Ballistix Tactical Tracer 2666 MHz DDR4

[newtekie1]

Yes, I do actually look at my case a lot. If you don't care or don't have a window, etc, you're not going to get such RAM kits regardless. There's PLENTY of options already out there.

I'm sure you do look at it a lot, but are you seriously using it as a diagnostic tool to detect RAM usage? I doubt it.

 

[Woomack]

8x8GB @3800 CL20 1.45V on X299 ( 2 kits of 32GB )
I have only 4 slot X399 mobo but it works at 3600 20-23-23 1.35V stable, and for tests max is 3733. Can't count on much tighter timings but performance is still not bad.

 

[coonbro]

I've always been of the opinion that heatspreaders on RAM do nothing because the ICs don't get hot enough to require them... in fact, they might actually be detrimental to performance/longevity because they don't allow air to circulate over the ICs. IMO, an article proving or disproving this hypothesis would be really useful - i.e. get a kit of RAM with heatspreaders, put a temperature probe as close to one of the ICs as possible, run some benchmarks and overclocking, then remove the heatspreaders and repeat the tests.

I'd be willing to do such a test myself, but I don't have a DDR4 system.

I pulled the sinks off on some that did not fit as I liked under a cpu cooler , I did not see where it affected the memory at all . thing is I do feel its more about brands having a way to hype up the looks ,, [WOW!!! THAATS SO COOL !!} than you know next is a place to splash and hype the brand and name ..

in the end practical and function is not the first priority . like I like to say its all about hype and gimmicks over support function and service , and led lighting is now the strongest selling points of todays hardware

then maybe all that 20 way user programmable rbg LED's add to a more massive and stable overclock over the same that naked OEM looking sticks ? if not then just set up the led lighting and play some 1970's disco music and enjoy the inside of your case.

most sinks are applied with thermal tape / glue looking stuff and careful effort the sinks peel off . but ya, like you its more about show then anything to do with go

was it back in the ddr2 days Samsung had plain jane OEM sticks that kicked A$$ if you got a hold of some you had something

 

[cadaveca]

I've always been of the opinion that heatspreaders on RAM do nothing because the ICs don't get hot enough to require them... in fact, they might actually be detrimental to performance/longevity because they don't allow air to circulate over the ICs. IMO, an article proving or disproving this hypothesis would be really useful - i.e. get a kit of RAM with heatspreaders, put a temperature probe as close to one of the ICs as possible, run some benchmarks and overclocking, then remove the heatspreaders and repeat the tests.

I'd be willing to do such a test myself, but I don't have a DDR4 system.

What's really going with heatspreaders are this:

They aren't really to cool per se.

You see, memory has differing refresh rates depending on temperature. If the ICs are of differing temps, they could require different refresh rates, something not possible, so the heatspreader makes sure that all ICs are the same temperature, and ergo, require the same refresh. It's NOT about keeping the sticks cooler than they would be otherwise; it's an effort to ensure the sticks stay stable from differing temps within the stick.

8x8GB @3800 CL20 1.45V on X299 ( 2 kits of 32GB )
I have only 4 slot X399 mobo but it works at 3600 20-23-23 1.35V stable, and for tests max is 3733. Can't count on much tighter timings but performance is still not bad.

Yeah, getting that overall latency down helps for sure, but those bandwidth numbers aren't anything to call home about for 3800 MHz.

you have the 16 GB sticks, eh?

 

[coonbro]

''You see, memory has differing refresh rates depending on temperature. If the ICs are of differing temps, they could require different refresh rates, something not possible, so the heatspreader makes sure that all ICs are the same temperature, and ergo, require the same refresh. It's NOT about keeping the sticks cooler than they would be otherwise; it's an effort to ensure the sticks stay stable from differing temps within the stick.''

that could be a interesting point of view on that , then your back to what I asked abot removing the sinks in the review to then see if in fact there is say thermal tape on each ic in contact with the sink ?

so to say prove the sinks are practical and not sales hype , sure seems like the ones I removed were just for show and just thermal tape for holding them on at each end ? [its been some time ago I did that , and I do suffer from brain cramps from time to time. well maybe a lot ..lol.. ]

like here with these are the sinks practical for that or just there to hold the LED lighting in ?

 

[cadaveca]

to then see if in fact there is say thermal tape on each ic in contact with the sink ?

There is great contact with these sticks.

so to say prove the sinks are practical and not sales hype , sure seems like the ones I removed were just for show and just thermal tape for holding them on at each end ? [its been some time ago I did that , and I do suffer from brain cramps from time to time. well maybe a lot ..lol.. ]

Like, I said, high temperatures. chances are your sticks haven't breached that like 72c mark. You see, those temps aren't that high. There are within a reasonable range for memory in a hot climate at the peak of summer, so much so that part of a stick may be say, 69c, and OK with normal refresh, but the other may be 75c, and needs higher refresh now... the heatspreader spreads the heat. it is NOT a heatsink.

like here with these are the sinks practical for that or just there to hold the LED lighting in ?

LEDs are affixed to the PCB. This is shown in one of the pictures.

 

[coonbro]

got .ya.

remember when g-skills top kits came with a 2 fan bracket for there cooling , [do they still do that ? ]

so I guess with this its not a spreader if using the fans on them , your back to whats hype and whats not

https://www.newegg.com/Product/Product.aspx?Item=N82E16820231808

like here from the brand reviewed its 3ed in reasons for them and #1 is looks and style . must not be that critical even to them not being the primary thing in there reasons

https://ballistixgaming.com/articles/three-benefits-of-ballistix-heat-spreaders.html

the best

'' Since newer technology is more energy-efficient,''

guess that's why todays hardware needs all the fancy coolers dual and triple fans and cases need 6 - 120mm fans all over the place . compare to the old egg frying power sucking hungry parts of the past that did not ?

lol...

in the end it all about hype , gimmicks and sales pitch ..


i'll leave it at that ..

good review and enjoy . your points are valid to me on the spreader function I can se that as workable reason

 

[Woomack]

Yeah, getting that overall latency down helps for sure, but those bandwidth numbers aren't anything to call home about for 3800 MHz.

you have the 16 GB sticks, eh?

Bandwidth is not bad just that latency is not as low as we wish. I guess that max bandwidth in AIDA64 also depends on the motherboard as I see some differences between brands and some boards have AIDA64 boost option in BIOS what I don't have on ASRock boards.
At least bandwidth on X299/X399 doesn't look bad comparing to some other kits ( even Samsung based ) on XMP profiles. I'm not talking about tweaked subs etc. On the other hand, most users won't even touch these settings.
Also keep in mind that my results are on i9-7900X/TR 1920X while results on 16 cores will give higher bandwidth.

I have 8GB sticks, as I said in my post. I was testing 8x8GB so 2 kits on X299 and 4x8GB on X399 as I have ASRock X399M Taichi which has 4 memory slots.
Here is 3733 on ASRock X399M Taichi.

 

[cadaveca]

very interesting that you are getting these clocks, but it is with quite the voltage boost (which we know should be fine, but, ykwim)

I have 8GB sticks, as I said in my post. I was testing 8x8GB so 2 kits on X299 and 4x8GB on X399 as I have ASRock X399M Taichi which has 4 memory slots.

Ah. I was confused because you said 4 sticks and two 32 GB kit and it threw me for a loop.

What I posted earlier on the X399M taichi was from different kit, not this one. What's interesting to see is how the numbers we got are similar, although timings are different, as are CPUs. Cores don't really make that much of a difference that I have seen, neither on Intel nor AMD. I have 7900X and 7980XE I use for testing memory on X299, and the memory numbers are similar there for me too.

Internal timings for memory and such definitely do change from board to board; even BIOS to BIOS at times.

 

[Assimilator]

What's really going with heatspreaders are this:

They aren't really to cool per se.

You see, memory has differing refresh rates depending on temperature. If the ICs are of differing temps, they could require different refresh rates, something not possible, so the heatspreader makes sure that all ICs are the same temperature, and ergo, require the same refresh. It's NOT about keeping the sticks cooler than they would be otherwise; it's an effort to ensure the sticks stay stable from differing temps within the stick.

Huh, that makes sense! Thanks!

 

[coonbro]

Huh, that makes sense! Thanks!

all you can do is buy them and pull the ''spreader'' off and see if there all connected to it to transfer the heat from one to the next or what ever or if its just a cover they slapped on for there logos .

[sorry , I said before I was done but, ]
like if you got a bank row of 6 IC's 1-2-3-4-5-6- and 6 is the hottest how does the heat transfer to say 1 if its the coolest one clear on the other end of the stick ?? seems the heat would disapate faster then it could move that far now its a sink [???]

that one I pulled off just seemed a cover with a thermal tape at each end to hold it on then maybe it was to trap heat to keep then all warm like a blanket ? then if thermal tape is used to hold them on ? then is it on all ic chips with out a gap in between them / a strip or a pad on each one individually ? is the tape or what ever a material that can move the heat from the hottest IC all the way across to the coldest one

thing at ? here on that is whats fact and whats hype or gimmick unless you but them take them apart and look and think ''well that will work or that a load of crap . out side of that your just taking there word on it all . in the review with out that shown apart to see its just a gamble on there word it can or just a fancy logo placeholder .

I would think top brands and there top memory is as expected but these days , I just don't know anymore like I say gimmicks and sales hype supersede everything else in the end with all this hardware . the best you can hope for is that it works as advertised and expected or it don't and back to newegg . that's the true end result of it all spreader or not .

good luck

heres a jim dandy spreader job for example

http://www.legitreviews.com/corsair-pc2-5400ul-ddr2-memory-review_204/2

extra info

http://www.legitreviews.com/do-memory-heat-spreaders-make-a-difference_299

 

[CheapMeat]

I'm sure you do look at it a lot, but are you seriously using it as a diagnostic tool to detect RAM usage? I doubt it.

That's EXACTLY what happened with a X79 board I bought. I was getting a memory error on the diagnostic Dr. Debug. It would not go further in the post. Dr. Debug just shows you there's an error & the nature of it, not where the exact error is. Some of newer boards have DIMM slot LEDs to show such errors but X79 boards don't. The Tracers LEDs showed me instantly which DIMM slot was the culprit. It allowed me to figure out which slots worked and which didn't. And it helped me narrow down the problem to being a bent LGA pin. Part of the charm of the rig is the aesthetics but it DID serve as a diagnostic tool in this situation. Do you think data center technicians stare at the racks constantly as well? No. But do they still include blinkenlights? Yes they do.




Also at coonbro, they often use heatspredders on server DIMMs too. I'm sure those idiots are getting paid top dollar to waste money like that. Have you ever touched a DIMM after heavy usage? Is it cool to touch? Because when I have, they're quite hot. You can induce soft errors in RAM with heat. As mentioned by cadaveca, they're not always used as a heatsink, but a heat spreader. But they're not useless as alluded to.

http://bluesmoke.sourceforge.net/heat_gun.html

https://arstechnica.com/information...y-turns-assumptions-about-errors-upside-down/

 

[Tsukiyomi91]

I like the "the RGB is strong with this one" remark. xDD Kinda agree with that too. the 3D printed logo makes it perfect for demo systems or LAN party goers who has both speed & style.

 

[basco]

sorry off topic but plz tell me how i can open that window with spd timings in aida64 like woomack + lots others are showing?
TiA

 

[cadaveca]

sorry off topic but plz tell me how i can open that window with spd timings in aida64 like woomack + lots others are showing?
TiA

you need the engineer edition of AIDA64.