What's new
Carbonite

South Africa's Top Online Tech Classifieds!
Register a free account today to become a member! (No Under 18's)
Home of C.U.D.

Going DICE

VPII

Legendary Member
Rating - 100%
39   0   0
Joined
Dec 10, 2010
Messages
1,976
Reaction score
695
Points
5,965
It's been almost 18 months since I last ran any benchmarks using DICE or LN2. I thought it best to get rid of all the cobwebs by first trying out DICE again. The system used will be:

Ultracooled i7 4770k @ do not want to count my chickens before they've hatched
Gigabyte G1 Sniper 5 with F7 bios
2 x 8GB Corsair Vengeance DDR2400
1 x MSI GTX 780 flashed with modded bios
1 x GIGABYTE GTX 780 will also be flashed with modded bios if I end up running sli
Antec HCP 1200 Watt

Materials used to insulate the motherboard:
Art eraser (not sure of the brand)
2 ply toilet paper
Foam insulation (not sure about the technical term)

Cooling Pot used will be the Dragon F1 Gemini from Kingpin Cooling. I should have used the pot I designed as the Dragon F1 Gemini stand a chance to leak acetone as it does not seal properly. I've used tape to try and seal it but as they say the proof will be in the pudding.



The art eraser is used to insulate around the cpu socket up to the heat sinks for the motherboard, VRM and up to the first memory slot. Before I applied the art eraser I've actually pressed toilet paper tightly underneath the heat sink for the VRM to ensure that any moisture is caught by the toilet paper before getting in contact with the components.



Before mounting the pot the first memory module is slotted into the memory slot closest to the cpu and additional art eraser is applied against the memory module for added insulation. In the next picture you'll notice that I've put toilet paper over the memory modules which I've done just before I secure the cooling pot for the toilet paper to be pressed between the foam insulation around the pot and the memory module. Afterwards I've secured the pot fastening the bolts across and then diagonally across.





Luckily my case is of such a nature that I could use it as a bench table meaning no drives need to be removed and it is open for me to put dry ice in the pot. Underneath the motherboard is a whole sheet of foam insulation taped on the sides to the motherboard. The bracket for the pot is not touching the board as the sheet of foam insulation is fastened between the motherboard and fastening bracket.

Benching will start tonight and Ill be adding results soon....

3DMark11


3DMark Fire Strike
 
Last edited:
I cannot really say whether I expected more from the chip but it was fun playing around and seeing what the chip can do. I'm a little disappointed as I only managed one SLI run and that was running Vantage. I doubt whether I'll get few more runs in as my little boy will be waking up soon. But here are the results....

Superpi 1M


Superpi 32M


Cinebench11


Wprime - both 32 and 1024 ran after each other


Aquamark (I did not fine tune or even clock the GPU - just ran it to see what it does)


3Dmark03 - need more cpu speed


3Dmark05 - need more cpu speed


3Dmark06 - need more cpu speed


3Dmark Vantage - single


3DMark Vantage - SLI


3dmark11 and Fire Strike results I've added to the first post
 
Last edited:
aww yeaaa go VPII!!

Remember to monitor your voltage, 1.6 in the bios would mean 1.65 real if not more, and around 2.3 vrin should be nice.

hope results are awesome
 
Has it REALLY been that long since we had a benching session?! o.0 The Geminis don't leak TOO badly, if you've tapped it up and wrap it with toilet paper you should be fine. I really, really love that board's colours. As I said last night, very reminiscent of DFI's old INTEL boards yet better.

Anyway, I'm looking forward to the results!

Oh, and Mr AMD wants to try that other setup again. I still have my old D9GMH that did DDR2-1200 3-3-3-9 at 3.375v :p
 
Thanks guys.... I just went to pick up 15kg dry ice which I think should be enough. Will start benching tonight and see what is possible with the chip.
 
Nice! Happy Chappy Benching! Thanks for pics, keep em coming. :cool:

Hope that pot doesn't leak, a kingpin product with such defects. Sigh. Marketing i tell you
 
Hi Seeker, thanks, but it is not a defect, the pot is basically a LN2 pot and LN2 won't leak as it evaporates to quickly to leak.
 
Oh i see now. My knowledge is just just next to none existent when it comes to LN2 stuff. Looking forward to results :)

18months must have been a long long time. Good to see you back for more
 
Nice! Happy Chappy Benching! Thanks for pics, keep em coming. :cool:

Hope that pot doesn't leak, a kingpin product with such defects. Sigh. Marketing i tell you

Vince's pots are some of the best around and have incredible build quality. Don't knock things you don't understand ;)
 
You should see the looks I used to get at the shop when paying for a tub of Vaseline and two rolls of toilet paper......... :(
 
I know absolutely nothing about this but, wouldn't plumbers tape work better than toilet paper? If you trying to seal something that is.

Sent from my HTC One X using Tapatalk 4
 
Hi Crack, even though plumbers tape seal well when used on threaded pipe joints the problem is that due to the difference between the ambient temperature and temperature of the pot or inside the pot you dit with alot of condensation that runs down the pot and this is where toilet paper makes the difference catching the condensation running down.

Hope that makes sense

Sent from my GT-I9100 using Tapatalk 2
 
Done with results updated on first two posts

Sent from my GT-I9100 using Tapatalk 2
 
Nice! though my eyes hurt when trying to see cpuz , evens via the enlarged image site itself :(

Do i need glasses? What CPU Freq and volts were common across the bench runs?
 
How did I miss this?
Damn. Looking alright man. Good job :cool:
some fine tuning you can do to those scores to boost the numbers a bit. Inspector is your friend when it comes to this as you know.
Haswell chips are touch and go, get that chip to Vivi and have him pull the IHS off and replace it with something useful and your clocks should improve on the CPU.
 
[MENTION=3888]VPII[/MENTION] Well done mate ! , I haven't benched in a long while as-well .. and it seems like your catching up to me on the BOT . Must submit something again soon ... arg .. if only cash wasn't such a needed aspect in benching .. Oh well .. Nice runs .. keep up the good work .
 
Hi ShockG, thanks.... the 4770K was already went under Vivi's hands before I took it sub-zero. I'll need to get back into actually tweaking for runs before going LN2, which would hopefully be soon.

Tweak-venetica, well it does not take all that much to get back into it. If it wasn't for my understanding better half it may not have happened for me so I'm fortunate in that regard, now just for some LN2 fun.....
 
Okay.... it's been a while since I last did a good overclocking session with dry ice and on Friday I thought maybe I should try some this weekend. There has been a few hardware changes from the the last results I've posted so here with the hardware:

Intel i7 5930K
Asus Rampage V Extreme Usb 3.1
8 x 4GB DDR4 Kingston HyperX - speeds in results posted
Evga GTX980ti Hybrid
Corsair HX1200i

I ran 3 sessions, Friday night the first, Saturday night the second and this morning I did another session. At first I realized that pot was not properly fitted to the cpu, which is why I went with another run on Saturday. Some advice from [MENTION=64]ShockG[/MENTION] helped me get the cpu to 5.3ghz and stable.

Wprime 1024

VPII`s wPrime - 1024m score: 1min 32sec 231ms with a Core i7 5930K

XTU - this was the benchmark I ran to see where I could be stable as it really heats up the cores

VPII`s XTU score: 2181 marks with a Core i7 5930K

Cinebench 11.5

VPII`s Cinebench - R11.5 score: 17.42 points with a Core i7 5930K

Cinebench 15R

VPII`s Cinebench - R15 score: 1613 cb with a Core i7 5930K

On the 3D bench side, I did not run much as most of the 3d benchmarks you actually need to take the GPU sub-zero or more cpu speed to get good gains.

3DMark03 - just needed more cpu speed here

VPII`s 3DMark03 score: 189859 marks with a GeForce GTX 980 Ti

3DMark11
http://hwbot.org/submission/3237906_vpii_3dmark11___performance_geforce_gtx_980_ti_25003_marks
 
Okay.... it's been a while since I last did a good overclocking session with dry ice and on Friday I thought maybe I should try some this weekend. There has been a few hardware changes from the the last results I've posted so here with the hardware:

Intel i7 5930K
Asus Rampage V Extreme Usb 3.1
8 x 4GB DDR4 Kingston HyperX - speeds in results posted
Evga GTX980ti Hybrid
Corsair HX1200i

I ran 3 sessions, Friday night the first, Saturday night the second and this morning I did another session. At first I realized that pot was not properly fitted to the cpu, which is why I went with another run on Saturday. Some advice from [MENTION=64]ShockG[/MENTION] helped me get the cpu to 5.3ghz and stable.

Wprime 1024

VPII`s wPrime - 1024m score: 1min 32sec 231ms with a Core i7 5930K

XTU - this was the benchmark I ran to see where I could be stable as it really heats up the cores

VPII`s XTU score: 2181 marks with a Core i7 5930K

Cinebench 11.5

VPII`s Cinebench - R11.5 score: 17.42 points with a Core i7 5930K

Cinebench 15R

VPII`s Cinebench - R15 score: 1613 cb with a Core i7 5930K

On the 3D bench side, I did not run much as most of the 3d benchmarks you actually need to take the GPU sub-zero or more cpu speed to get good gains.

3DMark03 - just needed more cpu speed here

VPII`s 3DMark03 score: 189859 marks with a GeForce GTX 980 Ti

3DMark11

http://hwbot.org/submission/3237906_vpii_3dmark11___performance_geforce_gtx_980_ti_25003_marks

Firestrike

http://hwbot.org/submission/3238513_vpii_3dmark___fire_strike_geforce_gtx_980_ti_19684_marks

Firestrike Ultra

http://hwbot.org/submission/3238514_vpii_3dmark___fire_strike_ultra_geforce_gtx_980_ti_5382_marks
 
Last edited:
That CPU I suspect is really good, perhaps 5.7 or 5.8GHz good under LN2.
Sucks that I'm this far but would have definitely joined for LN2 session.
16GB as discussed will work way better. Focus on CL rather than the other two. They are necessary, but you won't be able to bring down twcl (which matters a lot) with all timings low.
You'll find you're better off with 12-18-18-36 than 14-14-14-28 many times, I see this with XTU as well. XTU is plenty of mem tweaking more so than CPU.
So far looking good though. CPU needs more cold :) or more bravery on your side. Maybe 5G NB and try higher CPU.
Memory on your board should be good for 3200 tight.
 
That CPU I suspect is really good, perhaps 5.7 or 5.8GHz good under LN2.
Sucks that I'm this far but would have definitely joined for LN2 session.
16GB as discussed will work way better. Focus on CL rather than the other two. They are necessary, but you won't be able to bring down twcl (which matters a lot) with all timings low.
You'll find you're better off with 12-18-18-36 than 14-14-14-28 many times, I see this with XTU as well. XTU is plenty of mem tweaking more so than CPU.
So far looking good though. CPU needs more cold :) or more bravery on your side. Maybe 5G NB and try higher CPU.
Memory on your board should be good for 3200 tight.
Yup.... I really need to see where I can get a Dewar to fetch Ln2 from usual spot. Next session will be 16gb ram instead with higher speed. But I am happy with this chip it seems good for way more under cold.

Sent from my SM-G925F using Tapatalk
 
So [MENTION=29046]Schnubab[/MENTION] and I did some Dice running on Friday evening. Unfortunately we struggled at first to get the system going, partly due to it reverting back to the 2nd bios, which was 1701 instead of 3101, so it was old. We ultimately for everything running, but for some reason the results just did not reflect what was expected with a 500 to 550mhz bump in cpu clock. The one run we did which was not so bad was Unigine Heaven Extreme Present:

VPII`s Unigine Heaven - Xtreme Preset score: 6404.98 DX11 Marks with a GeForce GTX 1080

The 3dmark11 and 3dmark Fire Strike runs we did that night was a little lower than expected but not all that bad. I did end up cleaning the board the following morning and refitting the pot and temp probe to see if I could get better results. Unfortunately FS Ultra seems to be a little temperamental when running the 1080, or that may be just my experience and maybe a dirty windows. But in the end I could not surpass my best result 5925 which I got with the cpu running 4750mhz.

Below is the other results I we got:

VPII`s 3DMark03 score: 202832 marks with a GeForce GTX 1080

VPII`s 3DMark Vantage - Performance score: 75388 marks with a GeForce GTX 1080

VPII`s 3DMark11 - Performance score: 27670 marks with a GeForce GTX 1080

VPII`s 3DMark - Fire Strike score: 21479 marks with a GeForce GTX 1080

I was expecting a little higher 3dmark11 and 3dmark FS run, but it is fine for now.
 
Hmm, something is wrong with your 3DMark11. I can't confirm FS, but even that should be a little higher.
On that 3DMark11 run, please if you can run AIDA64 and check your mem read, write, copy and latency results.
You should be over 28K.
 
Hmm, something is wrong with your 3DMark11. I can't confirm FS, but even that should be a little higher.
On that 3DMark11 run, please if you can run AIDA64 and check your mem read, write, copy and latency results.
You should be over 28K.
That is what I thought...

Sent from my SM-G925F using Tapatalk
 
It was a good session and thank you the chance to learn some from you [MENTION=3888]VPII[/MENTION]

Hopefully you can solve whatever is holding back your scores as you definitely deserve higher.
 
I know very little about these extreme clocking enthusiasts and what they do.

But Ihave picked up that condensation is a serious concern. Why not just have the motherboard upside down?

And moisture will then drip away from it.

Am I missing something?

Sent from my SM-N910F using Tapatalk
 
I know very little about these extreme clocking enthusiasts and what they do.

But Ihave picked up that condensation is a serious concern. Why not just have the motherboard upside down?

And moisture will then drip away from it.

Am I missing something?

Sent from my SM-N910F using Tapatalk

How will you keep the LN2 or DICE in the pot? :D:D:D:D:D

Also, condensation sticks to things. It won't just fall off.
 

Users who are viewing this thread

Latest posts

Back
Top Bottom