 x Noctua NF-S12A FLX (front, intake) Windows 10 Pro 64 bit 
OSOSMonitorMonitor
Windows 10 TP KDE Neon Developer Stable HP ZR2440w Dell U2311H 
KeyboardPowerCaseAudio
Microsoft Ergonomic keyboard 4000 Corsair AX760 (Platinum); EVGA SuperNova 750 G2 Corsair Air 540 Asus Xonar Essence STX II 
AudioOtherOther
Sennheiser HD 600 Aquaero 5LT  THRUSTMASTER GPX LIGHTBACK FERRARI F1 GAMEPAD 
  hide details  
Reply
post #20746 of 22603
Yes like that, it's best illustrated.
You can leave the threads at auto or just set them to 2 when testing 2 cores, shouldn't make much of a difference.
Yes I just switch it manually while the test is running or you can run some long tests and run each on different set of cores. Also you have HT so you can do 2 threads per core and select the HW threads for that core to be tested, just as you did it seems.

Of course compared to simply testing all cores at once (say 1h for that) on a single frequency you are in for:
4 single core tests = 4h
6 dual core tests = 6h
4 triple core tests = 4h
1 quad core test = 1h
Total = 15h

When you want to test all the possible combinations when you use per core OC. 15h vs 1h is quite a difference and 1h is quite short it's used just for simplicity of calculation.
But I just cheat it, I use my 46x4 stable, I can't do 47x4 stable without delid, so I set 47,46,46,46 which is hard to test due to turbo being lowered easily to 46 on all, so set it to 47,7,6,6 which keeps without lowering much more easily and quickly test that, and apply another cheat where any combination of two cores is enough for me lowering the number from 6 needed combinations to test down to 3, 1-2, 1-3, 1-4, and the triple down to 2 needed 1-2-3, 2-3-4, now when I set 47,7,7,6 it doesn't pass anymore, same with 48,7,6,6, and core #2 being the weakest and where the 48 won't pass and 47,7,7,6 being too much at the voltage I want to keep, leaving me with 47,7,6,6 as a highest usable option.
Using the shorter but still probably enough number of tests it's down to 4+3+2+1 = 10h instead of 15h.

But still if you run a proper test of 8h you are in for 8h vs 80h demand for testing biggrin.gif
And ain't nobody got time for that.

You can run Cinebench R15 on single core but it's more for fun than any stress testing, it does show a proper score for single core performance as long as you don't run anything else so the turbo mode stays on the max frequency for 1 core load. I could run 48,7,6,6 Cinebench R15 but otherwise it's not stable.
Edited by JackCY - 6/14/16 at 9:57am
post #20747 of 22603
@JackCY, okay thanks for all the information!

In my system, this specific chip, has completed 10 loops at 4.8 GHz and 5 loops at 4.9 GHz. Only five loops at 4.9 because back when I was testing it 5 loops of the x264 were (and still are) good enough for me. Very recently I've switched in running 10 loops. So, I am sure that my chip can run at least 4.8 GHz. It can do 4.9, as well, without any issues. These profiles were set and saved in my BIOS and they were used for quite some time, each, always doing all of my regular / everyday tasks, of course. This is why you "hear" me sounding so sure about them.

Anyway, I consider the Per-Core OC the best possible overclock a user can have. I used to set just one core active in the BIOS and test this first core only but now I realize this was not correct.

Anyway, I have been using this per-core, x 50 x50 x 49 x 48, cache x 44, for a couple of days now and I will keep using it until it will crash. So far, I have done everything I usually do on a daily basis without absolutely any issue. I have not gamed yet but tomorrow most probably I will install and play some of my games. I don't think I will face any issue. My Core Max has an average of 32 Celsius and a maximum of 60 C, with ambient of 28 C. My fans, shown in my screenshot above, spin around 1000 RPM and never exceed 1500 RPM, whereas my cooling solution for my entire system is just a Corsair H110 AIO mounted on the front panel of my chassis, as intake. No other fans are used in my system this period of an ambient of 28 Celsius. [I've also stressed my GPU with AIDA64 and the results were absolutely fine].

I really like, I adore, the simplicity of this system.

Thank you.
Edited by LostParticle - 6/14/16 at 12:31pm
    
CPUMotherboardGraphicsRAM
Intel Core i7-4790K ASRock Z97 OC Formula; ASUS MAXIMUS VII HERO; G...  EVGA GTX 1060 SC GAMING 6 GB G.SKILL F3-1866C8D-16GTX 
Hard DriveHard DriveHard DriveHard Drive
Samsung 850 EVO 250 GB SanDisk Extreme Pro 240GB Samsung 840 PRO 128GB Plextor PX-128M5S 128GB 
Hard DriveCoolingCoolingCooling
WD 1002FAEX 1TB Caviar BLACK (in enclosure) Noctua NH-D15S Corsair H110 2 x Noctua NF-A14 industrialPPC-2000/3000 PWM ... 
CoolingCoolingCoolingOS
Case Fan: 1 x Noctua NF-A14 PWM (rear, exhaust) Case Fan: Noctua NF-S12A PWM (front, intake) Case Fans: 2 x Noctua NF-S12A FLX (front, intake) Windows 10 Pro 64 bit 
OSOSMonitorMonitor
Windows 10 TP KDE Neon Developer Stable HP ZR2440w Dell U2311H 
KeyboardPowerCaseAudio
Microsoft Ergonomic keyboard 4000 Corsair AX760 (Platinum); EVGA SuperNova 750 G2 Corsair Air 540 Asus Xonar Essence STX II 
AudioOtherOther
Sennheiser HD 600 Aquaero 5LT  THRUSTMASTER GPX LIGHTBACK FERRARI F1 GAMEPAD 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Intel Core i7-4790K ASRock Z97 OC Formula; ASUS MAXIMUS VII HERO; G...  EVGA GTX 1060 SC GAMING 6 GB G.SKILL F3-1866C8D-16GTX 
Hard DriveHard DriveHard DriveHard Drive
Samsung 850 EVO 250 GB SanDisk Extreme Pro 240GB Samsung 840 PRO 128GB Plextor PX-128M5S 128GB 
Hard DriveCoolingCoolingCooling
WD 1002FAEX 1TB Caviar BLACK (in enclosure) Noctua NH-D15S Corsair H110 2 x Noctua NF-A14 industrialPPC-2000/3000 PWM ... 
CoolingCoolingCoolingOS
Case Fan: 1 x Noctua NF-A14 PWM (rear, exhaust) Case Fan: Noctua NF-S12A PWM (front, intake) Case Fans: 2 x Noctua NF-S12A FLX (front, intake) Windows 10 Pro 64 bit 
OSOSMonitorMonitor
Windows 10 TP KDE Neon Developer Stable HP ZR2440w Dell U2311H 
KeyboardPowerCaseAudio
Microsoft Ergonomic keyboard 4000 Corsair AX760 (Platinum); EVGA SuperNova 750 G2 Corsair Air 540 Asus Xonar Essence STX II 
AudioOtherOther
Sennheiser HD 600 Aquaero 5LT  THRUSTMASTER GPX LIGHTBACK FERRARI F1 GAMEPAD 
  hide details  
Reply
post #20748 of 22603
Quote:
Originally Posted by dmfree88 View Post

Feeling like a boss. 5.0ghz at 1.4v no issues (except the peak of my cooling capabilities I reached 91c stress testing). Might be able to bring the voltage down but I see no reason to run at 5.0 for more then playing around. Probably get a solid 4.7ghz daily OC and leave it.

http://valid.x86.fr/9qmcdm

Not sure what everyone uses for stress testing but I just ran through IBT 10 times and it seems to run fine at 5.0. Feels good to break 5.0ghz stable. This was something I could never accomplish on my FX-8350 with a similar cooler.

Use asus realbench. Try 10x h264 in a loop. After than 1-4 hours of stress test. Then you are ready to go thumb.gif
post #20749 of 22603
You guys really shouldnt run stress tests for a crazy amount of time anyways. i know most of us wont keep our chips longer then 5 years anyways but it does seriously decline the life of a chip.

Probably doesnt matter anyways but worth noting haha.
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
R7 1700X asus crosshair 6 Sapphire R9 390 Nitro CMK16GX4M2B3200C16 
Hard DriveHard DriveCoolingCooling
Kingston HyperX 3K Seagate ST2000DM006 EKWB Supremacy Evo EK-XRES 100 Revo D5 - Acetal 
CoolingCoolingCoolingCooling
Alphacool GPX r9 390 m01 alphacool xt45 240 Alphacool xt45 360 Alphacool D5 PWM 
OSMonitorKeyboardPower
Windows 10 LG 29UM68 Corsair Strafe Corsair RM1000 
CaseMouseMouse Pad
Thermaltake X5 Corsair M65 Pro RGB Razer Goliathus Speed 
  hide details  
Reply
post #20750 of 22603
And your VRM lifetime as well.
FDR4-4790k
(22 items)
 
  
CPUMotherboardGraphicsRAM
I7 4790k @ 4.8 GHz / 1.28V Asus Maximus VI Hero MSI GTX 1070 X 32GB g.skill Sniper 2400 MHz, 11-13-13-31/ 1.65V 
Hard DriveHard DriveHard DriveHard Drive
Samsung 850 PRO  Samsung 850 EVO 1TB Western Digitial Caviar Black 2 TB Western Digital Caviar Blue in USB 3.0 enclosure 
Hard DriveHard DriveCoolingCooling
Western Digitial Caviar Green in USB 3.0 enclosure Samsung 960 Pro 512 GB Noctua NH-D15 (only middle NF-A15 Chromax) 4 x Noctua NF-A14 Chromax black swap 
CoolingOSMonitorMonitor
Noctua NF-F12 Chromax black swap Windows 10 Pro 64 bit NEC LCD2490WUXI-SV DELL P2217H 
KeyboardPowerCaseMouse
Cooler Master Quickfire Rapid TKL Seasonic Prime Ultra Titanium 750W Fractal Design R4 Logitech G602 
Mouse PadAudio
perixx dx-2000 On Board SupremeFX 
  hide details  
Reply
FDR4-4790k
(22 items)
 
  
CPUMotherboardGraphicsRAM
I7 4790k @ 4.8 GHz / 1.28V Asus Maximus VI Hero MSI GTX 1070 X 32GB g.skill Sniper 2400 MHz, 11-13-13-31/ 1.65V 
Hard DriveHard DriveHard DriveHard Drive
Samsung 850 PRO  Samsung 850 EVO 1TB Western Digitial Caviar Black 2 TB Western Digital Caviar Blue in USB 3.0 enclosure 
Hard DriveHard DriveCoolingCooling
Western Digitial Caviar Green in USB 3.0 enclosure Samsung 960 Pro 512 GB Noctua NH-D15 (only middle NF-A15 Chromax) 4 x Noctua NF-A14 Chromax black swap 
CoolingOSMonitorMonitor
Noctua NF-F12 Chromax black swap Windows 10 Pro 64 bit NEC LCD2490WUXI-SV DELL P2217H 
KeyboardPowerCaseMouse
Cooler Master Quickfire Rapid TKL Seasonic Prime Ultra Titanium 750W Fractal Design R4 Logitech G602 
Mouse PadAudio
perixx dx-2000 On Board SupremeFX 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › The Intel Devil's Canyon Owners Club