Sign in with
Sign up | Sign in
Your question
Solved

Which memory speed timings is faster for my memory?

Last response: in Memory
Share
March 28, 2014 1:20:13 PM

Dominator Platinum DDR3 PC3-17066 (2133Mhz)

Which is better in general (& gaming)?

9-9-9-24 @ 1333Mhz
or
9-11-11-31 @ 2133Mhz?

If you have any idea what I can get this memory to run at timings-wise at 1600Mhz or 1800 Mhz please let me know! It is hard to find information on this. Thank you!
a b } Memory
March 28, 2014 1:32:40 PM

You would have to test it with something like memtest86 to find out which has better memory bandwidth. The CAS timings can make a difference but the Mhz difference is hard to make up.
m
0
l
a b } Memory
March 28, 2014 1:49:55 PM

rswain823 said:
Dominator Platinum DDR3 PC3-17066 (2133Mhz)

Which is better in general (& gaming)?

9-9-9-24 @ 1333Mhz
or
9-11-11-31 @ 2133Mhz?

If you have any idea what I can get this memory to run at timings-wise at 1600Mhz or 1800 Mhz please let me know! It is hard to find information on this. Thank you!


It's platform specific but even so, I can't imagine any scenario where 1333 memory with those timings will outperform 2133 memory with those timings. Tbh, with such a drastic difference in frequency I can't imagine any timings (within reason) that would make up the difference in speed.
m
0
l
Related resources
March 29, 2014 9:05:56 AM

Thank you!
m
0
l
a b } Memory
March 30, 2014 6:21:30 PM

Did you even bother doing bandwidth testing?
m
0
l
a b } Memory
March 30, 2014 6:36:48 PM

Did you even bother reading the question?
m
0
l
a b } Memory
March 30, 2014 6:57:33 PM

Quote:
It's platform specific but even so, I can't imagine any scenario where 1333 memory with those timings will outperform 2133 memory with those timings. Tbh, with such a drastic difference in frequency I can't imagine any timings (within reason) that would make up the difference in speed.


this said by Deuce65 is wrong.
a 1333Mhz ram can perform faster than a 2400MHz frequency ram. This is against what you see on paper. a higher frequency ram should deliver much higher performance theoretically, but another factor comes into play. LATENCY.

latency refers to the time taken by your ram to respond to a request from the processor. For rams, as the frequency went higher the ram compromises on its latency, which means a higher frequency ram will take a long time to respond to the processor's request compared to a lower frequency one.

Thus for gaming a lower frequency ram us WAY BETTER than a higher frequency ram.

MY CHOICE FOR GAMING IS ANY RAM FROM A REPUTED COMPANY RUNNING AT 1866MHz ( because 1866 Mhz ram with CAS 9 performs the best and deliver more FPS) HAVING 16GB MEMORY.

for gaming 16GB will be enough, if you have a much deeper kitty then you can go upto 32GB. anything above 32GB will be a waste as it will be sitting there DOING NOTHING.

The entire case reversed when it comes to using multi threaded application. there a higher frequency ram is needed as to deliver large chunks of data to the applications.
m
0
l

Best solution

a b } Memory
March 30, 2014 8:04:06 PM

What you wrote is patently false. Here is a link to ACTUAL BENCHMARKS comparing low frequency, tight timings to high frequency, loose timings.

http://www.pugetsystems.com/labs/articles/Breaking-the-...

The high frequency memory out performs the low frequency memory in virtually every single test. Once could certainly argue that the minimal gains are not worth the increase in cost, but that wasn't the question.

And of course, in any case, the statement you made "Thus for gaming a lower frequency ram us WAY BETTER than a higher frequency ram." is completely wrong. Not only is it not "way better", it is actually worse.

Can 1333 memory outperform 2400 memory? Sure, given tight enough timings for the 1333 and loose enough timings for the 2400. Which is why I didn't categorically say 2400 is better. What I did say, is that in almost any realistically conceivable scenario that doesn't involve some ridiculously tight 1333 memory at like 700 dollars a stick, or someone using stupidly loose timings on the 2400, the 2400 will outperform the 1333 in every scenario. How much it outperforms depends on your platform and what you are doing.

tl;dr: gaming god doesn't know wtf he is talking about.
Share
a c 616 4 Gaming
a c 2051 } Memory
March 30, 2014 8:22:08 PM

DRAM performance is based on a combination of high freq and low latency, neither by itself decides things. A 1333/7 set equates roughly to 2400/11 DRAM though still would lose overall based on the much higher bandwidth of the 2400 sticks. If looking at performance DRAM take a stairstep method to compare (starting at 1600 which is basically entry level these days)

1600/7 1866/8 2133/9 2400/10 2666/11 with each step up there should be a slight increase performance wise - can then look at variations, i.e if looking at 2133/11 sticks with them being the same price as 1866/8 will prob see better performance from the 1866 sticks (also of note, most good sticks can OC at least 1 step higher with minimal timing changes - i.e. that same 1866/8 set can prob run at 2133/9 which is much better than 2133/11

m
0
l
a b } Memory
March 30, 2014 9:23:56 PM

Deuce65 said:
What you wrote is patently false. Here is a link to ACTUAL BENCHMARKS comparing low frequency, tight timings to high frequency, loose timings.

http://www.pugetsystems.com/labs/articles/Breaking-the-...

The high frequency memory out performs the low frequency memory in virtually every single test. Once could certainly argue that the minimal gains are not worth the increase in cost, but that wasn't the question.

And of course, in any case, the statement you made "Thus for gaming a lower frequency ram us WAY BETTER than a higher frequency ram." is completely wrong. Not only is it not "way better", it is actually worse.

Can 1333 memory outperform 2400 memory? Sure, given tight enough timings for the 1333 and loose enough timings for the 2400. Which is why I didn't categorically say 2400 is better. What I did say, is that in almost any realistically conceivable scenario that doesn't involve some ridiculously tight 1333 memory at like 700 dollars a stick, or someone using stupidly loose timings on the 2400, the 2400 will outperform the 1333 in every scenario. How much it outperforms depends on your platform and what you are doing.

tl;dr: gaming god doesn't know wtf he is talking about.


can a 1333 overtake a 2400
theoretically NO. but in real life YES

there is some thing called as latency - the time taken by your ram to respond to a request from the processor.
AS THE FREQUENCY GOES HIGHER THE LATENCY DECREASES. this is a theoretical proof. a higher frequency ram taken a lot of time to feed the processor with data, during that time the processor is literally DEAD.

a lower frequency ram can supply data quickly (though not as fast as a higher frequency ram) and can collect the next data quickly.

this is what is needed for gaming. (you can ask anyone about this)

on paper the 1333 can never perform better than a 2133 or 2400, but in real life the case CHANGES.
the 1333 can perform AS FAST AS A 2400 AND SOMETIMES (for some games) PERFORMS BETTER.

the best FPS for gaming is provided by 1866. No one can change this.


no offence
m
0
l
a c 616 4 Gaming
a c 2051 } Memory
March 30, 2014 9:40:52 PM

It's a given that gaming doesn't really require or utilize bandwidth that well, DRAM is is simply a data conduit in gaming, as far the opinion in the video, have seen that before , it floated around quite a bit with different IT folks I network with (for humorous reasons, how much faith do you giv a self proclaimed DRAM expert that stands there rubbing his fingers on the gold contacts of the sticks :)  ), though he makes the counter case that most IT folks know, higher bandwidth shines when muti-tasking (he mentions servers, which is where their strength lies) but even in a stand alone PC it's the same, it shines when muti-tasking and using memory intensive apps or large data sets
m
0
l
a b } Memory
March 30, 2014 10:54:52 PM

Gaming God said:
Deuce65 said:
What you wrote is patently false. Here is a link to ACTUAL BENCHMARKS comparing low frequency, tight timings to high frequency, loose timings.

http://www.pugetsystems.com/labs/articles/Breaking-the-...

The high frequency memory out performs the low frequency memory in virtually every single test. Once could certainly argue that the minimal gains are not worth the increase in cost, but that wasn't the question.

And of course, in any case, the statement you made "Thus for gaming a lower frequency ram us WAY BETTER than a higher frequency ram." is completely wrong. Not only is it not "way better", it is actually worse.

Can 1333 memory outperform 2400 memory? Sure, given tight enough timings for the 1333 and loose enough timings for the 2400. Which is why I didn't categorically say 2400 is better. What I did say, is that in almost any realistically conceivable scenario that doesn't involve some ridiculously tight 1333 memory at like 700 dollars a stick, or someone using stupidly loose timings on the 2400, the 2400 will outperform the 1333 in every scenario. How much it outperforms depends on your platform and what you are doing.

tl;dr: gaming god doesn't know wtf he is talking about.


can a 1333 overtake a 2400
theoretically NO. but in real life YES

there is some thing called as latency - the time taken by your ram to respond to a request from the processor.
AS THE FREQUENCY GOES HIGHER THE LATENCY DECREASES. this is a theoretical proof. a higher frequency ram taken a lot of time to feed the processor with data, during that time the processor is literally DEAD.

a lower frequency ram can supply data quickly (though not as fast as a higher frequency ram) and can collect the next data quickly.

this is what is needed for gaming. (you can ask anyone about this)

on paper the 1333 can never perform better than a 2133 or 2400, but in real life the case CHANGES.
the 1333 can perform AS FAST AS A 2400 AND SOMETIMES (for some games) PERFORMS BETTER.

the best FPS for gaming is provided by 1866. No one can change this.


no offence


I don't need to ask anyone about this. I can look at benchmark after benchmark that prove definitely that what you are saying is wrong. I linked my proof, so that anyone who is undecided can look and see the ACTUAL REAL performance they can expect from different timings. Those are not theoretical numbers, those are real life performance tests. You can continue posting your drivel, but it doesn't change the facts.

Incidentally, the original poster specifically asked which is better IN GENERAL.

Anyways, I'll leave this conversation now, as two things are obvious.
1. You're wrong.
2. Facts demonstrating you being wrong don't appear to have any influence on you, you've made up your mind and that's that. Which is fine, everyone is entitled to their own opinion, even if their opinion is obviously wrong.
m
0
l
a c 616 4 Gaming
a c 2051 } Memory
March 30, 2014 11:00:16 PM

Deuce65 said:
Gaming God said:
Deuce65 said:
What you wrote is patently false. Here is a link to ACTUAL BENCHMARKS comparing low frequency, tight timings to high frequency, loose timings.

http://www.pugetsystems.com/labs/articles/Breaking-the-...

The high frequency memory out performs the low frequency memory in virtually every single test. Once could certainly argue that the minimal gains are not worth the increase in cost, but that wasn't the question.

And of course, in any case, the statement you made "Thus for gaming a lower frequency ram us WAY BETTER than a higher frequency ram." is completely wrong. Not only is it not "way better", it is actually worse.

Can 1333 memory outperform 2400 memory? Sure, given tight enough timings for the 1333 and loose enough timings for the 2400. Which is why I didn't categorically say 2400 is better. What I did say, is that in almost any realistically conceivable scenario that doesn't involve some ridiculously tight 1333 memory at like 700 dollars a stick, or someone using stupidly loose timings on the 2400, the 2400 will outperform the 1333 in every scenario. How much it outperforms depends on your platform and what you are doing.

tl;dr: gaming god doesn't know wtf he is talking about.


can a 1333 overtake a 2400
theoretically NO. but in real life YES

there is some thing called as latency - the time taken by your ram to respond to a request from the processor.
AS THE FREQUENCY GOES HIGHER THE LATENCY DECREASES. this is a theoretical proof. a higher frequency ram taken a lot of time to feed the processor with data, during that time the processor is literally DEAD.

a lower frequency ram can supply data quickly (though not as fast as a higher frequency ram) and can collect the next data quickly.

this is what is needed for gaming. (you can ask anyone about this)

on paper the 1333 can never perform better than a 2133 or 2400, but in real life the case CHANGES.
the 1333 can perform AS FAST AS A 2400 AND SOMETIMES (for some games) PERFORMS BETTER.

the best FPS for gaming is provided by 1866. No one can change this.


no offence


I don't need to ask anyone about this. I can look at benchmark after benchmark that prove definitely that what you are saying is wrong. I linked my proof, so that anyone who is undecided can look and see the ACTUAL REAL performance they can expect from different timings. Those are not theoretical numbers, those are real life performance tests. You can continue posting your drivel, but it doesn't change the facts.

Incidentally, the original poster specifically asked which is better IN GENERAL.

Anyways, I'll leave this conversation now, as two things are obvious.
1. You're wrong.
2. Facts demonstrating you being wrong don't appear to have any influence on you, you've made up your mind and that's that. Which is fine, everyone is entitled to their own opinion, even if their opinion is obviously wrong.

_____________________

+1
m
0
l
a b } Memory
March 30, 2014 11:50:57 PM

Deuce65 said:
Gaming God said:
Deuce65 said:
What you wrote is patently false. Here is a link to ACTUAL BENCHMARKS comparing low frequency, tight timings to high frequency, loose timings.

http://www.pugetsystems.com/labs/articles/Breaking-the-...

The high frequency memory out performs the low frequency memory in virtually every single test. Once could certainly argue that the minimal gains are not worth the increase in cost, but that wasn't the question.

And of course, in any case, the statement you made "Thus for gaming a lower frequency ram us WAY BETTER than a higher frequency ram." is completely wrong. Not only is it not "way better", it is actually worse.

Can 1333 memory outperform 2400 memory? Sure, given tight enough timings for the 1333 and loose enough timings for the 2400. Which is why I didn't categorically say 2400 is better. What I did say, is that in almost any realistically conceivable scenario that doesn't involve some ridiculously tight 1333 memory at like 700 dollars a stick, or someone using stupidly loose timings on the 2400, the 2400 will outperform the 1333 in every scenario. How much it outperforms depends on your platform and what you are doing.

tl;dr: gaming god doesn't know wtf he is talking about.


can a 1333 overtake a 2400
theoretically NO. but in real life YES

there is some thing called as latency - the time taken by your ram to respond to a request from the processor.
AS THE FREQUENCY GOES HIGHER THE LATENCY DECREASES. this is a theoretical proof. a higher frequency ram taken a lot of time to feed the processor with data, during that time the processor is literally DEAD.

a lower frequency ram can supply data quickly (though not as fast as a higher frequency ram) and can collect the next data quickly.

this is what is needed for gaming. (you can ask anyone about this)

on paper the 1333 can never perform better than a 2133 or 2400, but in real life the case CHANGES.
the 1333 can perform AS FAST AS A 2400 AND SOMETIMES (for some games) PERFORMS BETTER.

the best FPS for gaming is provided by 1866. No one can change this.


no offence


I don't need to ask anyone about this. I can look at benchmark after benchmark that prove definitely that what you are saying is wrong. I linked my proof, so that anyone who is undecided can look and see the ACTUAL REAL performance they can expect from different timings. Those are not theoretical numbers, those are real life performance tests. You can continue posting your drivel, but it doesn't change the facts.

Incidentally, the original poster specifically asked which is better IN GENERAL.

Anyways, I'll leave this conversation now, as two things are obvious.
1. You're wrong.
2. Facts demonstrating you being wrong don't appear to have any influence on you, you've made up your mind and that's that. Which is fine, everyone is entitled to their own opinion, even if their opinion is obviously wrong.

http://www.youtube.com/watch?v=dWgzA2C61z4#t=259
m
0
l
a c 616 4 Gaming
a c 2051 } Memory
March 31, 2014 12:11:19 AM

Moving your youtube "explanation" doesn't change anything, as I mentioned earlier.


"It's a given that gaming doesn't really require or utilize bandwidth that well, DRAM is is simply a data conduit in gaming, as far the opinion in the video, have seen that before , it floated around quite a bit with different IT folks I network with (for humorous reasons, how much faith do you giv a self proclaimed DRAM expert that stands there rubbing his fingers on the gold contacts of the sticks :)  ), though he makes the counter case that most IT folks know, higher bandwidth shines when muti-tasking (he mentions servers, which is where their strength lies) but even in a stand alone PC it's the same, it shines when muti-tasking and using memory intensive apps or large data sets"
m
0
l
!