Overclock.net › Forums › Intel › Intel CPUs › == Haswell-E Overclock Leaderboard & Owners Club ==
New Posts  All Forums:Forum Nav:

== Haswell-E Overclock Leaderboard & Owners Club == - Page 1383

post #13821 of 22233
Quote:
Originally Posted by michael-ocn View Post

One of my sticks seems not like the others, the 4G module in slot B1 sucks, I think that's an appropriate technical description of the problem.

Come to think of it, maybe it's the slot and not the stick? Any bets on whether reseating slot B1 or swapping which stick is in which slot would help?

This would be the easiest way to find out.
Edited by GreedyMuffin - 10/25/15 at 2:23pm
post #13822 of 22233
Jeez didnt know cache had such an impact on power draw and temps

currently testing a 5960x @ 4.6 1.264v on a RVE

Cache @ 4.2 1.150v
CPU Input @ 1.904 LLC8

was doing a x265 encode with staxrip and was pulling 430watts from the wall max temp on the hottest core was 76c

x265 process crashed so i thought let me reset the cache to default and retest

when i next did the encode the max power i can see from my power meter was 380-90 watts and the max temp is 71c on the hottest core

40watts and 6c is a massive difference might just leave cache at stock it still encoding at the same FPS as it was with the cache at 4.2
just not worth the power and temps and the cache volts wasnt even that high only 1.150 alot of people run 1.25+
SM8 Water
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 5960x @ 4.5 1.23v Rampage V Extreme GeForce GTX 1080 Ti 32GB G.Skill Trident Z 3200 CL14 
Hard DriveHard DriveCoolingOS
960 Pro 512 GB | 850 Pro 2TB 4TB WD4004FZWX | 4TB WD40EFRX Custom Water Windows 10 Pro 64bit 
MonitorKeyboardPowerCase
Asus ROG Swift PG279Q Corsair K95 RGB  EVGA 850 P2 Caselabs SM8  
MouseAudio
Razer Deathadder Chroma Soundblaster ZX 
  hide details  
Reply
SM8 Water
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 5960x @ 4.5 1.23v Rampage V Extreme GeForce GTX 1080 Ti 32GB G.Skill Trident Z 3200 CL14 
Hard DriveHard DriveCoolingOS
960 Pro 512 GB | 850 Pro 2TB 4TB WD4004FZWX | 4TB WD40EFRX Custom Water Windows 10 Pro 64bit 
MonitorKeyboardPowerCase
Asus ROG Swift PG279Q Corsair K95 RGB  EVGA 850 P2 Caselabs SM8  
MouseAudio
Razer Deathadder Chroma Soundblaster ZX 
  hide details  
Reply
post #13823 of 22233
Quote:
Originally Posted by Sem View Post

-snip- /quote]'

Well, on the die itself the cache got some size to it as you can see here. Warning: Spoiler! (Click to show)

So it`s quite obvious that overvolting such a part from stock should easily add some heat tongue.gif Stock voltage is 0.8V on Haswell-E if i remember correctly?) So 1.150 is actually quite a bit compared to stock. smile.gif

EDIT: Quote error tongue.gif
post #13824 of 22233
Been tinkering again today.
So how's this for a memory overclock, I used memory try it to set the timings, voltages are 1.350v, it did try to set the VCCSA to 1.35, but I manually set it to 1.050v, seems to be the sweet spot for my setup.
Still at 100Mhz BCLK, I can't set it to 125mhz no matter what I do, guess it's the Gaming 7 IDK..

Edit: Nevermind wasn't stable at all, couldn't even get Realbench to start a run..lol
Warning: Spoiler! (Click to show)


Still not convinced 4.4Ghz over 4.3Ghz is worth the extra voltages though:

4.4Ghz Warning: Spoiler! (Click to show)

4.3Ghz Warning: Spoiler! (Click to show)

Edited by schoolofmonkey - 10/26/15 at 2:59am
post #13825 of 22233
I had a 5930k for a year before i got this 5960x

Spent over 4 weeks overclocking my 5930k and thought i had haswell-e overclocking nailed down

Remember testing cache with the 5930k i had mine at 4Ghz @1.2 and it didnt make any signifcant difference on that cpu

but it seems overclocking a 5960x is a totally different beast its almost like starting from scratch

had it for 2 weeks already and when i think i have it nailed something like this pops up


Quote:
Originally Posted by GreedyMuffin View Post

Quote:
Originally Posted by Sem View Post

-snip- /quote]'

Well, on the die itself the cache got some size to it as you can see here. Warning: Spoiler! (Click to show)

So it`s quite obvious that overvolting such a part from stock should easily add some heat tongue.gif Stock voltage is 0.8V on Haswell-E if i remember correctly?) So 1.150 is actually quite a bit compared to stock. smile.gif

EDIT: Quote error tongue.gif
SM8 Water
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 5960x @ 4.5 1.23v Rampage V Extreme GeForce GTX 1080 Ti 32GB G.Skill Trident Z 3200 CL14 
Hard DriveHard DriveCoolingOS
960 Pro 512 GB | 850 Pro 2TB 4TB WD4004FZWX | 4TB WD40EFRX Custom Water Windows 10 Pro 64bit 
MonitorKeyboardPowerCase
Asus ROG Swift PG279Q Corsair K95 RGB  EVGA 850 P2 Caselabs SM8  
MouseAudio
Razer Deathadder Chroma Soundblaster ZX 
  hide details  
Reply
SM8 Water
(14 items)
 
  
CPUMotherboardGraphicsRAM
i7 5960x @ 4.5 1.23v Rampage V Extreme GeForce GTX 1080 Ti 32GB G.Skill Trident Z 3200 CL14 
Hard DriveHard DriveCoolingOS
960 Pro 512 GB | 850 Pro 2TB 4TB WD4004FZWX | 4TB WD40EFRX Custom Water Windows 10 Pro 64bit 
MonitorKeyboardPowerCase
Asus ROG Swift PG279Q Corsair K95 RGB  EVGA 850 P2 Caselabs SM8  
MouseAudio
Razer Deathadder Chroma Soundblaster ZX 
  hide details  
Reply
post #13826 of 22233
I think it's the amount of Cache that matters.

Best I was able to muster on Aida64 was about 65Gbits Read, 70 Gbits Write and 70Gbits Copy on a 4.6 Core, 4.4 Uncore at 3200 Memory at 16-17-17-CR1.

Read by the stops scaling below 70Gbits on a 6-Core 15MB Cache processor on my tests.
post #13827 of 22233
Sub timings are important too, I can get 80k read and copy quite easily now even with cache at 4.2Ghz, when I needed 4.4Ghz at least before some tweaking.
CaseLabs Panda
(29 items)
 
Bench table
(18 items)
 
Fractal White
(17 items)
 
CPUMotherboardGraphicsRAM
5960X Rampage V Extreme KFA² GTX 980Ti HOF Watercooled Kingston HyperX HX426C13PB2K4/16 
RAMHard DriveHard DriveCooling
G.Skill TridentZ F4-3200C14Q-32GTZ Kingston v300 Intel SSD 535 EK-Supremacy EVO - Full Nickel 
CoolingCoolingCoolingOS
EK-Coostream XE 360 EK-Vardar F4-120ER EK-XRES 140 D5 Vario (incl. pump) Win10 / Win 8.1 / Win 7 / Linux Mint 
MonitorKeyboardPowerCase
Dell U2515H Ducky Channel Shine III MX black - white LED /... Enermax Platimax 1000w dimastech easy v3 
MouseMouse Pad
logitech G9X HyperX 900x400 / Steelseries QCK 
  hide details  
Reply
CaseLabs Panda
(29 items)
 
Bench table
(18 items)
 
Fractal White
(17 items)
 
CPUMotherboardGraphicsRAM
5960X Rampage V Extreme KFA² GTX 980Ti HOF Watercooled Kingston HyperX HX426C13PB2K4/16 
RAMHard DriveHard DriveCooling
G.Skill TridentZ F4-3200C14Q-32GTZ Kingston v300 Intel SSD 535 EK-Supremacy EVO - Full Nickel 
CoolingCoolingCoolingOS
EK-Coostream XE 360 EK-Vardar F4-120ER EK-XRES 140 D5 Vario (incl. pump) Win10 / Win 8.1 / Win 7 / Linux Mint 
MonitorKeyboardPowerCase
Dell U2515H Ducky Channel Shine III MX black - white LED /... Enermax Platimax 1000w dimastech easy v3 
MouseMouse Pad
logitech G9X HyperX 900x400 / Steelseries QCK 
  hide details  
Reply
post #13828 of 22233
memory throughput is dependant on uncore frequency yes. People have tended to get a little too concerned with this on HWE though. There's plenty enough memory bandwidth on X99 for this to be a non issue as far as performance is concerned
post #13829 of 22233
Quote:
Originally Posted by Silent Scone View Post

memory throughput is dependent on uncore frequency yes. People have tended to get a little too concerned with this on HWE though. There's plenty enough memory bandwidth on X99 for this to be a non issue as far as performance is concerned
Indeed, except bigger number on Aida, I haven't seen much if any difference in normal use. Even on 3D benchmark, the difference is meaningless.
CaseLabs Panda
(29 items)
 
Bench table
(18 items)
 
Fractal White
(17 items)
 
CPUMotherboardGraphicsRAM
5960X Rampage V Extreme KFA² GTX 980Ti HOF Watercooled Kingston HyperX HX426C13PB2K4/16 
RAMHard DriveHard DriveCooling
G.Skill TridentZ F4-3200C14Q-32GTZ Kingston v300 Intel SSD 535 EK-Supremacy EVO - Full Nickel 
CoolingCoolingCoolingOS
EK-Coostream XE 360 EK-Vardar F4-120ER EK-XRES 140 D5 Vario (incl. pump) Win10 / Win 8.1 / Win 7 / Linux Mint 
MonitorKeyboardPowerCase
Dell U2515H Ducky Channel Shine III MX black - white LED /... Enermax Platimax 1000w dimastech easy v3 
MouseMouse Pad
logitech G9X HyperX 900x400 / Steelseries QCK 
  hide details  
Reply
CaseLabs Panda
(29 items)
 
Bench table
(18 items)
 
Fractal White
(17 items)
 
CPUMotherboardGraphicsRAM
5960X Rampage V Extreme KFA² GTX 980Ti HOF Watercooled Kingston HyperX HX426C13PB2K4/16 
RAMHard DriveHard DriveCooling
G.Skill TridentZ F4-3200C14Q-32GTZ Kingston v300 Intel SSD 535 EK-Supremacy EVO - Full Nickel 
CoolingCoolingCoolingOS
EK-Coostream XE 360 EK-Vardar F4-120ER EK-XRES 140 D5 Vario (incl. pump) Win10 / Win 8.1 / Win 7 / Linux Mint 
MonitorKeyboardPowerCase
Dell U2515H Ducky Channel Shine III MX black - white LED /... Enermax Platimax 1000w dimastech easy v3 
MouseMouse Pad
logitech G9X HyperX 900x400 / Steelseries QCK 
  hide details  
Reply
post #13830 of 22233
Quote:
Originally Posted by Kimir View Post

Sub timings are important too, I can get 80k read and copy quite easily now even with cache at 4.2Ghz, when I needed 4.4Ghz at least before some tweaking.

5820K right?

DS vs SS stick contribute too right?

I only have SS sticks so, tongue.gif
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Intel CPUs
Overclock.net › Forums › Intel › Intel CPUs › == Haswell-E Overclock Leaderboard & Owners Club ==