Six Low-Voltage Dual-Channel 8 GB Memory Kits, Overclocked

[hannibal]

These were allso the best kits that each manufacturer has in this segment. Not surprised that we did not see bad performance anywhere. The best thing is that these run fine timings in 1.35 volt, so there is no need to punish your CPU and still get desent memory bandwide.
Ivy and Haswel and CPU from this on are going to be not so tolerant to high voltages, because the node size is getting smaller and smaller.

 

[daglesj]

[citation][nom]Aegean BM[/nom]Yes, I agree that the memory tests didn't show a clear winner. I'm still glad Thomas did the tests even if they confirmed what I already thought. I admit I was curious if low voltage RAM was somehow different.However, just because I wasn't impressed with a clear winner in this batch of hopefuls doesn't mean that there aren't some losers I could buy. I don't think ALL RAM is the same.[/citation]


Yeah my point was I've seen so many RAM reviews over the past 4-5 years and they all draw the same conclusion. In not so many words...it doesn't really matter what you buy you won't get any significant performance increases from buying good stock ram to the fancy picked LED be-decked ram costing $30 extra.

Now back in the day buying 2GB of DDR500 could make a nice difference (when we were pushing 30FPS with our 6800GT's) by giving us an extra frame or two but now its not so important.

Better off spending the time, effort and money elsewhere IMO. Leave the ram at stock and tinker elsewhere.

 

[modex20]

would it be safe to assume that the 16GB kit would perform just as well as the 8GB kit?

 

[modex20]

Is it safe to assume the 16GB kits would behave the same way?

 

[modex20]

is it safe to assume the 16GB kits would behave the same way?

 

[daglesj]

Man, this posting system still needs some serious fixing.

It's 2013 people! Most sites have managed to install decent comments systems for years.

 

[Crashman]

[citation][nom]modex20[/nom]Is it safe to assume the 16GB kits would behave the same way?[/citation]You'll see one in action in the next SBM, coming soon!

 

[fullspectrum]

Maybe I'm reading it wrong, but the 3D game charts is labeled GB/s instead of FPS.

 

[Aegean BM]

[citation][nom]allan_hm[/nom]Pointless analysis for too little performance AND price difference...[/citation]Article intent was to find modules within Intel's 1.55V limit that pulled top numbers in frequency and/or latency. Some were found, article is a success.

I understand how frustrating it must get when the author writes My goal is X, and then the commenters say X was pointless. You should have done Y. Frustrating or not, that's still no excuse for declaring an article a success if the author thinks he hit his goal. The article is a success only to the extent that tom's readers say it is.

Interestingly, the article never said "top numbers in frequency and/or latency". Instead it said "Can any of today’s lower-voltage RAM deliver world-class performance without breaking past that limit?" In the end, the article failed to identify what I would call world-class performance. Loved that you did the tests, but I have to totally agree with allan_hm's Pointless analysis for too little performance AND price difference. Not a success for this reader.

 

[Crashman]

Interestingly, the article never said "top numbers in frequency and/or latency". Instead it said "Can any of today’s lower-voltage RAM deliver world-class performance without breaking past that limit?" In the end, the article failed to identify what I would call world-class performance. Loved that you did the tests, but I have to totally agree with allan_hm's Pointless analysis for too little performance AND price difference. Not a success for this reader.

You just contradicted yourself. If you wanted to see larger performance scaling and didn't, the article successfully pointed out that the big performance scaling wasn't there to be had. If you were looking for proof that DDR3-2133 could be had at within Intel's recommended voltage levels, you've found it. If you were looking for proof that DDR3-2400 couldn't be reached at 1.50V, you've found that. If you've failed to find any point to the information presented, the question becomes this: Who failed?

 

[Aegean BM]

You just contradicted yourself.

You're probably right. It's easy to talk past each other. I suspect we're using the same words for different things, and I'm quite sure our goals are different. I don't know if the following will help. FWIW I like your rebuttal.

Performance to me isn't if DDR3 frequency is 2133 or 2400. Even in FPS, I'm not going to notice 3 FPS difference or 5 seconds on a 2 minute batch job.

I love tom's tests, regardless of the results, but I hate it when tom's declares winners when the results are marginal. How can the article successfully point out that the big performance scaling wasn't there to be had and then crown some winner as if it were?

I read tom's as a buying guide. If the professional tester sat down at a machine with stock memory to work and play for a day, and then that night elves silently swapped the memory with the low voltage overclocked memory winner, would the tester notice a difference without stopwatches, spreadsheets, and the usual Seismic Hertz Inference Tool? The subjective experience answer has to be the same as the objective numbers answer to be a good buying guide.

 

[Crashman]

You just contradicted yourself.

You're probably right. It's easy to talk past each other. I suspect we're using the same words for different things, and I'm quite sure our goals are different. I don't know if the following will help. FWIW I like your rebuttal.

Performance to me isn't if DDR3 frequency is 2133 or 2400. Even in FPS, I'm not going to notice 3 FPS difference or 5 seconds on a 2 minute batch job.

I love tom's tests, regardless of the results, but I hate it when tom's declares winners when the results are marginal. How can the article successfully point out that the big performance scaling wasn't there to be had and then crown some winner as if it were?

I read tom's as a buying guide. If the professional tester sat down at a machine with stock memory to work and play for a day, and then that night elves silently swapped the memory with the low voltage overclocked memory winner, would the tester notice a difference without stopwatches, spreadsheets, and the usual Seismic Hertz Inference Tool? The subjective experience answer has to be the same as the objective numbers answer to be a good buying guide.

People often pay 10% more for a 3% performance gain, which probably makes sense when the 10% more you're paying for the product is less than 3% of the total system cost. Anyway, there needs to be some way to differentiate, otherwise nobody would bother responding to the articles.

 

[Giovanni-L]

In Brazil, computer hardware is incredibly junky. Our "highest end mainstream" parts are Kingston Value Ram chips... I know, it's sad. We feel pretty good equipped with any 1600mhz DDR3 amassing 8gig. Be it Markvision, Kingston Value or Patriot entry-level modules.

Only the wealthy ones can afford to buy anything better. To even consider having any module featured in this article is unrealistic and dream-like. Not to mention that Kingston Value 2Gig 1333mhz modules are 45US$ A PIECE!

 

[xcaninox]

All the graphics of page 12 are implying that graphics is a comparison of FPS (Frames per second) but on the bottom right of each graphics it is stated that GB/s (gigabytes per second) i believe this is a mistake cuz the text implies that it is in fact FPS, but you guys should fix this, it is very confusing especially if it is a noob who is reading.