DDR3-2133...CAS Latency important?

[Rhezuss]

Yo! Starting to think about upgrading to Ivy Bridge so need CPU/Mobo/DDR3...

So, while upgrading why not take the "best" there is since the prices are considerably low.

Well, i'm leaning toward getting 8gb (2x4gb) of DDR3-2133 since my guess is that i'll have plenty of speed with this but my question is concerning CAS Latency.

8gb od 2133 with CAS 11 cost 49CAD (G.Skill Sniper ones) but I saw the G.Skill Ripjaws 2133 with CAS 9 for 62CAD...

Is it really worth it or the DDR3-2133 CAS 11 will be plenty of punch for let's say a i5-3450 or 3570k?

 

[kleinkinstein]

Don't even toil with 2133. If you want the best, get some 1.35v Samsung sticks and OC the shit out of 'em. The DIMMS are dynamite, very little if anything truly compares.

 

[OVerLoRDI]

I don't have a good answer, but I remember CAS latency being important in the DDR1 days. CAS 2 was epic and OCZ had some DIMMs that could do 1.5 (with extra voltage). It seems as bandwidth has increased CAS latency has increased and no one cares anymore.

Even when people did care the performance difference was only like 1-3% in memory heavy applications. These days, I bet that gap is smaller and only shows up in very particular situations.

 

[tweakboy]

Latency is more important cuz 2133 is plenty bandwidth.

The higher you go in mem clock the higher the timings will be.

You can set it to 1600 or 1700Mhz @ 8-8-8 maybe.... just a thought. Youll have snappier system. gl

 

[thelastjuju]

Will it be faster? Yes, technically..

..but we're talking faster, in the milliseconds, in artificial benchmarks.

higher numbers in the specs ≠ higher performance, when it comes to this.

 

[Termie]

Depends on the test: http://www.xbitlabs.com/articles/memory/display/ivy-bridge-ddr3_3.html#sect0

I would personally also recommend the Samsung sticks, which can easily do 2133 cas11 for less than the cas11 G.Skill set you found. That being said, 2133 cas9 is a big jump, and I'd be tempted to go for the $62 set you found, because I'm not sure that the Samsungs can hit 2133 cas9 without more voltage than is advisable. I'm sure those G.Skill units are very high voltage, though, which makes me think on second thought I'd skip them.

 

[kleinkinstein]

Let's just do the math, the frequency is expressed in Hertz, which means "cycles per second". So, the DDR3 2133 will perform 2133 cycles a second while the DDR3 1600 will do, well, 1600. You, of course, know this.

Now the CAS latency is given in cycles. So, a CAS8 DIMM will take 8 cycles to respond and the CAS11, 11 cycles.

Now putting it all together - the DDR3 2133 CAS11 will take 11/2133 seconds, which is equal to 0.00516 seconds, to respond while the DDR3 1600 CAS 8 will take 8/1600, which is equal to 0,005 seconds, to respond. Thus, the 1600 DIMM is faster. For your dilemma, you're contemplating 0.000516 versus 0.00422 or timing difference of 0.00094 seconds! How fast are your reflexes and how long are you willing to wait for your memory to respond? Yes, I'm being an ass. You'll never ever see, feel or sense a difference.

You'll also want to carefully check the timings, since the higher the MHz the looser they are. I've seen 2133 DIMMS with 11-14-28-30, simply junk. As well as voltage, avoid anything higher than 1.5V. The only way many manufacturers get a respectable CAS at high MHz is to crank the voltage to 1.65V. Hence is why the 1.35V Samsung DIMMS are so smashing. Also, less DIMMs generally means less load to the memory controller. Therefore, 2x8GB is better than 4x4GB.

 

[Dark Shroud]

Yeah I paid more than I care to admit right now for my 1600 low latency ram over a year ago. The 8GB (2x4) kits of this Samsung ram locked with low latency are not much cheaper.

http://pcpartpicker.com/part/samsung-memory-mv3v4g4dus

I'm waiting for Samsung to release 8GB sticks of that ram before I upgrade again.

 

[jjmIII]

I would just get some name-brand 2x8gb DDR1600 for the same price!

 

[Termie]

Yeah I paid more than I care to admit right now for my 1600 low latency ram over a year ago. The 8GB (2x4) kits of this Samsung ram locked with low latency are not much cheaper.

http://pcpartpicker.com/part/samsung-memory-mv3v4g4dus

I'm waiting for Samsung to release 8GB sticks of that ram before I upgrade again.

Wrong product - that's DDR-1333, cas9, 1.5v, 40nm.

This is the one we're talking about:

http://www.newegg.com/Product/Produc...82E16820147096

 

[BonzaiDuck]

Some models of RipJaws rated at 2133 require higher RAM voltage approaching 1.65V to run at their spec CAS 9.

Always -- increased speed yields net gains over tighter latencies, so an overclocked RAM set at the next looser latency settings will still provide better benchmark latency scores.

But you want RAM that will run below 1.58V for these skt-1155 processors. The lower, the better.

So get the Samsungs, which I understand can be volted to just under 1.50V over their stock 1.35V. If you can trim the latencies with the overclock, keep the same latencies or even loosen them, it's the best way to go. And frankly it was my understanding that the stock latencies are looser than they need be, but if they fail stress-testing, you've got as much as 0.15V room for adjustment.

 

[Blain]

Well, i'm leaning toward getting 8gb (2x4gb) of DDR3-2133 since my guess is that i'll have plenty of speed with this but my question is concerning CAS Latency.

8gb od 2133 with CAS 11 cost 49CAD (G.Skill Sniper ones) but I saw the G.Skill Ripjaws 2133 with CAS 9 for 62CAD...

Is it really worth it or the DDR3-2133 CAS 11 will be plenty of punch for let's say a i5-3450 or 3570k?

If the rated voltage is over 1.5v, pass on the memory.

If I were shopping for high quality memory, I would use the following parameters as a guide...

* DDR3 rated at 1.5v or lower
* DDR3 rated at the lowest CAS I could afford
* DDR3 rated at the highest clock speed I could afford
* Limit the scope of my purchease to G.Skill, Mushkin, Corsair XMS or Crucial (non-Ballistix)

While not wavering on the voltage point, I would balance the other issues with my budget.

Remember my goal is not pure "benchmarking" performance, but simply finding the highest quality memory I can afford. ^_^

The only reason I pay a premium for low latency, high speed, low voltage memory is...
Quality and quality alone.

1.5v is the JEDEC DDR3 voltage standard.
Stay with 1.5v or less if you can afford it..

What he said.

 

[Rhezuss]

So this would be great: http://www.newegg.ca/Product/Product...82E16820231477

1.35V
DDR3-1600
CAS Latency 9

Pretty standard specs but low voltage.

 

[kleinkinstein]

You bet! Just make certain they will fit under your CPU cooler, those tall heat-sink fins can cause issues.

 

[BonzaiDuck]

So this would be great: http://www.newegg.ca/Product/Product...82E16820231477

1.35V
DDR3-1600
CAS Latency 9

Pretty standard specs but low voltage.

I'm rather stunned by your "Egg" find here. I had been looking at 2x8GB kits from G.SKILL for the last three months or so, before I decided to put in a second kit to make 4x4GB using their "-GBRL" Ripjaws.

I could swear I never saw any G.SKILL kits volted at 1.35V, but I must have missed these 2x4GB kits. But it makes sense: some unspecified number of G.SKILL products use Samsung black parts -- the same parts used to make the touted Samsung sticks that run at 1.35V.

It would be interesting to see how much you have to push the voltage on these to keep the same timings for DDR3-1866. What would the timings have to be to reach 2133? 10-10-10 . . ??

 

[Cerb]

I don't have a good answer, but I remember CAS latency being important in the DDR1 days. CAS 2 was epic and OCZ had some DIMMs that could do 1.5 (with extra voltage). It seems as bandwidth has increased CAS latency has increased and no one cares anymore.

No, it's that the real access times are perfectly fine. We have enough cache, and it works well enough, that tinkering with RAM is a last-mile performance optimization, modern RAM is pretty good, and we don't need fast RAM to OC, anymore.

Access times matter, but that has led to access patterns used by PCs being very important as part of the memory specifications. IoW, think of it like this: part of the reason we don't have to care much today is that we did have to care in 1999, and so many good engineers have worked to make cheap RAM good enough, and cheap RAM controllers good enough, so that we don't have to seek out special low-latency RAM sticks.

Even when people did care the performance difference was only like 1-3% in memory heavy applications. These days, I bet that gap is smaller and only shows up in very particular situations.

The difference can be more than that, but you've got to cherry pick your benchmarks. For most uses, the differences are well into the neglible range. Way back then, memory controllers and CPU caches still had room for improvement, and you might actually be able to perceive timing differences.

http://www.anandtech.com/show/4503/sandy-bridge-memory-scaling-choosing-the-best-ddr3

 

[tweakboy]

Ive studied latency and DPC latency ever since last decade.

I rather have 1700Mhz @ 8-8-8

Why don't you do benchmark. Get AIDA free and go and benchmark your latency figure. Then take the RAM to 2133 11-11-11 and test the ram latency last tab in app open it and click on memory latency and click refresh green icon and wait a sec for results. I bet 1700 with tighter timings gives say 50ns ,, but the 2133 would give more then that.

In AIDA ,, AMD 64 is still latency champ @ 47ns

 

[Rhezuss]

And if I don't overclock RAM (I really don't ever overclock RAM) i'm guessing this kit would be good: http://www.newegg.ca/Product/Product...Desktop-Memory

 

[blastingcap]

So, while upgrading why not take the "best" there is since the prices are considerably low.

http://www.anandtech.com/show/4503/sandy-bridge-memory-scaling-choosing-the-best-ddr3/6

Speed and CAS latency are related, and it's not a big deal anymore to have low latency RAM as long as you are at or above 1333 MHz, preferably 1600 MHz+. But even 1333 MHz won't see much of a hit in many applications. Anything over 1600 MHz is not really necessary. I'd save the money and put it towards a bigger SSD or better case or something else.

 

[Rhezuss]

Thanks blastingcap. I don't know much about RAM but all of this make sence. I'm gonna keep that in mind.

 

[Ben90]

So, the DDR3 2133 will perform 2133 cycles a second

lol

the DDR3 2133 CAS11 will take 11/2133 seconds, which is equal to 0.00516 seconds, to respond while the DDR3 1600 CAS 8 will take 8/1600, which is equal to 0,005 seconds, to respond.

lol

How fast are your reflexes and how long are you willing to wait for your memory to respond? Yes, I'm being an ass. You'll never ever see, feel or sense a difference.

I'd be pissed as hell if I had to wait 0.005 seconds for my memory to respond.

 

[masteryoda34]

Let's just do the math, the frequency is expressed in Hertz, which means "cycles per second". So, the DDR3 2133 will perform 2133 cycles a second while the DDR3 1600 will do, well, 1600. You, of course, know this.

Now the CAS latency is given in cycles. So, a CAS8 DIMM will take 8 cycles to respond and the CAS11, 11 cycles.

Now putting it all together - the DDR3 2133 CAS11 will take 11/2133 seconds, which is equal to 0.00516 seconds, to respond while the DDR3 1600 CAS 8 will take 8/1600, which is equal to 0,005 seconds, to respond. Thus, the 1600 DIMM is faster. For your dilemma, you're contemplating 0.000516 versus 0.00422 or timing difference of 0.00094 seconds! How fast are your reflexes and how long are you willing to wait for your memory to respond? Yes, I'm being an ass. You'll never ever see, feel or sense a difference.

This information and analysis is completely incorrect.

 

[lamedude]

The numbers need doubled since actual clock speed is half the data rate but otherwise it looks correct to me.

 

[kpo6969]

http://www.newegg.com/Product/Produc...Desktop-Memory
10-11-10-28 1T @2133 1.5V XMP
2x4GB
$59

 

[jwilliams4200]

The numbers need doubled since actual clock speed is half the data rate but otherwise it looks correct to me.

Besides being off by a factor of 500,000 .... sure, looks correct.