ROG RealBench Chat and Comments

[Nodens]

Mmmm... Closed beta, yum yum. Winter time, cold nights, rain falling on the windows, fireplace, beta testing new RB Am I romantic?

You won't need a fireplace if you're benching..case exhaust will fix you right up..

[meankeys]

You won't need a fireplace if you're benching..case exhaust will fix you right up..

Yeah Baby RB will be HOT!

Keep up the great work Nodens. Looking forward to version II. If I only knew 1/2 of what you guy's forgot about this stuff I would be a very smart and a happy camper

[meankeys]

Not sure. Was getting 725-735 range before with 4.6GHz. Now, I can't break 720 with fresh RB, the same settings, tweaks, crossed fingers and facial deformations. Disconnecting my external HD helped a little, but I'm missing something. Maybe I'm not holding my tongue right...<shrugs>.
Yep it's the 4770k. I think I might go back and play with my P8Z77 Pro-V with the 3770k for couple days and see if I can coax some kind of epiphany to come to me by pretending I don't care.

IM I am far from being the expert here but this is what I would try

1. Go back to the basic bios settings best know version.
2. set the board up to your fastest stable OC
3. Run your memory up to it's fastest stable know speed @ default timings
4 Run RB with the latest know tweaks at least 2 times post your score
Oh and # 5 swoop your dimm's A to B and B to A

The reason for step 5 is to stop you from thinking in to this to hard. RELAX you know what works do it.

[IM2L844]

Thanks, MK. I think I'm back in the saddle. I don't know how all the Windows 8.1 performance dynamics work, but the official release left a whole bunch of crap files that needed cleaned up and it created two recovery partitions on my SSD. After getting all that sorted out, my RB scores picked up. This is with the settings I had saved to HWBot - XTU that previously produced a 735 score and I think these results are within an acceptable margin of varience from run to run:



Now, I can go ahead and see what 47 & 48 will do, but I think I'm gonna just set this machine up with a nice hefty adaptive oc for everyday use now that the temperatures are under control. I think I'll set my sights on building a dedicated benching rig so I don't have to constantly fiddle with settings per separate intended uses. I'm hooked on this hobby so I might as well plan on getting set up to maximize long term enjoyment of it.

[meankeys]

Lookin much better IM. I am glad you got it all sorted out. That score looks better for sure. Yes this hobby is very addicting and it make it that much more fun when your setup with a bench and dedicated rig. good luck my friend I look forward to seeing an 800 RB score shortly

[solaufein_gr]

You people are very kind. Thank you, i had a great time. A place to learn and have a good time and make friends

[Nodens]

IM, while what you're planning is the best course of action, I wanted to let you know that you can use the profiles saving feature for your settings. It's really there for benchmarking heh. As long as you're int he same UEFI version your can save and load as you please

[Abyssic]

isn't rog supposed to be gaming oriented? why is this a workstation benchmark?

[Zka17]

Welcome, Abyssic, to the ROG Forum!

The thing is that ROG is already over the gaming... the top/Extreme boards are for overclocking...

This is only the first version of RealBench - as I understood, the next version will have more detailed testings, maybe it will include graphics too... which would be a good pre-test for gaming...

[Nodens]

Unfortunately the OpenGL test did not make it in v2. Due to time constraints and lots of changes in tests and stress test plus other functions such as online leaderboard, automated score uploading, etc. We want to tackle the OpenGL test properly and that will need a version on its own most likely

Public beta very soon now. Should be this week. (We are not doing a closed one..public directly).

PS You can not classify RB as a workstation benchmark. It's a performance benchmark based on realistic usage scenarios. The concept behind it is to see how your system performs in a task that you actually may find yourself doing instead of running a few algos and display a performance bar. That allows a clear comparison of system performance that is understandable to any user, contrary to calculating something like Gigaflops..or other arbitrary number that no normal user can translate into tangible system performance differences. Eventually it will cover the *gaming* scenario as well. But we don't want to do a 3dmark type of test for that. We have something else in mind for it.

EDIT: Btw, with v2 you won't even need to run CPU-Z tabs for your screenshots. And it also has a builtin screenshot function with superior image quality output