Official Haswell-E thread

[JoeRambo]

Performed more stability testing. 4.2GHz (40 x 105MHz) @ 1.25V with the memory at 2520 using 14-14-14 timings @ 1.20V. High LLC was also used. Room was at 75° F. The CPU temps will never get this high during normal use, so I think it did pretty good. I was also browsing the web during the earlier parts of the test which caused the GFlops to go down a bit. Went from 162 Gflops at stock to 228 Gflops using the same LinX AVX settings of 14 threads and 24GB of memory. A nice increase. Going to do long duration testing before I call it stable.

I'd like to point out several errors in Your testing:

1) You are using wrong number of threads - Linpack is FP type load that does not scale with HT at all. By running 14 threads you are at mercy of Windows scheduler to swap them around. The proper way to do it is to run Linx, set threads to CPU physical cores AND in task manager, set Linx affinity to physical CPUs only (basically all odd or all even selected).

2) Linx 0.6.4 is ancient, simply put you are getting less GFlops than Haswell quad on 4.5Ghz and thermal load is not full either. Getting Linx with updated Linpack libraries ( latest is LinX 0.6.5-11.2.0 ) is great idea. ( tho you still need to apply threading affinity from (1).

3) I would not recommend starting testing AVX2 on high voltage / freq, drop some and start from there. Also using tools like PerfMonitor 2 is highly suggested, that way you can see unhalted CPU cycles from each core and if they drop during main Linx calcs phase - you have throttling.


P.S. AVX2 testing is as real world as any other stress test out there, same caveats apply.

 

[AdamK47]

Thanks for pointing out several errors in my testing.

Edit: I'll use the latest version of LinX tonight. I was just using the version I already had for quick testing while being able to use the system to do other things. I've done previous testing with LinX by assining threads to physical cores, but I've had better luck finding errors by assigning the number of threads near the number of logical cores. Full stress testing will begin soon. I'll also be testing using LinX in conjunction with 3DMark / Unigine loops. If anything fails, it's usually during such a test.

 

[Maximilian]

Just about done. Need to copy a bunch of stuff from my backup HDD to the 4TB Samsung EVO array before getting into overclocking and testing.

Shiieeeettt!

That is sweet! Nerd porn right there :thumbsup:

 

[Techhog]

Ladies and gentlemen, a golden 5960X! https://www.youtube.com/watch?v=VnQHm9ZRXw0

Keep in mind that there's a slight chance that Intel is cheating and this chip uses epoxy instead of TIM like the retail chips will. (Though there's also a chance that only the 6-core chips are using TIM while the 5960X uses epoxy.)

 

[FiLeZz]

Did amazon raise the price on the 5960x http://www.amazon.com/Intel-i7-5960X...keywords=5960x $$1,155

http://camelcamelcamel.com/Intel-i7-5960X-Haswell-E-Processor-BX80648I75960X/product/B00MMLXIHM I guess they started at a higher price then everyone else.




Newegg has been $1049 from the start..

 

[AdamK47]

Well... version 0.6.5 of LinX with the latest libraries is certainty performing a bigger pounding on my 5960X. I'm seeing temps around 5 degrees higher.

 

[dmoney1980]

Well... version 0.6.5 of LinX with the latest libraries is certainty performing a bigger pounding on my 5960X. I'm seeing temps around 5 degrees higher.

AdamK47,

What are your initial thoughts on the motherboard? I'm holding out until the rampage comes back in stock, but I might grab a asus deluxe from micro center

 

[ehume]

I've been looking at those Gigabyte boards. Notmally I go with their boards, but they seem to be running 4 of the x16 slots with three x1 between them. It's OK for the 4-PCIE card users, but hard on those of use who want to use an NH-D14 or NH-D15.

 

[AdamK47]

I feel like a guinea pig. Need some other people posting on here with their experiences!

There were several things that drew me to the Gigabyte board. I like the layout of the PCI-E slots for 4-Way SLI. When using 4-Way SLI, the PCI-E speeds are 16X/8X/8X/8X. The Gaming G1 comes with the Sound Core3D with rear panel headphone amplification. A bonus is that it has upgradable op amps.

I found a way to get higher memory speeds on the X99 Gigabyte boards. Needed to use the 1.25X base multiplier. The system won't post with the 1.25X multiplier alone though. I remembered that with my ASRock Extreme11 X79 board that I had to disable all C states in order to use the 1.25X multipler. Sure enough it worked. The memory is now at 2750.

 

[FiLeZz]

I feel like a guinea pig. Need some other people posting on here with their experiences!

There were several things that drew me to the Gigabyte board. I like the layout of the PCI-E slots for 4-Way SLI. When using 4-Way SLI, the PCI-E speeds are 16X/8X/8X/8X. The Gaming G1 comes with the Sound Core3D with rear panel headphone amplification. A bonus is that it has upgradable op amps.

I found a way to get higher memory speeds on the X99 Gigabyte boards. Needed to use the 1.25X base multiplier. The system won't post with the 1.25X multiplier alone though. I remembered that with my ASRock Extreme11 X79 board that I had to disable all C states in order to use the 1.25X multipler. Sure enough it worked. The memory is now at 2750.

I will post when I get my hands on the board and RAM.. Sux that I can't find the Asus extreme v in stock anywhere. Also cant find the gskill 3000 ram in stock anywhere. So I just for the moment have a lonely 5960x

 

[Grooveriding]

I feel like a guinea pig. Need some other people posting on here with their experiences!

There were several things that drew me to the Gigabyte board. I like the layout of the PCI-E slots for 4-Way SLI. When using 4-Way SLI, the PCI-E speeds are 16X/8X/8X/8X. The Gaming G1 comes with the Sound Core3D with rear panel headphone amplification. A bonus is that it has upgradable op amps.

I found a way to get higher memory speeds on the X99 Gigabyte boards. Needed to use the 1.25X base multiplier. The system won't post with the 1.25X multiplier alone though. I remembered that with my ASRock Extreme11 X79 board that I had to disable all C states in order to use the 1.25X multipler. Sure enough it worked. The memory is now at 2750.

My CPU got here today, came from Buffalo, but my board/ram is coming from California, should be here by Thursday. I'll be sure to post how it goes using the X99 Deluxe from Asus.

 

[AdamK47]

Tested at an even 4GHz with 1.20V using LinX 0.6.5 / Linpack 11.2.0.003. The memory is at 2666 with 1.20V and 15-15-15 timings. For 4.25GHz I'll need between 1.275V and 1.30V. Don't know if I wan't to push my luck with the temps. The Tj Max on these CPUs is lower than the mainstream Haswell. When I put the system together I was shooting for around 4GHz anyway. I want the peace of mind that my system is stable. I'll probably play with uncore speeds and memory timings next.

 

[EvilNodZ]

I got my system built lastnight but had some rebooting problems.

System was solid when in windows but kept getting stuck in a reeboot cycle anytime i restarted the pc rebooting before i could get into the bios.
I think i sorted it with a bios update.
This was on the gigabyte X99 gaming 5 at stock.

I done a quick test lastnight before i went to bed. I managed to get my i7 5820K at 4.0ghz on all cores with all settings on auto in bios setting which bumped the vcore upto 1.192V using NH-D15 temps were mid 70degc after 25runs of LinX.

Ill do proper manual testing tonight tho when i have time.

 

[FiLeZz]

Please post your final overclocks along with your CPU you are using, your S-spec and your batch#. I will update the post with everyone's info.

No overclock for me as of yet, (waiting on parts) but here is my CPU info.
CPU: 5960x
S-spec SR20Q
Batch# 3424B585

 

[Bubbleawsome]

I'm just going to throw this out here. If anyone happens to already own BeamNG (or want to buy it I guess) and has got the 5960x would you mind benching this? http://www.beamng.com/threads/8595-BANANA-Benchmark-Results-Comparison

No one has broken 100mbeams. Fastest we've got is a 3930k at 5.1Ghz.

 

[EvilNodZ]

CPU: 5820K
S-spec SR20S
Batch# 3420C075

3.9Ghz
@ 1.175V Fail
@ 1.180V Stable

4.0 Ghz
@ 1.185V Fail
@ 1.190V Stable

4.1 Ghz
@ 1.195V Fail
@ 1.200V Fail
@ 1.205V Fail
@ 1.210V Fail
@ 1.215V Fail
@ 1.220V Pass

4.2Ghz
@ 1.225V Fail
@ 1.230V Fail
@ 1.235V Fail
@ 1.240V Fail
@ 1.245V Fail
@ 1.250V Fail

Im not going to push this OC any higher as temps were hitting high 80s during LinX on air.

 

[AdamK47]

What is the reported Tj Max of the 5820K? Same as the 5960X?

 

[Grooveriding]

http://www.youtube.com/watch?v=6yBUKBIhga0&list=UUgTTyN54Hljm7TqD13q8FEg

Decent video for some average expected overclocked numbers out of HW-E chips.

 

[EvilNodZ]

TJ Max is showing 94Deg C for the i7 5820K

wow didnt realise it was 89Deg C for the i7 5960X

 

[dmoney1980]

4.0 Ghz
@ 1.185V Fail
@ 1.190V Stable

4.1 Ghz
@ 1.195V Fail
@ 1.200V Fail
@ 1.205V Fail
@ 1.210V Fail
@ 1.215V Fail
@ 1.220V Pass

That's a big jump in voltage for only 100mhz!

 

[Kenmitch]

That's a big jump in voltage for only 100mhz!

.03v's doesn't seem like a big jump to me.

 

[EvilNodZ]

Yesterday my Tj Max was 94degC according to core temp, why does it say 118DegC now?

Anyone any ideas?

I have changed some stuff in the bios but surely that wont affect Tj max?

 

[AdamK47]

Testing out 4.125GHz @ 1.25V with 8 threads (affinity set) and 28GB of memory. I'll see if the system is still processing when I get home from work.

 

[biostud]

Asrock Extreme 4 + 5820K 4.4Ghz @ 1.35V

http://www.vmodtech.com/main/articl...rming-with-asrock-x99-extreme-4-review/all/1/

 

[AdamK47]

Asrock Extreme 4 + 5820K 4.4Ghz @ 1.35V


http://www.vmodtech.com/main/articl...rming-with-asrock-x99-extreme-4-review/all/1/

4.4GHz @ 1.35V may be good for quick benchmarking. I take overclock results with a grain of salt until I see proof of stability testing.