Sign in with
Sign up | Sign in
Your question
Solved

Best RAM timings for Ivy Bridge 3570K

Last response: in Memory
Share
June 30, 2013 1:25:48 PM

I was looking to buy some Corsair Vengeance Blue memory and I noticed that the 8 GB (2 x 4GB) kit had timings of 9-9-9-24 and said it was for Intel Sandy Bridge. I looked up a 16 GB (2 x 8BG) kit and it had timings of 10-10-10-27. Is this one better for an Ivy Bridge processor?
a b à CPUs
a b } Memory
June 30, 2013 1:30:46 PM

9-9-9-24 is faster than 10-10-10-27
m
0
l
June 30, 2013 1:38:23 PM

daswilhelm said:
9-9-9-24 is faster than 10-10-10-27


I know it's faster, but do different architectures run better with different timings because I read that Sandy Bridge ran best with 9-9-9-24
m
0
l
Related resources

Best solution

a c 227 à CPUs
a c 150 } Memory
June 30, 2013 1:55:07 PM

9-9-9-24 is standard for DDR3 1600. The 10-10-10-27 timings are due to the sticks being 8GB. Generally the denser the stick the higher the timings will be. This is still true but lower timed higher density sticks are becoming more common and cheaper. As long as you stick with 1.5v sticks you can use pretty much anything with Sandy or Ivy Bridge. The best bang for your buck with Sandy is DDR3 1600 and with Ivy is DDR3 1866. The memory controller is on the CPU and just can not use or need anything faster.
Share
a b à CPUs
a b } Memory
June 30, 2013 2:11:45 PM

anort3 said:
9-9-9-24 is standard for DDR3 1600. The 10-10-10-27 timings are due to the sticks being 8GB. Generally the denser the stick the higher the timings will be. This is still true but lower timed higher density sticks are becoming more common and cheaper. As long as you stick with 1.5v sticks you can use pretty much anything with Sandy or Ivy Bridge. The best bang for your buck with Sandy is DDR3 1600 and with Ivy is DDR3 1866. The memory controller is on the CPU and just can not use or need anything faster.


doesn't sandy use a 1333mhz memory controller, and ivy uses 1600mhz?
m
0
l
a c 199 à CPUs
a b } Memory
June 30, 2013 2:19:13 PM

skyline4727 said:
I know it's faster, but do different architectures run better with different timings because I read that Sandy Bridge ran best with 9-9-9-24


They don't have different architectures, just different timings. Statements such as "Sandy Bridge ran best with 9-9-9-24" are generally parroted from other posts out of context. Faster RAM and RAM with lower timings is ALWAYS faster than slower RAM or RAM with bigger timings. Where the parroting leads to confusion is that depending on the application, the difference may not be significant enough to justify the price increase. It's kinda like saying a Ferrari is just as fast as a Yugo because you timed them both on a road with a 55 mph speed limit. They both might travel the same distance on that road in about the same time, but one will let ya accelerate into traffic from a dead stop a helluva lot easier. Gaming rarely provides the opportunity for RAM to flex its muscles but there are instances where it does happen.

http://www.anandtech.com/show/2792/12

22.3 % (SLI) increase in minimum frame rates w/ C6 instead of C8 in Far Cry 2
18% (single card) / 5% (SLI) increase in minimum frame rates w/ C6 instead of C8 in Dawn of War
15% (single card) / 5% (SLI) increase in minimum frame rates w/ C6 instead of C8 in World in Conflict

Also see http://www.bit-tech.net/hardware/memory/2011/01/11/the-...

Civ 5 shows a 1 fps increase at DDR3-1600 w/ CAS 7 instead of 8 which doesn't sound like much but it is a 10% increase.



m
0
l
a c 227 à CPUs
a c 150 } Memory
June 30, 2013 2:24:06 PM

daswilhelm said:
anort3 said:
9-9-9-24 is standard for DDR3 1600. The 10-10-10-27 timings are due to the sticks being 8GB. Generally the denser the stick the higher the timings will be. This is still true but lower timed higher density sticks are becoming more common and cheaper. As long as you stick with 1.5v sticks you can use pretty much anything with Sandy or Ivy Bridge. The best bang for your buck with Sandy is DDR3 1600 and with Ivy is DDR3 1866. The memory controller is on the CPU and just can not use or need anything faster.


doesn't sandy use a 1333mhz memory controller, and ivy uses 1600mhz?


Yes. But in real world testing going up a speed translates to the best overall performance.
m
0
l
a c 199 à CPUs
a b } Memory
June 30, 2013 2:24:44 PM

anort3 said:
9-9-9-24 is standard for DDR3 1600.


Someone should tell Intel. See links below. Generally, the faster the RAM and the more RAM per module, the harder it is for manufacturers to maintain low timings. CAS 9 is routinely chosen for DDR3-1600 because beyond that prices start to climb more rapidly. CAS 10 was previously the most frequently chosen for 1866 but now you are stating to see more CAS 9 than CAS 10

Quote:
As long as you stick with 1.5v sticks you can use pretty much anything with Sandy or Ivy Bridge.


This myth just won't go away

1.5 volts is for the JDEC profiles .... but most RAM speeds are advertised and labeled based upon their XMP profiles......most XMP profiles for i5 and i7 RAM is 1.65. In fact, over 2/3 of the RAM on Intel's XMP compatible list are over 1.50 volts.
http://www.intel.com/content/www/us/en/gaming/gaming-co...

Quote:
Intel® Extreme Memory Profile (Intel® XMP) allows you to overclock compatible DDR3 memory to perform beyond standard specifications. It’s designed to enhance the gaming features built into Intel® technology–based PCs. If you like to overclock and squeeze as much performance from your PC as possible, then memory based on Intel XMP gives you that extra edge you need to dominate—without breaking a sweat.

Predefined and tested Intel XMP profiles can be loaded via BIOS or a specific tuning application through a computer’s operating system. Often the easiest way to load Intel XMP profiles is using a tuning utility, which may be available depending on the particular board manufacturer. To learn whether a tuning utility is available on your system, you should contact the board manufacturer.


Again, most listed compatible i5 / i7 RAM is 1.65 .... at least according to Intel's compatibility lists

http://www.intel.com/content/www/us/en/gaming/gaming-co...
http://www.intel.com/content/www/us/en/gaming/gaming-co...










m
0
l
a b à CPUs
a b } Memory
June 30, 2013 2:26:59 PM

anort3 said:
daswilhelm said:
anort3 said:
9-9-9-24 is standard for DDR3 1600. The 10-10-10-27 timings are due to the sticks being 8GB. Generally the denser the stick the higher the timings will be. This is still true but lower timed higher density sticks are becoming more common and cheaper. As long as you stick with 1.5v sticks you can use pretty much anything with Sandy or Ivy Bridge. The best bang for your buck with Sandy is DDR3 1600 and with Ivy is DDR3 1866. The memory controller is on the CPU and just can not use or need anything faster.


doesn't sandy use a 1333mhz memory controller, and ivy uses 1600mhz?


Yes. But in real world testing going up a speed translates to the best overall performance.


all the real world testing ive seen shows little to no improvement by increasing mhz.
m
0
l
a c 227 à CPUs
a c 150 } Memory
June 30, 2013 2:27:55 PM

JackNaylorPE said:
skyline4727 said:
I know it's faster, but do different architectures run better with different timings because I read that Sandy Bridge ran best with 9-9-9-24


They don't have different architectures, just different timings. Statements such as "Sandy Bridge ran best with 9-9-9-24" are generally parroted from other posts out of context. Faster RAM and RAM with lower timings is ALWAYS faster than slower RAM or RAM with bigger timings. Where the parroting leads to confusion is that depending on the application, the difference may not be significant enough to justify the price increase. It's kinda like saying a Ferrari is just as fast as a Yugo because you timed them both on a road with a 55 mph speed limit. They both might travel the same distance on that road in about the same time, but one will let ya accelerate into traffic from a dead stop a helluva lot easier. Gaming rarely provides the opportunity for RAM to flex its muscles but there are instances where it does happen.

http://www.anandtech.com/show/2792/12

22.3 % (SLI) increase in minimum frame rates w/ C6 instead of C8 in Far Cry 2
18% (single card) / 5% (SLI) increase in minimum frame rates w/ C6 instead of C8 in Dawn of War
15% (single card) / 5% (SLI) increase in minimum frame rates w/ C6 instead of C8 in World in Conflict

Also see http://www.bit-tech.net/hardware/memory/2011/01/11/the-...

Civ 5 shows a 1 fps increase at DDR3-1600 w/ CAS 7 instead of 8 which doesn't sound like much but it is a 10% increase.





Jack your memory scaling article is useless since it's LGA 1366. Sandy and Ivy have significantly different memory controllers than first generation core I chips. They are mostly insensitive to latency.
m
0
l
a c 199 à CPUs
a b } Memory
June 30, 2013 2:33:36 PM

anort3 said:
Jack your memory scaling article is useless since it's LGA 1366. Sandy and Ivy have significantly different memory controllers than first generation core I chips. They are mostly insensitive to latency.


Gotta read the whole post, purpose of which was to show results are consistent across platforms ....bit-tech article is Sandy Bridge and Civ 5 shows 10% increase .... as I said, some games will show improvement, some won't.

Your anadtech article also says same thing

Quote:
In the end, we have to recommend what kits our users should be looking for. Taking the DDR3-1333 C9 kit as a base, it seems a no-brainer to go for the DDR3-1600 C9 kit for $5 more. The boost across the board for a negligible difference in price is worth it. The jump up to the G.Skill 1866 C9 kit also provides enough of a measurable boost, although the leap in price from 1600 C9 is another $15, which could be harder to swallow.

As we move into the 2133 C9 kit we tested today, we again across our test bed see a tangible jump in performance. This jump is not as much as moving from 1333 to 1600, but it is there and users wanting peak performance will be happy with this kit, though the size of the user pockets will also have to match.


Another thing to note is that the noted performance increase is for the whole system .... the comparison should not be x % performance increase as compared with RAM cost, it should be x % performance increase as compared with system cost as it's the entire system that benefits.

m
0
l
a c 227 à CPUs
a c 150 } Memory
June 30, 2013 2:34:34 PM

daswilhelm said:
anort3 said:
daswilhelm said:
anort3 said:
9-9-9-24 is standard for DDR3 1600. The 10-10-10-27 timings are due to the sticks being 8GB. Generally the denser the stick the higher the timings will be. This is still true but lower timed higher density sticks are becoming more common and cheaper. As long as you stick with 1.5v sticks you can use pretty much anything with Sandy or Ivy Bridge. The best bang for your buck with Sandy is DDR3 1600 and with Ivy is DDR3 1866. The memory controller is on the CPU and just can not use or need anything faster.


doesn't sandy use a 1333mhz memory controller, and ivy uses 1600mhz?


Yes. But in real world testing going up a speed translates to the best overall performance.


all the real world testing ive seen shows little to no improvement by increasing mhz.


Sandy.

http://www.bit-tech.net/hardware/memory/2011/01/11/the-...

http://www.bit-tech.net/hardware/memory/2011/01/11/the-...

http://www.bit-tech.net/hardware/memory/2011/01/11/the-...

http://www.bit-tech.net/hardware/memory/2011/01/11/the-...

Conclusion.

" our testing shows that memory rated at over 1,866MHz doesn't give much extra performance. Worse still, in some applications only 1,333MHz memory gives a performance penalty, meaning that 1,600MHz memory is fine. "

Ivy.

http://www.anandtech.com/show/6372/memory-performance-1...

Just read the whole article.



m
0
l
a c 227 à CPUs
a c 150 } Memory
June 30, 2013 2:45:00 PM

Quote:

This myth just won't go away

1.5 volts is for the JDEC profiles .... but most RAM speeds are advertised and labeled based upon their XMP profiles......most XMP profiles for i5 and i7 RAM is 1.65. In fact, over 2/3 of the RAM on Intel's XMP compatible list are over 1.50 volts.
http://www.intel.com/content/www/us/en/gaming/gaming-co...


Just google '1.65v RAM voiding Intel CPU warranty' :) 

Or just PM 4Ryan6 and ask him. You use 1.65v RAM in Sandy or Ivy and Intel will not honor the CPU warranty. Official specs are 1.5v +/- 5% so you are officially safe to 1.575v.
m
0
l
a c 535 à CPUs
a c 1942 } Memory
June 30, 2013 7:14:07 PM

Jack and Anort, Thank you, you are both making good points and it's a spirited conversation, so if I could I'd like to throw out a few things that both of you might find interesting....

First things is on DDR3 voltages - when JEDEC released the DDR3 specs it was so declared theat the standard for voltages is 1.5...which is a good thing...but here's the kicker on voltages...the original specs only called for DDR3 to go up to 1600....and Intel likes to keep the MC and DRAM voltage within about 0.5 to 0.6 of each other. So as DRAM makers started making faster sticks, 1866, 2000, 2133, these freqs required an OC of the CPU and often the MC (memory controller). Early faster/high freq sticks had higher CLs (CAS Latency's) since they couldn't effectly run the higher freqs at 1.5, now that they have been working more with them they are bring the CL's lower by using higher DRAM voltage (1.6, 1.65) which is keeping in line with the vCore and MC voltage which goes up as you OC the CPU....The higher density (8GB sticks and sets of 32 and 64GB) have also been affected which is why at times (especially initially they ran higher CLs also), better ICs and more involvement in balancing the Advanced DRAM timings is allowing for the higher freq/high density sticks/sets to be tightened down....Today it has sort of balanced out, but there are still many out of place sticks and sets...

I for one would not touch 1600 sticks of CL10 or higher, or those that call for voltage of over 1.5, voltage asside you can get 1333/CL8 sticks that will outperform 1600/CL10 and cheaper.....

Performance sticks can generally be classified as 1333/7, 1600/7-8, 1866/8, 2133/9, 2400/10, 2666/11 and of course anything lower is better CL wise....For GP sticks take the above example and add 1 to the CL i.e. 1600/9 1866/9, 2133 /10, 2400, 11 and anything higher than those combos, sort of like drugs, just say No

2133/11 can be outperformed by 1866/9 and cheaper....

Another good thing with the 'performance' sticks, most any good ones (I pretty much go GSkill about 90% of the time) will go up up a step freq wise, as easily as raising the three primary base timings by one, i.e. 1866/8 @ 8-8-8-24 to 2133 at 9-9-9/24 along with adding about + 0.05 to the DRAM voltage and maybe the same to the MC voltage

Another thing, just for info, it's never a good idea to mix sticks out of different packages, XMP is programmed by the packaged set, so while base timings remain constant, advance timing can vary greatly, i.e. 2x8GB 2133/9 may need a tRFC of 208...put that with another set, chances are good the BIOS will be lost as to what tRFC for 4 sticks - buy a 4 stick package, enable XMP and Presto! you find the tRFC is 278

Jus t some food for thought ;) 
m
0
l
!