G.Skill RipjawsX F3-2133C9Q-16GXL 2133 MHz 4x 4 GB DDR3

[cadaveca]

T-Topology on nearly every Z78 motherboard makes filling all four of your motherboard's memory slots an interesting option. G.Skill's new G.Skill F3-2133C9Q-16GXL kit uses four sticks, comes in red, and has great heatsink compatibility, but does using single-sided sticks hurt performance?

Show full review

 

[Delta6326]

Great review as all way's!

It shows that just pure MHz++++ doesn't mean everything.

 

[Nirutbs]

cheaper is better...5555+

 

[Sabishii Hito]

These use Samsung ICs, you can always tell which G.Skill uses by the serial number: http://www.xtremesystems.org/forums/showthread.php?283666-Figuring-out-G.Skill-s-SNs

 

[DualCpuUser]

What's the big deal here? I've been running 2133-9-11-10-28-T1 for ~ 2 years now.

 

[Aquinus]

Is there really much different between this (G.Skill RipjawsX F3-2133C9Q-16GXL) and the last G.Skill 2133 set you review a couple years ago (G.Skill RipjawsX F3-2133C9-32GXH) with the exception of the memory size and slightly looser timings? I guess I don't see the purpose of these DIMMs if they don't use less voltage or have tighter timings.

 

[cadaveca]

Is there really much different between this (G.Skill RipjawsX F3-2133C9Q-16GXL) and the last G.Skill 2133 set you review a couple years ago (G.Skill RipjawsX F3-2133C9-32GXH) with the exception of the memory size and slightly looser timings? I guess I don't see the purpose of these DIMMs if they don't use less voltage or have tighter timings.

Haswell support.

Also, I asked G.Skill specifically for single-sided DIMMs to investigate 2 DIMMs vs 4 DIMMs... and 4 DIMMs proves faster than 2 with single-sided sticks like these.

But otherwise, yeah they are probably very similar, with these sticks being the single-sided version of that older kit(which I reviewed nearly 18 months ago).

 

[hero1]

Nice review

Nice review. I have always read reviews and articles here but I finally decided to join.

BTT: I think this is a good kit for those who want to brag but the sensible lot will get a 1866 and OC it. For me, I think 1866 MHz kit, which I own one, is a way to go as it doesn't break the bank and performs just as good!

 

[DualCpuUser]

My post above is 1866 OC'd to 2133 and at them timings.

Nice review. I have always read reviews and articles here but I finally decided to join.

BTT: I think this is a good kit for those who want to brag but the sensible lot will get a 1866 and OC it. For me, I think 1866 MHz kit, which I own one, is a way to go as it doesn't break the bank and performs just as good!

 

[cadaveca]

My post above is 1866 OC'd to 2133 and at them timings.

I'll never recommend any user buy memory looking to OC it to a specific speed. IF you want a specific speed, buy that speed you want, as OC is never guaranteed, and as many brands have higher-spec'd models out, memory that might make the cut into a higher speed is usually sold for that speed.


In OC, nothing is guaranteed. You got a small OC, nothing special, 1866 to 2133 MHz, really. What so special with these sticks? Well, what's so special about yours? I got 1600 MHz stick that hit 2800 MHz, makes your 1866-to-2133 MHz OC look rather...well, unimpressive. The gains from 1866 to 2133 are so small...it's rather useless.

 

[MightyMission]

Good review dave, I was quite interested by the extra 17 FPS OCcing that ram appeared to free up in in Metro: Last Light, is that an anamoly or normal to gain so much by only adding 120 MHz to the RAM frequency?

I only ask because I use the bench in that game to tweak my system and try squeeze as much as I can out of it..i have the two stick version of the ram you reviewed also but as i'm running a sandy bridge cpu I think I'm unable to run up the FSB that much to do these tests on my own system.

 

[cadaveca]

Good review dave, I was quite interested by the extra 17 FPS OCcing that ram appeared to free up in in Metro: Last Light, is that an anamoly or normal to gain so much by only adding 120 MHz to the RAM frequency?

I only ask because I use the bench in that game to tweak my system and try squeeze as much as I can out of it..i have the two stick version of the ram you reviewed also but as i'm running a sandy bridge cpu I think I'm unable to run up the FSB that much to do these tests on my own system.

It's both the ram and the added CPU speed. Metro: Last Light seems very sensitive to memory performance changes with the settings I use, but do keep in mind that because I am focusing on CPU/memory performance, I use the lowest settings possible with a resolution of 1280x800 selected.

 

[DualCpuUser]

Absolutely nothing is special; you misinterpret my comments to be what you want to think they are.. bragging of something special. It is not; my comments are more towards the memory being discussed nothing more.

I'll never recommend any user buy memory looking to OC it to a specific speed. IF you want a specific speed, buy that speed you want, as OC is never guaranteed, and as many brands have higher-spec'd models out, memory that might make the cut into a higher speed is usually sold for that speed.


In OC, nothing is guaranteed. You got a small OC, nothing special, 1866 to 2133 MHz, really. What so special with these sticks? Well, what's so special about yours? I got 1600 MHz stick that hit 2800 MHz, makes your 1866-to-2133 MHz OC look rather...well, unimpressive. The gains from 1866 to 2133 are so small...it's rather useless.

 

[MightyMission]

It's both the ram and the added CPU speed. Metro: Last Light seems very sensitive to memory performance changes with the settings I use, but do keep in mind that because I am focusing on CPU/memory performance, I use the lowest settings possible with a resolution of 1280x800 selected.

I've just re run the bench and dropped cas to 10 to be able to boot and stably run the bench at 5038mhz(105.5x48(2250-10-11-10-32 1T)) from 5000 (100x50(2133-9-11-9-30 1T)), i'll go deeper into fettling now as i'm intrigued..i did wonder if the IGPU was starved at 2133 and that's why it seemed to perform better with a neglible increase.

Either way it's given food for thought and now a real interest in memory timings.

 

[cadaveca]

Absolutely nothing is special; you misinterpret my comments to be what you want to think they are.. bragging of something special. It is not; my comments are more towards the memory being discussed nothing more.

You used your memory as a comparison.

What's the big deal here? I've been running 2133-9-11-10-28-T1 for ~ 2 years now.

There is no big deal, simply this is a new 4x 4 GB Haswell-approved single-sided kit. This kit has not reviewed anywhere else yet, either; G.Skill gave me the "exclusive". Performance is pretty good, considering (if you have 2666 MHZ memory like I do, you'd understand how good this memory actually is, just write performance is a bit lower than 2666 MHz), and for C9 stuff, it's decently priced compared to other options on Newegg.


I'd rather review stuff you guys would buy, rather than high-end stuff. I just test, put up the numbers and my thoughts, the rest is up to you to decide on whether you agree...or not, or simply don't care. Next review is a 3100 MHz Avexir kit, then a 1600 MHz LP/LV Crucial kit.

Corsair sent me an extra 2666 C10 kit, I am considering redoing the review on that kit, too, adding in some dual-sided 4 GB stick numbers into the current results I have in this review. If there are specific items you'd like to see me review, let me know, and I'll do my best to get 'em. I asked for this kit, well, a kit kinda like this, specifically based on user requests to look at 4 sticks VS two, and with single-sided sticks. When I approached G.Skill for options they had that fit what I was looking for, this is the kit that was suggested.


Honestly, I really really like this kit. Almost more than my 2666 MHz kits.