Massman

Nice one Alexis!

Latest version with Broadwell-E support: http://hwbotdownloads.s3.amazonaws.com/downloads/benchmarks/XTU-Setup-6128.exe

OMG!

WHOAH! Is this the real Phase?!

Congrats @Strong Island, epic score!

I know the feels. Catastrophic failure for HWBOT happens only when @richba5tard is on travel ...

Niiiice!

Pretty efficient system for Wprime!

@FUGGER ?

Ah, this is a FSU, not extreme. Moved

Niiiice!

Giving this a shot. Not sure what you'd want exactly. Below is a list of the highest average performance in Cinebench R15 per clock/core. Formula is: 10000 x [score / (cpu mhz x cpu cores)]. Not sure if that's what you were looking for. [table=head;autonum=1;autonumtitle=Pos;width=300px]CPU Model | Average Performance Core i3 6320|556.04 Core i3 6300|547.12 Core i3 6100|546.26 Core i7 6700K|546.21 Core i3 4370|514.55 Core i3 4340|512.48 Core i7 4770K|504.70 Core i3 4150|504.63 Core i7 4790K|501.26 Core i7 4790|496.64 Core i3 4130|496.59 Core i7 4770|495.11 Core i7 5960X|493.86 Core i7 5820K|490.15 Core i7 5930K|487.79 Core i7 4960X|449.19 Core i3 3220|449.17 Core i7 3770K|445.92 Core i7 4820K|437.34 Core i3 2120|431.77 Core i7 3960X|431.22 Core i7 2600K|429.36 Core i7 4930K|427.05 Core i7 3970X|426.45 Core i7 3930K|425.83[/table] SELECT CONCAT(cpu_model.model,"|",ROUND(10000*AVG(result.score/(result.cpu_oc*result.effective_cores)),2)) FROM result JOIN cpu_model USING (cpu_id) WHERE (application_id=94 AND result.cpucooling_id<5 AND cpu_oc<6000 AND cpu_oc>3000 AND cpucooling_id!=0 AND status_id<10) GROUP BY cpu_id HAVING COUNT(DISTINCT user_id)>=10 ORDER BY AVG(result.score/(result.cpu_oc*result.effective_cores)) DESC LIMIT 25 // edit: I guess you don't really care about the per MHz performance. Formula for below is: [score / cpu cores)]. [table=head;autonum=1;autonumtitle=Pos;width=300px]CPU Model | Average Performance Core i7 6700K|264.40 Core i3 6320|257.73 Core i3 6100|254.56 Core i7 4790K|242.92 Core i3 6300|240.73 Core i7 4770K|232.38 Core i7 5960X|228.66 Core i7 5820K|224.09 Core i7 5930K|223.87 Core i7 3770K|216.00 Core i7 4820K|212.74 Core i7 3970X|212.64 Core i7 3960X|212.26 Core i7 2600K|209.16 Core i7 4960X|208.31 Core i7 3930K|207.58 Core i3 4370|206.55 Core i7 4790|203.31 Core i7 2700K|202.20 Core i7 3820|201.34 Core i7 4930K|199.79 Core i5 6600K|197.98 Core i7 4770|197.34 Core i3 4340|194.53 Core i3 4150|184.14[/table] SELECT CONCAT(cpu_model.model,"|",ROUND(AVG(result.score/(result.effective_cores)),2)) FROM result JOIN cpu_model USING (cpu_id) WHERE (application_id=94 AND result.cpucooling_id<5 AND cpu_oc<6000 AND cpu_oc>3000 AND cpucooling_id!=0 AND status_id<10) GROUP BY cpu_id HAVING COUNT(DISTINCT user_id)>=10 ORDER BY AVG(result.score/(result.effective_cores)) DESC LIMIT 25

I had a look to find out the top-25 CPUs with the highest average overclock in the Cinebench benchmarks with air or water cooling. A table for your reference, [table=head;autonum=1;autonumtitle=Pos;width=450px]CPU Model | AVG OC in CB R11.5/R15 with Air/Water Core i7 7700K|5191 MHz FX-9590|5100 MHz FX-8370|5099 MHz FX-9370|5072 MHz FX-8350|4977 MHz Core i7 2700K|4971 MHz Core i7 2600K|4938 MHz Core i5 2500K|4929 MHz Core i7 3970X|4918 MHz Core i7 3960X|4892 MHz A10-6800K|4882 MHz Core i7 3930K|4869 MHz Core i7 4790K|4854 MHz Core i7 3770K|4848 MHz Core i7 4820K|4838 MHz Core i7 6700K|4835 MHz FX-8320E|4819 MHz FX-6350|4810 MHz Pentium 4 631|4807 MHz FX-6300|4783 MHz Core i7 3820|4779 MHz Core i5 6600K|4767 MHz FX-8320|4737 MHz FX-4100|4708 MHz Core i5 4690K|4704 MHz[/table] SELECT CONCAT(cpu_model.model,"|",ROUND(AVG(result.cpu_oc))," MHz") FROM result JOIN cpu_model USING (cpu_id) WHERE (application_id=94 OR application_id=48) AND result.cpucooling_id<5 AND cpu_oc<6000 AND cpu_oc>3000 AND cpucooling_id!=0 GROUP BY cpu_id HAVING COUNT(DISTINCT user_id)>=10 ORDER BY AVG(cpu_oc) DESC LIMIT 25

Beast mode

The list is getting more complete

If I find this one in Guang Hua, I'll pick one up as well

How come not faster than V3 ?

Savage

Savage

Insane

Hey guys, Just wanted to share the picture of our booth at Computex 2016. We used a much space we could find to promote overclocking and its culture.

From what I heard, all custom cards have the same problem at the moment. The highest clock I've heard about is ~2600 MHz, which is not nearly enough to top the leaderboards. No custom card has been shown with a "fix". Really, really strange.

Yes, and sometimes it helped. But not consistently.

http://www.overclocking-tv.com/content/news/24604/hwbot-world-tour-2016-asia-stream/

I wonder if it's easier to get high frequency in Windows 7. My chip can only do 120MHz Pifast 1/100 runs or so (usually stuck at 118 MHz). @The Stilt: I'm quite confident that the super high Vcore is only to get the high BCLK stable. I can run 117x38 at 1.5v with ~ 10c 32M stable, but had it up at 1.7V for my final runs. It's really hard to tell where the X4 845 would be if unlocked. Thanks for the info! I wonder if the training is happening properly, actually. @l0ud_sil3nc3 mentioned this already, but on my end I see no actual benefit from dialing in timings from the BIOS. In fact, apart from tCL and tRCD I see no change in performance or stability adjusting any of the other timings. Usually adjusting the timings makes things worse performance wise. I guess I'm going to pack things up for now and wait for Bristol Ridge to appear. But seriously, WTF AMD ...