Are lower numbers better for timing?

esCob4r · 03-31-12, 01:43 AM

Title says it all?
I have G.Skill RipjawsX 1600Mhz DDR3 8-8-8-24 2T memory. Does this mean my memory is faster then those set to 9-9-9-24?

redduc900 · 03-31-12, 03:18 AM

Yes... the smaller the number = the fewer the clock cycles = the faster the memory.

Janus67 · 03-31-12, 06:26 AM

Yes, although if those set to 999 are running 1t instead of 2t you are effectively running the same speeds, as running 2t pretty much adds 1 to your main timings

esCob4r · 03-31-12, 11:24 AM

ok so 8-8-8-24 2t is basically really 9-9-9-24? If so then what is the exact point of having lower 8-8-8-24 2t? I understand I could probably manually adjust it to run 1t maybe? Or maybe not?

Either way thanks. Trying to learn as much as I can with building pcs, overclocking, and what not.

redduc900 · 03-31-12, 11:30 AM

No... Janus was comparing the effect of 1T and 2T Command Rates to looser (9-9-9) and tighter (8-8-8) timings.

johan851 · 03-31-12, 12:08 PM

Quote:

ok so 8-8-8-24 2t is basically really 9-9-9-24?

8-8-8-24-2T is roughly equivalent to 9-9-9-24-1T. 8-8-8-24-1T is faster than both.

esCob4r · 03-31-12, 04:16 PM

Got yah. So basically it would make no real difference to use 8-8-8-24 2t vs 9-9-9-24 1t. I might have to look into that laters.

johan851 · 03-31-12, 04:30 PM

Right. But if you're using the subtimings, you can sometimes get to something like 8-9-8 @ 1T instead of 9-9-9 @ 1T, when maybe 8-8-8 @ 1T isn't quite stable. So there's a bit more flexibility.

We're talking tiny fractions of performance difference at that level though.

freeagent · 03-31-12, 04:38 PM

Chances are you wouldnt see a difference running 6-6-6-20 1T vs 8-8-8 or 9-9-9, except in benches, or if you are a maniac who never leaves his pc

Uncore is where its at

esCob4r · 03-31-12, 05:50 PM

I see so for gaming only it makes no real difference. Got ya. Good to learn and now.

johan851 · 03-31-12, 08:45 PM

Well... 6-6-6-20 1T maybe. But good luck finding memory that'll do that at speed.

I get your point though - it really doesn't make a difference. But it's one more thing to tweak, and the tweaking part is fun.

freeagent · 04-01-12, 07:58 AM

Quote:

Originally Posted by johan851

Well... 6-6-6-20 1T maybe. But good luck finding memory that'll do that at speed.

I get your point though - it really doesn't make a difference. But it's one more thing to tweak, and the tweaking part is fun.

I know Ive looked high and low for another set that can do em, no dice

These old Hypers are the last of a dieing breed

Sentential · 04-01-12, 08:01 AM

Quote:

Originally Posted by esCob4r

Got yah. So basically it would make no real difference to use 8-8-8-24 2t vs 9-9-9-24 1t. I might have to look into that laters.

It can make a substantial difference depending on the chip itself; the easiest way to think of it is that the timings there are how long the dram chip takes to respond to data queries so technically 8ns vs 9ns is aprox 10% faster.

8-8-8 @ 900mhz
9-9-9 @ 1000mhz

^^ Should yield the same throughput as one another.

Bobnova · 04-01-12, 11:23 AM

I have yet to notice a real life difference when it comes to ram speed/timings.
Even comparing 1333-9-9-9-2t to 2133-7-10-7-1t.

The thing to remember about 1T/2T is that the number is how many clocks each command is sent for. In essence 2t adds one to every timing, not just the main 3 to 4 timings.
It is also worth noting that some platforms care more about timings than others and some care about frequency more than others.
Ram on the whole is a strange game, one that the 24/7 user really doesn't need to think about much IMO.

Now the benching crowd is a different story, ram is critical there.

BugFreak · 04-01-12, 12:12 PM

Great thread here. Should be a sticky!

johan851 · 04-01-12, 03:53 PM

Quote:

Originally Posted by BugFreak

Great thread here. Should be a sticky!

It's true that this is something that those of us who have been around for a while take for granted, but it's a little confusing at first if you're new.

03-31-12, 01:43 AM	Thread Starter #1
esCob4r Member Join Date: Feb 2012 Location: South Florida	Are lower numbers better for timing? Title says it all? I have G.Skill RipjawsX 1600Mhz DDR3 8-8-8-24 2T memory. Does this mean my memory is faster then those set to 9-9-9-24? __________________ Intel i5-2550K @ 4.6Ghz 1.275Vcore Asus P8Z68-V Gen 3 G.Skill RipJawsX 8G 2x4G DDR3 1600Mhz 8-8-8-24 Evga GTX 560 SC(Non Ti) Cooler Master Hyper 212+ Cooler Masters Haf 912 Cooler Masters 120MM Sickle Flow Fan x4 Cooler Masters 120MM 1200RPM Fan x2 Seagate Barracuda 7200RPM 160G HDD Seagate Barracuda 7200RPM 500G HDD Asus 48X Sata DVD/RW Drive Corsair HX1000 Power Supply
	QUOTE Thanks

03-31-12, 03:18 AM	#2
redduc900 Super Moderator Join Date: Dec 2000 Location: WCG@home in Portland	Yes... the smaller the number = the fewer the clock cycles = the faster the memory. __________________ "Taking Your Posting to the Next Level" "Unofficial Posting Guidelines" "Forum FAQs, Guidelines, and Rules" "Signatures: Content, Size and readability"
	QUOTE Thanks

03-31-12, 06:26 AM	#3
Janus67 Benching Team Leader Wannabe Admin Senior Join Date: May 2005 Location: Ohio State	Yes, although if those set to 999 are running 1t instead of 2t you are effectively running the same speeds, as running 2t pretty much adds 1 to your main timings
	QUOTE Thanks

03-31-12, 11:24 AM	Thread Starter #4
esCob4r Member Join Date: Feb 2012 Location: South Florida	ok so 8-8-8-24 2t is basically really 9-9-9-24? If so then what is the exact point of having lower 8-8-8-24 2t? I understand I could probably manually adjust it to run 1t maybe? Or maybe not? Either way thanks. Trying to learn as much as I can with building pcs, overclocking, and what not. __________________ Intel i5-2550K @ 4.6Ghz 1.275Vcore Asus P8Z68-V Gen 3 G.Skill RipJawsX 8G 2x4G DDR3 1600Mhz 8-8-8-24 Evga GTX 560 SC(Non Ti) Cooler Master Hyper 212+ Cooler Masters Haf 912 Cooler Masters 120MM Sickle Flow Fan x4 Cooler Masters 120MM 1200RPM Fan x2 Seagate Barracuda 7200RPM 160G HDD Seagate Barracuda 7200RPM 500G HDD Asus 48X Sata DVD/RW Drive Corsair HX1000 Power Supply
	QUOTE Thanks

03-31-12, 11:30 AM	#5
redduc900 Super Moderator Join Date: Dec 2000 Location: WCG@home in Portland	No... Janus was comparing the effect of 1T and 2T Command Rates to looser (9-9-9) and tighter (8-8-8) timings. __________________ "Taking Your Posting to the Next Level" "Unofficial Posting Guidelines" "Forum FAQs, Guidelines, and Rules" "Signatures: Content, Size and readability"
	QUOTE Thanks

Are lower numbers better for timing?

Internet Services

Website Development

Tech Hardware

03-31-12, 04:16 PM	Thread Starter #7
esCob4r Member Join Date: Feb 2012 Location: South Florida	Got yah. So basically it would make no real difference to use 8-8-8-24 2t vs 9-9-9-24 1t. I might have to look into that laters. __________________ Intel i5-2550K @ 4.6Ghz 1.275Vcore Asus P8Z68-V Gen 3 G.Skill RipJawsX 8G 2x4G DDR3 1600Mhz 8-8-8-24 Evga GTX 560 SC(Non Ti) Cooler Master Hyper 212+ Cooler Masters Haf 912 Cooler Masters 120MM Sickle Flow Fan x4 Cooler Masters 120MM 1200RPM Fan x2 Seagate Barracuda 7200RPM 160G HDD Seagate Barracuda 7200RPM 500G HDD Asus 48X Sata DVD/RW Drive Corsair HX1000 Power Supply
	QUOTE Thanks

03-31-12, 04:30 PM	#8
johan851 Insatiably Malcontent Senior Member Join Date: Jul 2002 Location: Seattle, WA	Right. But if you're using the subtimings, you can sometimes get to something like 8-9-8 @ 1T instead of 9-9-9 @ 1T, when maybe 8-8-8 @ 1T isn't quite stable. So there's a bit more flexibility. We're talking tiny fractions of performance difference at that level though. __________________ ASRock Z68 Extreme3 Gen3 \| 2500K @ 4.6GHz \| 2x4GB Samsung DDR3 \| GTX460 768mb 120GB Crucial M4 \| 2x 2TB Samsung F4 \| Seasonic S12 600w AV-710 --> y2 DAC --> Custom M^3 --> Custom LM3875 ChipAmp --> Modula MTs Dual Dell 2007WFP \| Watercooled
	QUOTE Thanks

03-31-12, 04:38 PM	#9
freeagent Member Join Date: Sep 2004 Location: Winnipeg!	Chances are you wouldnt see a difference running 6-6-6-20 1T vs 8-8-8 or 9-9-9, except in benches, or if you are a maniac who never leaves his pc Uncore is where its at __________________ Xeon X5690 E.S. @ Stock+Turbo Corsair H100, 4x 38x120 Panaflo Asus Rampage III Formula 3x2GB SuperTalent Chrome @ 1333Mhz 6-6-6-18 1T, 1.6v Asus GTX580 Matrix Platinum @ Stock\|LG 42" Auzentech Prelude\| NAD, Mirage, Mission, Monitor Audio OCZ RevoDrive 120GB, Intel 520 120GB 2x WD 1TB Black, Hitachi 300GB, RaptorX Corsair TX850\|CM Stacker STC-01
	QUOTE Thanks

03-31-12, 05:50 PM	Thread Starter #10
esCob4r Member Join Date: Feb 2012 Location: South Florida	I see so for gaming only it makes no real difference. Got ya. Good to learn and now. __________________ Intel i5-2550K @ 4.6Ghz 1.275Vcore Asus P8Z68-V Gen 3 G.Skill RipJawsX 8G 2x4G DDR3 1600Mhz 8-8-8-24 Evga GTX 560 SC(Non Ti) Cooler Master Hyper 212+ Cooler Masters Haf 912 Cooler Masters 120MM Sickle Flow Fan x4 Cooler Masters 120MM 1200RPM Fan x2 Seagate Barracuda 7200RPM 160G HDD Seagate Barracuda 7200RPM 500G HDD Asus 48X Sata DVD/RW Drive Corsair HX1000 Power Supply
	QUOTE Thanks

03-31-12, 08:45 PM	#11
johan851 Insatiably Malcontent Senior Member Join Date: Jul 2002 Location: Seattle, WA	Well... 6-6-6-20 1T maybe. But good luck finding memory that'll do that at speed. I get your point though - it really doesn't make a difference. But it's one more thing to tweak, and the tweaking part is fun. __________________ ASRock Z68 Extreme3 Gen3 \| 2500K @ 4.6GHz \| 2x4GB Samsung DDR3 \| GTX460 768mb 120GB Crucial M4 \| 2x 2TB Samsung F4 \| Seasonic S12 600w AV-710 --> y2 DAC --> Custom M^3 --> Custom LM3875 ChipAmp --> Modula MTs Dual Dell 2007WFP \| Watercooled
	QUOTE Thanks

04-01-12, 11:23 AM	#14
Bobnova Air Superiority Senior Member Benching Team Leader Join Date: May 2009 Location: Humboldt	I have yet to notice a real life difference when it comes to ram speed/timings. Even comparing 1333-9-9-9-2t to 2133-7-10-7-1t. The thing to remember about 1T/2T is that the number is how many clocks each command is sent for. In essence 2t adds one to every timing, not just the main 3 to 4 timings. It is also worth noting that some platforms care more about timings than others and some care about frequency more than others. Ram on the whole is a strange game, one that the 24/7 user really doesn't need to think about much IMO. Now the benching crowd is a different story, ram is critical there. __________________ "Two things are infinite: the universe and human stupidity; and I'm not sure about the the universe." -- Einstein (maybe) How to check your PSU with a multimeter. 17bXw5t51rEBXGavJFMJsC8g7HQgThUGc7 Heatware.
	QUOTE Thanks

04-01-12, 12:12 PM	#15
BugFreak Join Date: Apr 2010	Great thread here. Should be a sticky! __________________ AMD Phenom 1055T @ 3.6 ASUS M4A89TD Pro @ 301 Bus Mushkin Redline 996999 PC31600 7,8,7,24 1T eVGA nVidia 460 GTX XSPC Rasa, MCP 655, Thermochill 120.1, XSPC RX240, EK FC460 GTX Two things are infinite: the universe and human stupidity; and I'm not sure about the universe. -Albert Einstein
	QUOTE Thanks

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search