That's one area that I haven't gotten into too far (Modded Vbios alterations) mainly because I have enough to deal with when using the ones that are available.
Maybe one day I'll try attempting more on that.

Not talking about using that kepler editor thing. Talking about actually hex editing the changes. Hex editing the bios is not even remotely easy. lol. And to many try to copy and paste solutions that they did not figure out them selves.

Moved to #12 HOF Fire Strike.
 

Using Windows 10 Pro Insider build 10130, I WOULD have improved my Fire Strike score by 504 points, But Futuremark doesn't approve these 353.30 drivers yet!
 

Got a Minor improvement in Valid Fire Strike Score using 353.06/Windows 8.1.
Still waiting for Futuremark to approve the 353.30 drivers and also the Retail release of Windows 10.
 

Windows10 release will be a good day for benchers everywhere.  OS is just fast, plain and simple.
 
For your benching system, do you go through the OS and strip out everything not needed either by disabling functions or disabling services to have the least background overhead when you are benching, or do you just run it stock?

I have always ran my OS's used for Benching complete with All updates.
After the issues we discussed concerning the odd operation of PX & 980Ti Vbios's, I'm undecided WHAT to do next.  

Slightly higher, But at Least the Benchmark Scores don't vary much.
 

EVGATech_DaveB
Windows10 release will be a good day for benchers everywhere.  OS is just fast, plain and simple.
 
For your benching system, do you go through the OS and strip out everything not needed either by disabling functions or disabling services to have the least background overhead when you are benching, or do you just run it stock?

Windows 10 is awesome, I run the technical preview on my machine and love it.
 
When I was super into it, I did strip out some OS's, but I don't bench anymore, simply no time for it right now. When I finish college I plan to get a dual stage phase, a couple SS's, some 160L LN2 dewars, and go to town though.
 
It doesn't make a huge difference, I mainly found it helped in wprime and superpi for obvious reasons. For 3D benching stock is probably fine, it might lower CPU score by a couple hundred points but no huge impact is made.

Ahh ok.  I was curious.  It makes sense at some level, as the real performance differences of a lot of the adjustments made in subzero benching are small but they still make more points, like dropping your ET .05 seconds, that can be the difference of getting a win right there.  So I had figured OS's would be stripped as bare as possible for those extra few points.  But that can be rather tedious to go through and do.
 
Looking forward to seeing more subzero benching being done here.  You said when you finish college, is that soon?  
NEED   MORE     SUBZERO    BENCHING