FRONTPAGE G.Skill TridentX 8GB DDR3-2933 Memory Kit Review

[Woomack]

Overclockers are really low % of all computer users. Even lower % are extreme overclockers and memory like that is generally dedicated for this maybe 1% of all users who want something way above average even if it gives low performance gain and cost a lot.
It's not so big difference as to pay $1k for Nvidia Titan card ... similar % of the market.

 

[Ivy]

Thats true, but the question is the "why", finally its not free of cost releasing such product. Guess still worth it in term its "overpriced".

 

[Woomack]

Availability of so high clocked memory is usually limited. In countries where almost no one is buying stuff like that ( look Poland ) it's almost impossible to buy anything from the top lines.
Look at Corsair, prices are sometimes much higher and some kits are available only as an order directly in their office.
These highest series are good marketing tool. They're award winners, something to show that they're the best ... most clients see it and grab some lower kits from the same series.
You rarely see reviews of standard products but most being sold are actually these cheapest series.
Highest series are overpriced but for small % of all users it's still worth it. Limited quantity also makes price higher as dedicated group of employees have to work on this product and additionally test it. Lower sales but higher profit from each stick. On the other hand you pay more to be sure that you get best product for your needs.

Memory like that can't be recommended for every user as it's not meant to be used by regular user.

 

[Ivy]

Nah, i wouldnt even buy it in term i got 1 million dollar. Because a good timed 2400 stick owns more than that for my needs. But its always waste of bucks i feel, the CPU is limiting the effective RAM performance and Aida 64 is truly awesome but i got something more funny to do than that. The only useful program making some visible use of those RAM speed is probably 7 Zip but the difference is in the few % range, not serious enough in order to make me invest into those "cherry pick" RAMs.

 

[hokiealumnus]

+1 to Woomack's post. Ivy, you're not the intended market. This kit is solely for people that enjoy clocking memory and want to have fun with it. Anybody else needs to get a cheaper kit; this would be a bad investment for them.

 

[Ivy]

What market? They got insane clocks but barely able to perform better with almost any of the "real programs", its like buying a super fast car and then having to limit the speed by the road signs. The problem is that the processors are limiting the performance unless we use some "memory-only" synthetic benches with almost any focus on RAM but its not realistic. Oh well, i dont need to understand because im probably a person with to much practical mentality. At least Overclockers wasnt making it a "recommended" deal because it would exclude 99% of the people.

But no matter what, those RAMs are a (put word inside, i dunno what to say) because not balanced well, in almost any task, even synthetic ones, a balanced one owns the machine.

 

[EarthDog]

Let me translate for you Ivy: You guys agree with each other, essentially.

These speeds are not meant for the average joe user like you and 99% of others. But for people that benchmark and push the limits. His conclusion said that and his post above said that as well.

At the risk of opening a can of worms, what do you mean by 'not balanced well'??

 

[hokiealumnus]

What market?

This kit is solely for people that enjoy clocking memory and want to have fun with it. Anybody else needs to get a cheaper kit; this would be a bad investment for them.

 

[Ivy]

At the risk of opening a can of worms, what do you mean by 'not balanced well'??

Balanced means that the timing is good at the same time and not MHz only. For example when running games the reality is that the best RAMs are the ones with a good balance and not the ones with super timings or super clocks. Thats why the 2600 kit can beat the other 2 but i didnt actually see the timings on the graph (1 kit only, would make stuff easyer visible when all timings revealed directly). But that counts for almost any program. Of course there is nearly endless things able to test but the only stuff truly able to create visible or feelable improvements is usualy the synthetic stuff. Sure i mean when some benchers using Aida 64 and then pushing everything to the "absolute max" in order to compete... why not, but the users doing so are probably in the one thousandth of a percent range.

Of course, i do understand the "intended" use, its simply for benchers who enjoy every percent at all cost. Well i mean, i do understand when people feel proud about theyr peformance "achievements" but some stuff is simply not any practical at all, not even to benchers. Its just not my world but what matters most is simply to respect it, i try hard doing so.

For sure, im thankful for all the stuff i was able to learn while browsing those forum, i got my "practical use" already.

 

[hokiealumnus]

As asked & answered previously in the thread, as well as in the "Test System" portion of the review, the 2600 kit's four primary timings were 10-12-12-31.

 

[Bobnova]

You hit the HWBot frontpage Hokie! Here's the forum thread for said frontpage post: http://forum.hwbot.org/showthread.php?p=255969#post255969

 

[hokiealumnus]

Nice; thanks for sharing that.

The best part about it is that Massman talked to G.Skill and they actually gained respect for us thanks to this review. When you can give a company a less than glowing review and they respect you for it, that is a company that's worthy of respect.

 

[Sabishiihito]

I want to see reviews of the double-sided high clocked MFR kits. G.Skill has the SKUs but so far has anyone seen them?

 

[hokiealumnus]

The double sided sticks are all 8 GB, which won't clock like these. It would be nice to try though!

 

[Woomack]

Expect ADATA 2600 11-13-13 / DS MFR review soon.

 

[Sabishiihito]

Just in my testing, most DS MFR can do 2933 but pushing past that is tough.

 

[hokiealumnus]

Even that's pretty impressive for high density sticks.

 

[dumo]

Awesome review as always Hokie

Just want to share my test here with same kit on air/cpu on H20 @ 3500

Will try for 3600+

 

[hokiealumnus]

Thanks! Amazing clocks as always. Your memory fu owns mine.

 

[Sabishiihito]

Just an example, 2x8GB Corsair Vengeance Pro 2400C10 Hynix MFR at DDR3-3000.