So what is the best memory for good overclocks yet is still stable? I originally tried a pair of Corsair TWINX1024-4000 Pro XMS sticks and my DFI LANParty UT nForce4 Ultra-D board would fail to post. I swapped the XMS for two 512MB sticks of Kingston ValueRam DDR333 c2.5 and my board has been stable ever since and faster than anything else I have. Any suggestions for an OC newbie who can't afford to simply keep trying different sticks until something works?
An interesting comparison would be an Athlon64 w/ DDR500 and a close to equivilently clocked Athlon64 w/ the HTT running at 250 to make the memory DDR500.
#32, man if i could get those here in the uk (at that price).
so Zebo, what do you mean that you can't run a64 in sync with memory. for example: if i get those:
http://www.ebuyer.com/customer/products/index.html... and a dfi lanparty ultra d with a 3000 venice, would i be able to get a decent overclock?
Quite right Zebo, what this shows is that even the dual-core A64 processors when running multiple threads get little benefit from faster memory even at the same tight timings. Also as you say there is no such thing as a synchronous memory speed that might provide a performance advantage, as all memory speeds are in reality a divider from the CPU core speed. Running good PC3200 memory at or as much as it'll go above DDR400 2-2-2 is probably the best option. And if you want 2GB, get 2x 1GB modules so you can still run them at 1T command-rate as that's a better bet than four single-sided 512MB modules.
One thing to bear in mind is if you enable Cool 'n' Quiet, that the memory may actually run faster at the lower CPU multipliers when it is set to other than DDR400 in the BIOS (DDR400 ensures the memory divider is always equal to the CPU multiplier).
#31 -- sure at a price, in the case of UTT is a pretty serious one if you ask me.. high volts.. high noise from fans.. high cost.. which may not be best usage of funds.. maybe that $150 saved is better spent on 7800GTX instead of 6800Ultra..or something like that.
I'm a long time overclocker so a couple percent faster means something. If the option is there and it runs stable then go for it. All these little percentages add up in a heavily tweaked system. Overclocking the video, CPU, memory, and bus can all add up.
PS I recommend Crucial sticks (not the 8T).. they have micron G abord..same as Ballistix for half price (but not speed binned so no guarantees).. Search around though.. many many including myself hit 2-2-2 with both 1024 x 2 and 512 x 2 configs.
1. Keep in mind anand kept 2-2-2, low latency up to 250Mhz and *still* saw little difference.. fact is only one type, well two if you include anceint but still the best BH-5, can run this bandwidth and latency, Windbond UTT. This ram seems failure prone, watchout! The volts required, eg 3.5+, to attain those LL and bandwidth are extremely dangerous without active cooling on the ram. This ram is also "untested" (UTT) from windbond adding in an extra layer of uncertainty...Add in 2-3x the cost of regualar value muskin/corsiar/OCZ/Crucial which *can* all hit 2-2-2@200 with ~2.8V make this choice pretty lame considering the marginal payoffs. If you like cyber olympics and compete on the margins 0-5% by all means go for it...but 99% of y'all could find much better application of your funds.
2. All other high bandwidth ram run crappy timings @ 250Mhz which will get stomped by LL @ 200. I've shown in forums you basically need 100mhz extra (not even 50) to hang with 200Mhz 2-2-2 when running ram 3-4-3, aka loose timings. Not worth buyin the high bandwidth stuff either when value muskin/corsiar/OCZ/Crucial which *can* all hit 2-2-2@200 with ~2.8V.
3. Overclcokers who want to run 1:1... there is no such thing as 1:1 in A64 archtecture.. all memory run async. So no problemo, no performance hit using 166/150/133 memsetting with value muskin/corsiar/OCZ/Crucial and shooting for 2-2-2@200 while clocking CPU to high heaven with proper HTT/FSB adjustments.
4. We see why AMD is'nt dieing for DDR2 this last year+.. and *when* they finally jump on board it will be at 667 instead of 533Mhz.. They can't afford the performance hit 533 will give, I'd be very surprised if 667 DDR2 is faster for AMD unless you run Crucials which is capable of 3-2-2.
In sum -- don't believe the hype. Get good cheap safe X2/A64 and DDR -- you'll be fast for a very long time.
Is it possible to measure this stuff on a multi socket board? One of AMD's big selling points of HyperTransport is that other CPU's in other sockets can get to other memory faster. I wonder what impact these faster memories would have on socket to socket memory access?
In other words socket one is direct connected to mem slot 1, the data in mem slot one is needed by a CPU in socket two, transiting the Hypertransport. How much benefit do the faster memories present?
Well, I could say this would have been a great test if only one thing: to show the effect of TIMINGS on this. I know the OCZ had sickest tight timings, but im talking about the difference of say, using 4 single gig memeory modules and then messing around with the timings, showing which is best. I know at 1 gig, most of those l33t ram flashy types of dual channel matches drop off and kinda go into hum-drum land.
If i had an X2 system it would have 2 gigs on two chips or 4 gigs on 4 chips. I couldnt see a dual proc system built for simply faster gaming, it's meant to be a cookie monster of processes.
"The ceil() function is a pretty basic mathematical function that returns the smallest integer value greater than its argument..."
Think that should be "greater than or equal to." ceil(10.0)=10. Not that it exactly matters much in this context. Anyhoo...
I'm glad AMD is further ahead of supporting mem speeds than needed, as opposed to playing catch-up. Additional options and capability for their future items will pay off down the road if not now.
#2 - The OCZ PC3500 GX is specially binned BH5. Anand had planned to use VX memory but it was out of stock as this article was being developed. OCZ binned some BH5 at 3.3V for 2-2-2 operation at DDR500 to meet the requirements of this review and turned it around in a few hours. The point is that OCZ PC3500 GX will NOT all operate at 2-2-2 at DDR500. The OCZ VX and Mushkin Redline, on the other hand, are both rated 2-2-2 at DDR500 at 3.5V. The VX/Redline are based on Winbond CH5 blanks.
#20 - There is a new official DFI BIOS in the works that is to post soon. However, as I mentioned in the nF4 Ultra roundup there are over 60 BIOS revisions avaialble for the DFI nF4 boards. Oskar Wu of DFI has posted many of the BIOS' at www.dfistreet.com, the DFI Forum website, or in the Forums at www.xtremesystems.org.
[In keeping with our recently growing practice,] "We chose three CPUs to investigate the impacts of these new memory dividers: [The three you are least likely to own]: the Athlon 64 X2 4800+ (2.4GHz/1MB L2), the Athlon 64 X2 4200+ (2.2GHz/512KB L2) and the Athlon 64 FX-57 (2.8GHz/1MB L2)."
I know this is going to sound like a dumb question, so here it is. Where did you get the BIOS update for the DFI board? I have that board, and their website only lists an update throung March, nothing as recent as July...
"According to [H] it is the San Diego core which has the improved (fixed) memory controller (see quote below). So do you have to be careful to get San Diego or is Venice ok?"
San Diego, Venice, DC Opterons, and the X2 all have the improved memory controllers...
A64s aren't bandwidth starved. We knew that much already :)
When you crank up the HT\FSB speeds you're normalyl trying to get the CPU clock speed up. Dividers just help if the memory can't keep up. :)
What's with all the OCZ+DFI love going on around here anyway? :)
According to [H] it is the San Diego core which has the improved (fixed) memory controller (see quote below). So do you have to be careful to get San Diego or is Venice ok?
"The San Diego core brings with it some very important things. Primarily, it has what AMD terms as a “more flexible memory controller.” We at HardOCP would prefer to call it a “fixed memory controller.” “Fixed” as in the older one was broken."
Well, I for one am just GLAD Anand is not dead... was begining to wonder... Maybe you got invited up to Redmond for a little chat due to you article that got pulled? At least you're alive and the M$ob did not get ya...
Great work on the article. I suppose as long as I do not play BF2 on one screen while rendering a scene in 3DStudioMax on another, I should be fine.
- Creathir
I do too, DDR333 was intended to be the last speed grade of DDR. Samsung and other memory makers had good yields of DDR400, and were having big problems with DDR2. Those have been fixed, and DDR2 is ready to go, having recently become even less expensive that DDR.
We’ve updated our terms. By continuing to use the site and/or by logging into your account, you agree to the Site’s updated Terms of Use and Privacy Policy.
37 Comments
Back to Article
Murthunder - Wednesday, July 13, 2005 - link
So what is the best memory for good overclocks yet is still stable? I originally tried a pair of Corsair TWINX1024-4000 Pro XMS sticks and my DFI LANParty UT nForce4 Ultra-D board would fail to post. I swapped the XMS for two 512MB sticks of Kingston ValueRam DDR333 c2.5 and my board has been stable ever since and faster than anything else I have. Any suggestions for an OC newbie who can't afford to simply keep trying different sticks until something works?Myrandex - Tuesday, July 12, 2005 - link
An interesting comparison would be an Athlon64 w/ DDR500 and a close to equivilently clocked Athlon64 w/ the HTT running at 250 to make the memory DDR500.semo - Tuesday, July 12, 2005 - link
#32, man if i could get those here in the uk (at that price).so Zebo, what do you mean that you can't run a64 in sync with memory. for example: if i get those:
http://www.ebuyer.com/customer/products/index.html...
and a dfi lanparty ultra d with a 3000 venice, would i be able to get a decent overclock?
Zebo - Monday, July 11, 2005 - link
It's really a shame anand did'nt do 3-3-3 like 99% of PC4200 sold runs @ 250...PrinceGaz - Monday, July 11, 2005 - link
Quite right Zebo, what this shows is that even the dual-core A64 processors when running multiple threads get little benefit from faster memory even at the same tight timings. Also as you say there is no such thing as a synchronous memory speed that might provide a performance advantage, as all memory speeds are in reality a divider from the CPU core speed. Running good PC3200 memory at or as much as it'll go above DDR400 2-2-2 is probably the best option. And if you want 2GB, get 2x 1GB modules so you can still run them at 1T command-rate as that's a better bet than four single-sided 512MB modules.One thing to bear in mind is if you enable Cool 'n' Quiet, that the memory may actually run faster at the lower CPU multipliers when it is set to other than DDR400 in the BIOS (DDR400 ensures the memory divider is always equal to the CPU multiplier).
Zebo - Monday, July 11, 2005 - link
http://forums.anandtech.com/messageview.aspx?catid...512 sticks
www.newegg.com/Product/Product.asp?Item=N82E16820146532
1024 sticks
www.newegg.com/Product/Product.asp?Item=N82E16820146545
#31 -- sure at a price, in the case of UTT is a pretty serious one if you ask me.. high volts.. high noise from fans.. high cost.. which may not be best usage of funds.. maybe that $150 saved is better spent on 7800GTX instead of 6800Ultra..or something like that.
AdamK47 3DS - Monday, July 11, 2005 - link
I'm a long time overclocker so a couple percent faster means something. If the option is there and it runs stable then go for it. All these little percentages add up in a heavily tweaked system. Overclocking the video, CPU, memory, and bus can all add up.Zebo - Monday, July 11, 2005 - link
PS I recommend Crucial sticks (not the 8T).. they have micron G abord..same as Ballistix for half price (but not speed binned so no guarantees).. Search around though.. many many including myself hit 2-2-2 with both 1024 x 2 and 512 x 2 configs.Zebo - Monday, July 11, 2005 - link
1. Keep in mind anand kept 2-2-2, low latency up to 250Mhz and *still* saw little difference.. fact is only one type, well two if you include anceint but still the best BH-5, can run this bandwidth and latency, Windbond UTT. This ram seems failure prone, watchout! The volts required, eg 3.5+, to attain those LL and bandwidth are extremely dangerous without active cooling on the ram. This ram is also "untested" (UTT) from windbond adding in an extra layer of uncertainty...Add in 2-3x the cost of regualar value muskin/corsiar/OCZ/Crucial which *can* all hit 2-2-2@200 with ~2.8V make this choice pretty lame considering the marginal payoffs. If you like cyber olympics and compete on the margins 0-5% by all means go for it...but 99% of y'all could find much better application of your funds.2. All other high bandwidth ram run crappy timings @ 250Mhz which will get stomped by LL @ 200. I've shown in forums you basically need 100mhz extra (not even 50) to hang with 200Mhz 2-2-2 when running ram 3-4-3, aka loose timings. Not worth buyin the high bandwidth stuff either when value muskin/corsiar/OCZ/Crucial which *can* all hit 2-2-2@200 with ~2.8V.
3. Overclcokers who want to run 1:1... there is no such thing as 1:1 in A64 archtecture.. all memory run async. So no problemo, no performance hit using 166/150/133 memsetting with value muskin/corsiar/OCZ/Crucial and shooting for 2-2-2@200 while clocking CPU to high heaven with proper HTT/FSB adjustments.
4. We see why AMD is'nt dieing for DDR2 this last year+.. and *when* they finally jump on board it will be at 667 instead of 533Mhz.. They can't afford the performance hit 533 will give, I'd be very surprised if 667 DDR2 is faster for AMD unless you run Crucials which is capable of 3-2-2.
In sum -- don't believe the hype. Get good cheap safe X2/A64 and DDR -- you'll be fast for a very long time.
robster3323 - Monday, July 11, 2005 - link
Is it possible to measure this stuff on a multi socket board? One of AMD's big selling points of HyperTransport is that other CPU's in other sockets can get to other memory faster. I wonder what impact these faster memories would have on socket to socket memory access?In other words socket one is direct connected to mem slot 1, the data in mem slot one is needed by a CPU in socket two, transiting the Hypertransport. How much benefit do the faster memories present?
ElJefe - Monday, July 11, 2005 - link
Well, I could say this would have been a great test if only one thing: to show the effect of TIMINGS on this. I know the OCZ had sickest tight timings, but im talking about the difference of say, using 4 single gig memeory modules and then messing around with the timings, showing which is best. I know at 1 gig, most of those l33t ram flashy types of dual channel matches drop off and kinda go into hum-drum land.If i had an X2 system it would have 2 gigs on two chips or 4 gigs on 4 chips. I couldnt see a dual proc system built for simply faster gaming, it's meant to be a cookie monster of processes.
KristopherKubicki - Monday, July 11, 2005 - link
bupkus: DDR2 comes with the M2 socket next year. At least, thats according to the roadmap.Kristopher
fishbits - Monday, July 11, 2005 - link
"The ceil() function is a pretty basic mathematical function that returns the smallest integer value greater than its argument..."Think that should be "greater than or equal to." ceil(10.0)=10. Not that it exactly matters much in this context. Anyhoo...
I'm glad AMD is further ahead of supporting mem speeds than needed, as opposed to playing catch-up. Additional options and capability for their future items will pay off down the road if not now.
brownba - Monday, July 11, 2005 - link
#1"keeps getting better and better."
huh? did you look at the benchmarks?
what is better?
Anand just showed us to save our money, we don't the very fastest ram, so I guess that is better.
Wesley Fink - Monday, July 11, 2005 - link
#2 - The OCZ PC3500 GX is specially binned BH5. Anand had planned to use VX memory but it was out of stock as this article was being developed. OCZ binned some BH5 at 3.3V for 2-2-2 operation at DDR500 to meet the requirements of this review and turned it around in a few hours. The point is that OCZ PC3500 GX will NOT all operate at 2-2-2 at DDR500. The OCZ VX and Mushkin Redline, on the other hand, are both rated 2-2-2 at DDR500 at 3.5V. The VX/Redline are based on Winbond CH5 blanks.Wesley Fink - Monday, July 11, 2005 - link
#20 - There is a new official DFI BIOS in the works that is to post soon. However, as I mentioned in the nF4 Ultra roundup there are over 60 BIOS revisions avaialble for the DFI nF4 boards. Oskar Wu of DFI has posted many of the BIOS' at www.dfistreet.com, the DFI Forum website, or in the Forums at www.xtremesystems.org.yacoub - Monday, July 11, 2005 - link
[In keeping with our recently growing practice,] "We chose three CPUs to investigate the impacts of these new memory dividers: [The three you are least likely to own]: the Athlon 64 X2 4800+ (2.4GHz/1MB L2), the Athlon 64 X2 4200+ (2.2GHz/512KB L2) and the Athlon 64 FX-57 (2.8GHz/1MB L2)."hehehe
ksherman - Monday, July 11, 2005 - link
I know this is going to sound like a dumb question, so here it is. Where did you get the BIOS update for the DFI board? I have that board, and their website only lists an update throung March, nothing as recent as July...Stinger22 - Monday, July 11, 2005 - link
One small correction..There is also an Athlon 64 3000+ that is a Socket 939 and is Revision E.
Hacp - Monday, July 11, 2005 - link
#13, I know of one other mobo that supports 3.3 volts and many other ram sticks can get 2-2-2-5 at ddr500.wien - Monday, July 11, 2005 - link
#15: It's not like they will stop making s939 CPUs the instant they launch M2. You'll be fine for a couple of years for sure.PrinceGaz - Monday, July 11, 2005 - link
#13- Lots of mobos support 3.3V RAM voltage. Only problem is they take PC66/100/133 modules rather tahn DDR :)AnnihilatorX - Monday, July 11, 2005 - link
Well in the end AMD still planned to move to DDR2To me they said S939 would last long
But they are moving to socket M2 and DDR2 next year
It would have been much better if they stick to s939 and wait for DDR3 instead
#12 I think it's next year
Viditor - Monday, July 11, 2005 - link
"According to [H] it is the San Diego core which has the improved (fixed) memory controller (see quote below). So do you have to be careful to get San Diego or is Venice ok?"San Diego, Venice, DC Opterons, and the X2 all have the improved memory controllers...
elecrzy - Monday, July 11, 2005 - link
#11: how many mobo's do you know support 3.3V+ RAM voltage and how many RAM sticks to you know support DDR500 with 2225 timing?bupkus - Monday, July 11, 2005 - link
When does AMD's roadmap start using DDR2?JustAnAverageGuy - Monday, July 11, 2005 - link
A64s aren't bandwidth starved. We knew that much already :)When you crank up the HT\FSB speeds you're normalyl trying to get the CPU clock speed up. Dividers just help if the memory can't keep up. :)
What's with all the OCZ+DFI love going on around here anyway? :)
GTMan - Monday, July 11, 2005 - link
According to [H] it is the San Diego core which has the improved (fixed) memory controller (see quote below). So do you have to be careful to get San Diego or is Venice ok?"The San Diego core brings with it some very important things. Primarily, it has what AMD terms as a “more flexible memory controller.” We at HardOCP would prefer to call it a “fixed memory controller.” “Fixed” as in the older one was broken."
http://www.hardocp.com/article.html?art=Nzg3
creathir - Monday, July 11, 2005 - link
Well, I for one am just GLAD Anand is not dead... was begining to wonder... Maybe you got invited up to Redmond for a little chat due to you article that got pulled? At least you're alive and the M$ob did not get ya...Great work on the article. I suppose as long as I do not play BF2 on one screen while rendering a scene in 3DStudioMax on another, I should be fine.
- Creathir
Joepublic2 - Monday, July 11, 2005 - link
"i remember when ddr400 wasn't official..."I do too, DDR333 was intended to be the last speed grade of DDR. Samsung and other memory makers had good yields of DDR400, and were having big problems with DDR2. Those have been fixed, and DDR2 is ready to go, having recently become even less expensive that DDR.
http://www.xbitlabs.com/news/memory/display/200507...
sprockkets - Monday, July 11, 2005 - link
i remember when ddr400 wasn't official...Joepublic2 - Monday, July 11, 2005 - link
"Potentially as a backup plan, the Rev E chips include unofficial support for memory faster than DDR400, without overclocking the Hyper Transport bus"This is an appeal to overclockers, nothing more.
"why isn't jedec officially supporting ddr500?"
JEDEC will not approve any more speed grades of DDR without Samsung and Intel's blessing.
Hacp - Monday, July 11, 2005 - link
Seems to be a conspiracy OCZ+DFI.coomar - Monday, July 11, 2005 - link
why isn't jedec officially supporting ddr500?ryanv12 - Monday, July 11, 2005 - link
Good article. This will probably help me sleep better at night when Socket M2 comes out and I still have an X2 939 :preactor - Monday, July 11, 2005 - link
hmm kinda sad that it doesnt improve performance much, but then we already know low latecy is better for amd cpus.are we going to see a review on the ocz gx ram?
SilthDraeth - Monday, July 11, 2005 - link
keeps getting better and better.