New Posts  All Forums:

Posts by Sam OCX

The problem with 4Gbit BFR is that it's extremely temperature sensible. After I added a fan to my sticks I could drop the stable voltages by as much as 0.08V and stable tRCD by as many as 3 values (16 to 13 at 1333). I'm not sure if any makers, beside Kingston, will bother with generating new specs to accommodate these new ICs so you'll only see them on low-end stuff.
They are actually 4Gbit BFR, MFR is much larger.
Got to test the 2x8GB 1600C11 G.Skills the other day. I must say, I was impressed with the results given the
I believe the 1600C9 version of Fury to be still based on Micron D9QBJ. If this is still the case, you should be able to get them up to 1100-1200MHz using CL10 or CL11. It is absolutely normal that the lower end modules overclock the best, percentage-wise. Don't forget that all the expensive stuff is pretty much the same memory that is guaranteed to overclock up to certain speeds.
Got another set of Crucial values recently, has Micron C9BDG. Clocks much like D9RGQ/D9RGV but doesn't even POST at 1333MHz or higher. Also got 2133C14 HyperX Fury. Is Hynix-based but doesn't like high voltage (1.6+ at 1500 and 1.3+ at 1600).
Crucial values are, indeed, not a bad option. They're cheap and Micron chips on these sticks clock better than low-end Samsung that you are very likely to get elsewhere. I would, however, get a 4x4GB kit instead of 2x8GB because of the platform.
Of the two, I would go for the Kingston kit as it has both better specs and better overclockability (G.Skill is most likely Samsung-based, which is a bad thing, Kingston is most likely Hynix-based, which is a good thing). Also, if you have access to 2666C13-rated Kingston model, I would take that instead of the 3000C15-rated one. 2666C13 is a harder spec to achieve and usually guarantees better overclockability.
I can't remember any DDR3 with load indicators like one on the picture. The closest thing you get are LEDs on Avexir, Crucial Ballistix Tracer and Panram that go on and off once every couple of seconds, depending on the memory load.
From my experience, 1500c11 is around 0.6-0.8s faster compared to 1600c12. However, I'm testing at low uncore (
I had a kit recently, here is what it could do:It also depends on the local pricing, but in Germany you can get 3000C15 blue Ripjaws for just 10 Euros more (which will overclock better), so I wouldn't bother with 2400C14.
New Posts  All Forums: