G.Skill FlareX DDR4-3200 C14-14-14-34 1,35v Review

[websmile]

G.Skill FlareX DDR4-3200 C14-14-14-34 1,35v Review

 

 

Hi and welcome to our test and review of G.Skill´s new FlareX series.

 

Introduction: Motivation and a look into history

 

When you look back in history of memory, the name “Flare†might sound familiar to you. Around seven years ago, a year after AMD released its AM3 socket and processors, G.Skill decided to release a memory model series optimized for AMD systems, including overclocking profiles for these which could be loaded. This was the original Flare series, which was very successful not only because of their AMD optimization but because they used well binned PCS X ics which proved extremely useful also on all Intel sockets from 1156 up to 1150.

 

Now, after AMD released its AM4 socket with the new spectacular Ryzen processors, G.Skill decided to follow up on this again and to create a successor to the legendary Flare series, the AMD and Ryzen optimized FlareX series. Inspired by our own experience with the original Flares and because we already saw tests on AMD with these, we decided to make the cross check – what can new FlareX DDR4 do on Intel 1151 with Kabylake, and does it match our nostalgic experience we made at DDR3? Let´s make the check…

 

 

The contestant: G.Skill FlareX DDR4-3200 C14-14-14-34 2x8GB(F4-3200C14D-16GFX)

 

The memory is delivered in a classic blister packaging which contains the two modules and an inlay. Apart from the series name, “FlareXâ€, the inlay provides the information the memory kit is designed for AMD Ryzen platforms plus a specifications sticker and some promo explanation plus the information about G.Skill limited lifetime warranty.

The sticks itself come with a black matted PCB which has size (8GB) and speed printed on it, a pattern we know from G.Skill now from DDR4 start. The heatspreaders which cover pcb and ics is also designed in black colours, material used is aluminium and each side of the modules has a sticker which states the series name and that these are AMD compatible. One side also shows a sticker with exact specifications (DDR4-3200 Cl 14-14-14-34 1,35v, 2x8GB), the production date ( April 2017 in our case) on the hologram and the serial number which indicates ics used are made by Samsung (A500) and sticks were produced at week 14 of 2017. The sticks are build with 1GB ICs, this means there are eight chips used on only one side of the pcb, a clear indicator for us that at rating of DDR4-3200 C14, these are Samsung B-die chips, known for their good overclocking potential. Interesting fact, the sticks also state Intel XMP ready, means a profile on Intel platforms can be loaded. For further impressions you can check the pictures, personally we liked the design.

http://www.gskill.com/en/product/F4-3200C14D-16GFX

 

 

Tests and benchmarks

 

For the tests of the FlareX, we used the listed system

Intel Core i7 7700K air cooled

Asrock Z170M OCFormula

2x8GB G.Skill FlareX DDR3200 Cl14

MSI Geforce 7900 GTO

Mushkin AP-800 PSU

a)daily settings at 1,35v

We started with the rated specs of DDR4-3200 at C14-14-14-4-34 by loading the MXP on our board. This worked perfectly at the rated 1,35v, so we decided to go on with some popular specifications of higher binned kits which we emulated.

Moving up to DDR4-3600 16-16-16-16-36 1,35v worked perfectly, also DDR4-4000 17-19-19-39 was no problem within the 1,35v range. We then dialed in DDR4-4133 19-21-21-41 which also passed Memtest, Geekbench and Super pi 32m which we used as tests for possible daily settings. At next step we tried 4133 19-19 but failed, the fact we were able to end up at 18-20-20-41 showed s that this was due to TRCD/TRP limit at 1,35v. We show a couple of results with screenshots and it is a fair guess that this will give you a big variety of options to chose from, the fact these were confirmed with dos memtest and lots of results also were confirmed with full available 16GB at OS should guarantee you that these settings are 24/7 stable for the tester.

 

 

 

 

b) benchmark settings at high voltage

 

We started with some standard tests to evaluate needed voltage for different frequencies at Cas12-12 and 12-11.

Moving up from 3600 12-12-12-28 220 1T with tight timings we tested 3733 and 3866 up to DDR4-400012-11-11-28 1T 220.

During this process we noted that all settings worked, but that this kit we tested is for sure no low voltage kit. The 3866 12-11-11-28 setting already needed 1,85v set in BIOS, which is around 1,9-1,91v on our board due to the massive overvolting the MOCF does. We were quite surprised then to see that 400012-11-11-28 and 12-12-12-28 worked around 1.94-1,96v BIOS (2,01-2,03v real) depending on the benchmark, we tested geekbench3, Cinebench 11.5 and 15, XTU and GPUPI as well as Super Pi 32m including a test of this at Windows XP with the legendary copy waza to optimize the result and cut down time needed. All worked.

After the pile of kits we already tested in the past, optimism to go further was low, We had B-die based kits that opted out at at around 1,9v on the MOCF, average max voltage tolerated was around 1,95v which we already reached. Nonetheless we tried to achieve more.

To make a long story short, our FlareX kit opted out around 2,05v BIOS, which is way above 2,1v real voltage. We achieved around DDR4-4080 12-11-11-28 240 Super Pi 32m with waza, above DDR4-4100 12-12-12-28 geekbench around 2,03v BIOS and were able to run Super Pi 1m and cinebench 15 at DDR4-4133 Cl12-12-12-28 240 tight settings at the mentioned 2,05v set in BIOS, around 2,13v real. This was end of the road for us on the used system. This was in fact much more than we expected after our recent experiences with higher binned kits.

 

On a note, most of the tests were done on Windows 7 X64, if you ever dealt with high voltage/tight timing settings on Samsung B-die, you will know this needs maxmem set to around 4000MB. I mention this because I regularly get questions even from experienced overclockers and benchers which work with B-die for the first time, it is a ic specific problem that settings from 3400c12 upwards need this or you will lose stability at OS completely. Due to the limit on x32 operating systems, this is normally not needed there.

 

 

 

 

Conclusion and personal rating

 

It is never easy to rate memory kits nowadays. After the price explosion especially for higher rated kits over the last months, we needed to put all into a new perspective.

So we start with optics and workmanship. For us the FlareX are a good alternative to models offered before, they use an attractive design and low heatspreaders which is quite rare nowadays, the all black design fits nearly all mainboard and cooler combinations. As usual, this depends on the view of the potential buyer, but the classic design and colouring should attract a lot of potential customers. Crafting on out kit was flawless, quality of used components gave us no point we could criticize.

On the daily oc, we were able to run these on an Intel system at 4133 at specs of kits 25-35% more expensive. For benchmark use, all benchmarks including Super Pi 32m with waza above 4000 C12 and a some 2D and 3D legacy benchmarks even at 4133 c12 offer good performance. We recently tested a couple of kits rated a lot higher than 3200c14 which failed on this. Another advantage of these, at least theoretically, is their compability and profile for AMD Ryzen, how big this advantage is above other B-die kits with Intel only certification others will have to judge. We simply saw that these are good daily oc Rams and are clearly capable of running high benchmark frequency even for ambitious professional benchers, which we found not too often lately on all brands and bins.

The cheapest price at the moment at European price search engine geiizhals.de for the 3200c14 FlareX is 195 Euros, which is around 10 Euros more than similar rated kits and around 40-75 Euros cheaper than 4000 c19 and c18 rated kits in Europe for example. When I check the results obtained at the tests and our experiences with a lot of kits tested in 2017, we tend to give the FlareX a thumbs up, our sample provided solid daily and benchmark overclocking power.

The complete package of optics and performance, plus the price compared to kits with similar specifications out of the box justifies a recommendation for these, nonetheless it is fair to say that at least for Intel systems 3600 c16-16 rated like ripjaw5 series are a decent alternative. Personally I like the FlareX, they beat my expectations and if you search a decent all around benchmark kit, they are a serious contender.

 

 

Thanks for reading and to G.Skill for providing this retail sample and their faith

Edited April 12, 2017 by websmile

[Alan_Alberino]

Very nice review and very nice mems, great job... About mems, I didn't had any problems about running "non AMD Optimized B-Die" at 3200 C14 on Ryzen, so don't see the need of releasing new AMD compatible series when previous mems are 100% compatible.

[websmile]

Thanks for the informations on Ryzen, good to see "normal" mems work flawless on AM4 as well.

One major difference I saw was the choice of SPD entries which I forgot to mention explicitly. You can see at the screenshots that G.Skill added another bin as default profile, 2400 16-16 at 1,20v, labeled as JEDEC. On the Intel models, this is something I never saw before, the mems did boot 2400 by default indeed instead of the 2133 c15 I see usually.

[I.nfraR.ed]

You can see at the screenshots that G.Skill added another bin as default profile, 2400 16-16 at 1,20v, labeled as JEDEC. On the Intel models, this is something I never saw before, the mems did boot 2400 by default indeed instead of the 2133 c15 I see usually.

 

Well, that's the AMD "optimization", I guess Most important is the chip - B-die.

Both kits I have (old 3600C17 and 3733C17) work without a problem on Ryzen. I even tested for stability 4x8GB at 3200 14-14-14-34 1T (trying to mimic that G.Skill screenshot from some time ago) at 1.35V. Also managed 3800+ at C12 bench "stable".

 

But I guess it is good for people who just plug and play. It will offer a little better performance compared to 2133 JEDEC.

 

And, this FlareX doesn't seem too bad, considering many recent high bins fail on the same test. It could be much worse, plus voltage tolerance is higher than expected.

[GeorgeStorm]

Nice read

 

B-die (and DDR4 in general) is still a bit of a mystery so nice reading things like this.

[Splave]

Nice review, glad to see nice memory reviews like this again

[flanker]

good one, only as I said, you need also Ryzen for this specialized memory kit

[websmile]

I am sure Ryzen owners will test it and I heard some already did, my goal was to make the cross check like I did on DDR3 with Flares. Worked flawless and I had some fun. This kit will for sure not limit ambitions to get best possible OC on AM4 when I look at component and ic quality, and B-die seem to be best choice for all platforms involving DDR4 at the moment.

[raminr]

my friend thanks

 

these ram special for ryzen motherboard , has special new technology , you can overclock near 1000 mhz stable but only on rysen

i hope see your test on ryzrn specily gigabite ax370 gsming K7 ( can support 4000mhz ram)

 

these rysen mobo have xmp 2 ,,, gigabyte , asrock (you must enable in bios)

[flanker]

Agree

[Amd_ocer]

Well, that's the AMD "optimization", I guess Most important is the chip - B-die.

Both kits I have (old 3600C17 and 3733C17) work without a problem on Ryzen. I even tested for stability 4x8GB at 3200 14-14-14-34 1T (trying to mimic that G.Skill screenshot from some time ago) at 1.35V. Also managed 3800+ at C12 bench "stable".

 

But I guess it is good for people who just plug and play. It will offer a little better performance compared to 2133 JEDEC.

 

And, this FlareX doesn't seem too bad, considering many recent high bins fail on the same test. It could be much worse, plus voltage tolerance is higher than expected.

 

I just want to make sure, did you get the 3200 at 14 timing through 4x8gb even with the 3600 CL17 kit?? Can I achieve those numbers with a 2x2x8gb kit running at 3600C17?

 

Thanks for confirming

[Amd_ocer]

Well, that's the AMD "optimization", I guess Most important is the chip - B-die.

Both kits I have (old 3600C17 and 3733C17) work without a problem on Ryzen. I even tested for stability 4x8GB at 3200 14-14-14-34 1T (trying to mimic that G.Skill screenshot from some time ago) at 1.35V. Also managed 3800+ at C12 bench "stable".

 

But I guess it is good for people who just plug and play. It will offer a little better performance compared to 2133 JEDEC.

 

And, this FlareX doesn't seem too bad, considering many recent high bins fail on the same test. It could be much worse, plus voltage tolerance is higher than expected.

 

Was the 4x8gb 3200 C14 achieved with 4x8gb 3600C17 or a mix of the 3600 and 3733 kits? Please provide us with more details

[I.nfraR.ed]

I have 2 kits: One is older (Jan 2016) 3600C17 (17-18-18-38 XMP) and the other one is more recent 3733C17 (17-17-17-37 XMP).

The 3200C14 test was done with both kits in a 4x8GB combination. Both kits do 4000+ 12-11-11 on Apex motherboard.

[I.nfraR.ed]

So I thought it might be good to add my tests with the same model on Ryzen.

 

Test system was 1800X on air, running on Asus Crosshair VI Hero with bios 0082.

 

 

First I started with XMP. Worked without issues and I was able to lower the voltage down to 1.325V.

 

 

 

Next, I simulated a "standard" 3600C16 rated kit, which also worked flawless.

 

 

 

Last thing I tried was tightening the timings to XMP values on the same 3600MHz.

Obviously a bump in voltage was required, but I still consider this on the safe side for 24/7 operation.

 

 

 

This kit is weaker than my best B-Die, but still pretty decent.

More than 3600MHz on Ryzen are not really viable for 24/7 at the moment, even this is pushing it, because you need rather high bclk to achieve such speeds and might have problems with various devices attached to the PCI-E lanes.

 

I've also tried C12 and C11 - all worked with voltage increase.

Haven't checked them on Apex yet and don't know voltage tolerance and max Pi32M with wazza.

 

Overall it's a good memory, however if you own other good B-Die kit, there's no point changing it. You can also buy a higher bin B-die, they will still work at same timings on Ryzen.

Another thing is my IMC seems to be above average, so you might not be able to get same results at 3600, but 3200 should be no problem on 100MHz bclk with 3200 ratio or with higher bclk and lower ratio.

Edited May 4, 2017 by I.nfraR.ed

[websmile]

Good work and thanks for the results, really interesting to see how AM4 has now a decent start for AMD and DDR4

[XYchromosone]

Is the G.Skill FlareX DDR4-3200 C14 compatible with Intel Z170 MB like Asus Maximus VIII series? Or are these strictly AMD only? I can not seem to find this information so far. TIA for any help

[websmile]

Is the G.Skill FlareX DDR4-3200 C14 compatible with Intel Z170 MB like Asus Maximus VIII series? Or are these strictly AMD only? I can not seem to find this information so far. TIA for any help

 

The FlareX worked perfectly on Intel Z170, very good memory

[XYchromosone]

The FlareX worked perfectly on Intel Z170, very good memory

 

Thank you. I thought that was the case, from reading the OP second to last sentence, I quote: "Interesting fact, the sticks also state Intel XMP ready, means a profile on Intel platforms can be loaded."

 

However, others have posted information that made me question if these mems were intel compatible.