Recent content by drizzler

Yes received it but have not tested it yet (vor modified the BIOS). Will try it out soon.

After some quick testing (only some CB runs) my 1660v3 (8core xeon variant of the 5960x) @ 4ghz CPU and cache with 2400 CL 10 is delivering nearly the same numbers like my 4,5 ghz 1680v2 2600 cl11 is doing even at something like cb15 (no avx/2). Clock for clock (only did some runs since my...

Just did some benchruns with my 1680v2 before swapping to x99, 1.375v under load for 4758ghz is ok but coming from 1.225v for 4.5ghz it was not worth it in the past...

Thanks for your effort but even the 2nd one is charging 20 dollars tax so i am at round about 150 dollars again (and i even have to pay a bit more due to weak euro at the moment). I was able to get a 2697v3 for under 50 euros from business so i will give that thing a try.

I should have add "hard to find for a reasonable pricing in my area". I am from europe and a 2699v3 is about 160 euros while a 2969v3 is about 200 euros (and way less offers) or i buy from aliexpress with maybe taxes to add. Same goes for the 6950x that is used from private at about 250-299...

Hey, maybe you can provide some single core bench numbers please (all cb versions, geekbench, cpu-z and whatever else you got? Would be very nice since i am exactly thinking about doing this litte upgrade. I have a rampage v laying around with a 14core e5 v4 (sadly 2 memchannels are broken due...

You should try a different strap with none matching bclk. or try the 100 strap at least with bclck below 100. X99 suffers the same bugs with higher ram dividers like X79 does. I can tell by your lower readspead in comparison to your write and copy that you suffer from this.

That is the problem, you simply can not tell before buying since it is not labeld. CPU-Z can tell you by hitting the SPD tab. Server related ram usually is lableing the rank organisation, with consumer ram you have no indication, at least back in the ddr3 days. (dunno if this changed with ddr4)...

No, rank is how the ICs are organized on the pcb. To keep it simple : single rank results in a "busy" state for the whole ramsize of the dimm while a ram access is in progress while with dual rank (or quad / octa rank) only a part of the whole dimm is "busy" while other parts can do more ram...

Also have some dusty :D 3000 mhz cl12 running with my chia rig. The problem with this high ddr3 freq rams is that they are only good for high frequency runs / benches because they are all only single ranked modules. With 2400 / 2600 / 2800 there are some dual ranked modules you can find / use...

We also talked about this here, see https://www.techpowerup.com/forums/threads/x79-and-or-rampage-iv-ocing-thread-for-those-of-us-still-left.225209/post-4554290 ;D

Dude, see my posting before. Do not use 100 bclk with 2400 divider, it is bugged and you will not get any realiable results.

Before you test further you first should get rid of the write bug (first screen) caused by 100 bclk with the chosen ram divider. Best method from my findings is to use 125 strap and lower the bclk to round about 122. You will have to test which blck is working best for your system and is giving...

Ah i think some more people running this thread are using BE. 1.225 @ bios with 1.216 cpu-z idle and 1.2 load (vrm setts : LLC extreme with current inrush enabled). 4,514ghz with ddr3 2600. Tested 4,6 ghz with 1,26 vcore, 4,7ghz with under 1.3 but have to reduce ramspeed for it quite a lot...

Temp 3,4,5 are the optional temp probe readings which are "dead" without an attached probing cable/sensor, the temp2 i do not know but it is no vrm reading (i have watercooling and the item is not moving from 92degree), so i guess it is just a dead reading from hwinfo.