The benchmark results of Cinebench R23 can only be compared within this version and are not compatible or comparable with older versions of Cinebench such as Cinebench R20 or Cinebench R15. We use the single-core as well as the multi-core test here. It is based on the Cinema 4D application, which is used to generate 3D shapes. The biggest gain was increasing the Cache to 4.0 Ghz.Cinebench R23 is the latest version of Maxon's render benchmark. The Difference in my tests between 2666 Mhz and 3000 Mhz is only a 0.6 % increase in performance with Cinebench. Only if you are after higher benchmark scores. High Ram speed just isn't worth the extra money. It does support what I have been reading on the net. Is interesting though how small the gains in performance are with higher ram speeds. I know that Gigabyte has good quality hardware but their BIOS does let them down. I picked up the SOC Board and on the way home I went and bought the Rampage V Extreme. I ordered it in and by the time it came in I decided that Asus had the better BIOS. I have a brand new unopened Gigabyte X99-SOC-Force sitting on the floor here. Great stuff there and thanks for taking the time to run and post your test results, I have no idea if you could do this on other manufacturers motherboards or not, ASUS is very good with the bios's on there top end boards.įunny you should mention that. HW-E doesn't run well outside of the 100 Mhz Strap and BCLK. At least the score goes up with Ram Clock speed. Would probably need a Dram voltage of 1.5 which I don't really want to use. The review I saw said they couldn't get over 3000. Not bad for Ram that is rated for 2666 Mhz. I could probably tighten the timings more but the gains are not going to be much anyway. The Cinebench scores are averaged over 3 runs each. Not sure if 3000 at 1.380 is stable but it was only for a cinebench test anyway. I did have to bump the Dram voltage up to 1.3 though. The Cinebench scores now went up with 28. I just tried 28 at 100 Mhz Strap and BCLK. They all said above 2666 needed a Strap and BCLK of 125 Mhz. I shouldn't believe what I read on the Net. Thanks for mentioning I could change the Strap and BCLK to 100 Mhz for high Ram speed. Is that enough Proof that HW-E works much better with a Strap and BCLK at 100 Mhz ? When I tested 2800 Mhz Ram, performance was lower than 2400 Mhz Ram. Higher ram speed which forces a higher Strap and BLCK is actually slower.

It looks to me that HW-E is best with 2666 Ram and 100 Mhz Strap and BCLK. Quite a drop in Cinebench score with 125 Mhz CPU Strap and BCLK. As long as my testing procedure is accurate this proves that HW-E works much better with a 100 Mhz Strap and BCLK. Main and Sub and they were Identical in both tests. I just did some tests with the same Ram speed but different Strap and BCLK. With this Ram 2800 has a lower cinebench score compared to 26 with CPU Render Testing.īack to Cinebench score lowering with 125 Mhz Strap and BCLK. Maybe having the base clock on anything other than 100 causes some other issues with syncing the Ram and Cache together.Īnyone else noticed a similar thing when running 2800 compared to 26 ? My Cinebench score actually went down by 1.1% compared to the 2400 Ram ! It actually increased the CPU from 4.4 Ghz to 4.45 Ghz. The XMP profile upped the Ram voltage from 1.2v to 1.35v and set the CPU strap and baseclock. I then tried the Ram at 2800 using the XMP profile number 2. 2666 still used as base clock of 100 Mhz. I started off with 2666 and my cinebench score went up 0.7 % compared to the Gskill 2400.

And according to reviews it will run at 3000 by relaxing the timings a bit. This Ram even has an XMP profile for 2800, Even though it's labelled 2666. I went out and bought myself some Corsair Vengeance LPX 2666 so I could play with different speeds for myself and see if the difference is really not worth the extra money as I have read on reviews.

Good cinebench score