RAM Dividers - A Comparison for Core2 / DDR2 / 975X

[winston856]

Hey H_G, how's the testing coming along?

[Thraxz]

Actually, I feel that GPU GHz are king nowadays.

 

-leaves-

[Malatory]

at DDR800, the defautl divider, the memory is running at 800Mhz, but the cpu's FSB is running at 1066Mhz

 

EDIT for space...

 

 

Intel cpu's, since Pentium4 days, have had what is called a "Quad-Pumped Bus".

 

This means that the core2 cpu, is really only a 266Mhz FSB. But since we use DDR RAM, it's doubled to 533.

 

Normally you can imagine a clock cycle in a processor to be like a real clock's pendulum....it swings all the way left, then all the way.

 

This is one complete clock cycle (left to right back to original left)

 

Processors from the old days were able to execute instructions at the end of one complete clock cycle. So a 200Mhz cpu using 200Mhz RAM would only be an effective 200Mhz total in communication.

 

AthlonXP's and such moved on to DDR RAM.

 

DDR RAM was really only SDRAM. The difference is that processors were now able to execute instructions on each "edge" of the clock cycle.

 

For the 'edge' again imagine it like a real clock's pendulum, swinging back and forth.

 

 

Now Intel, in order to compete with the superior AthlonXP, decided to keep that same 200Mhz bus, but they designed the cpu to execute TWICE per clock edge.

 

so, on SDRAM, if you got 200Mhz FSB and 200Mhz RAM, you were at 200Mhz

 

on DDR SDRAM, if you had 200FSB, and 200Mhz RAM, you were at 400Mhz, because the cpu could execute instructions at each edge of the clock cycle, thereby "doubling" the data being manipulated (but we chose to explain it as a speed increase to make it easier to understand).

 

 

Intel's Pentium4, PentiumD, and Core2, they can execute TWICE at each edge of the clock cycle.

 

so if you have a 200Mhz FSB, 200Mhz RAM, you are at 800Mhz. Again, if you execute two instructions on each edge of the clock cycle, you have effectively 'quadrupled' the amount of data being manipulated.

 

The only real change now is the Intel cpu's are 266Mhz FSB

 

still quad-pumped bus

 

so

 

266x4 = 1066Mhz.

 

 

on the Core2 right now, the FSB is still 1066Mhz, but it is only accessing the memory controller @ 800Mhz (because honestly, it really only needs 533Mhz RAM)

 

so the RAM is running at 800Mhz, while the cpu is running at 1066 (reality only 533 but we'll let Intel have it their way for sake of explanation)

 

this is why you get the odd divider ratio of 4:3, 2:3, 5:4, etc. It means that the cpu and the memory are not running in synch (1:1, or 533/533).

 

But the point of all this is to test just like I did for Athlon64...how important are memory timings? How important is 1T vs 2T? How important is memory bandwidth vs cpu mhz?

 

So far, these processors are identical to the Athlon64 in the sense that CPU Mhz seems to be the king....but we'll see as I get more testing

 

Reading through this has help me pull my ram down to 3-3-3-9 timings (CPU stock speed) with a small differance in preformace. I know everyone goes for fast this and that .. but is a 100 points in pcmark05 memory test going to boost your score that much?

 

5-5-5-15 @ 800mhz memory score = 4834 (PcMark05 = 6125)

3-3-3-9 @ 533mhz memory score = 4758 (PcMark05 = 6104)

 

3dmark06 was a differance of 4 points (margin of error)

 

the differance comes more from Sandra bandwitdh testings

5-5-5-15 @ 800mhz int Buffer = 5604 float = 5596

3-3-3-9 @ 533mhz int buffer = 5252 flaot = 5254

[Angry_Games]

Hey H_G, how's the testing coming along?

 

ah it's coming...i keep getting sidetracked and having to put in different RAM for db entries etc

 

Sw4y and I just sat down and installed about 6 good games onto the rig for some 'real game' testing then we played a bunch of Tony Hawk instead and then I got sidetracked once again.

 

one problem I've encountered is these modded 84.56 drivers for SLI on non-SLI chipsets don't have SLI profiles for the newer games that came out after these drivers (like Company of Heroes) so I've had to test manual SLI profiles on them to get them to work, as well as having a slight issue with HL2 Lost Coast actually running SLI @ 1600x1200 (it worked once).

 

but I'm still on top of it so don't worry...it just takes me a while to get things done sometimes =/

[winston856]

ah it's coming...i keep getting sidetracked and having to put in different RAM for db entries etc

 

Sw4y and I just sat down and installed about 6 good games onto the rig for some 'real game' testing then we played a bunch of Tony Hawk instead and then I got sidetracked once again.

 

one problem I've encountered is these modded 84.56 drivers for SLI on non-SLI chipsets don't have SLI profiles for the newer games that came out after these drivers (like Company of Heroes) so I've had to test manual SLI profiles on them to get them to work, as well as having a slight issue with HL2 Lost Coast actually running SLI @ 1600x1200 (it worked once).

 

but I'm still on top of it so don't worry...it just takes me a while to get things done sometimes =/

 

Yeah I understand, I was just checking in for an update

[iamkoza]

so since I am up to 375x8, I am running 1:1, which means my RAM speed is doubled (again, using DDR), to 750Mhz.

 

Keep in mind however that my FSB numbers change in accordance to the cpu's stock FSB.

 

so at stock speed, I'm at 1066FSB....but at the overclock of 375x8, my FSB is now 1503Mhz.

 

Thanks for the explanation, I too am new to the intel OC side of things. With the A64 you wanted to keep your htt around 2000, is there anything like that with the new core2duo's? or is the CPU's FSB as high as you can get it to go stable?

 

I soon will have an e6600 to play around with, if i'm reading you right, I can try for say 400x7 and run the ram 1:1 (ddr2 800 ram)?

[Angry_Games]

overclocking Core2 is just like overclocking Pentium4 and AthlonXP, except dividers and RAM timing are not detrimental to your performance like they were with P4 and AthlonXP.

[Thasp]

On the ASUS P5W DH I have, I can run my DDR1000 4-4-4-12 RAM at 333 FSB, with the RAM at DDR999. It'll memtest all day.

 

At 445 FSB, where the RAM is DDR890, same timings, it fails memtest in a second. Or less.

 

Please tell me if this is the case with the DFI, whenever you have a chance and can get around to testing. I want to know if this is a lack-of-RAM-subtimings-in-the-crappy-ASUS-BIOS-issue, or an Intel issue where high FSB kills the RAM in general.

 

I can't wait to dump this awful Asus when the lanparty comes out, but if the infinity doesn't have these issues I may not even bother waiting for the new lanparty and just buy that DFI and sell the asus.. this thing has been annoying from day one. If you can test at high FSB/low multiplier and see if the RAM becomes crippled, this would be much appreciated and very helpful.

[Angry_Games]

there is no 333 divider on our board

 

there are 400, 533, 667, and 800 dividers

 

I can't say about 445 because I haven't hit that yet. 425x8 is as far as I've reached and now I got busy doing a few other things (god i hate xmas season)

[Thasp]

Oh, my bad.

 

I meant with FSB at 333, I had the RAM at the divider that gives me DDR999.

 

What I was trying to say is, at FSB 450, the CPU is stable at 3.6 GHz, the RAM is stable at DDR667. I can run prime blend test or prime CPU, or even prime RAM all day. But DDR677 sucks! I lose 400 KBps on winrar, and I use that a lot, along with other memory intensive apps. This is 1 GHz RAM, binned 4-4-4-12. I can't even run DDR890, 5-5-5-15, with loose TRRD/TRFC at this 3.6 GHz. Same for 3.52.

 

With 333 FSB, it's easy to be at 1 GHz. I don't even have to adjust TRFC and TRRD with memset to get it to be stable.

 

At 400 FSB, it gets harder to be at DDR 1 GHz. I have to adjust the TRRD from 6 to 9, and TRFC from 30 something to 42, but it works, so I have no complaints. I am not happy that I have to use a windows only program to set timings, but I'll live for now.

 

The higher the FSB, the worse the RAM. This is not something I'm used to in AMDland. The higher the FSB, so long as the CPU was stable, the RAM would work fine. There was no differential between 300 FSB with RAM 248 MHz, or 1:1 FSB 248 MHz and RAM 248 MHz. But now the higher the FSB, even if the CPU is stable and the RAM is at the same speed, the RAM's stability goes down the toilet.

 

This has happened with elpida chips binned DDR800 4-4-4-12, and with micron D9 binned for DDR1000 @ 4-4-4-12.

 

Right now I'm at 400 FSB and testing stability, but I really do want to use the CPU to its fullest. So if there are motherboards coming out, or out right now, that don't have this stupid problem, I would love to get them. I was asking if your current testing, or future testing, is showing this to be an issue.

[MACarter02]

Are we going to see this thread get updated, AG? I'm very interested... as are many others, I'm sure.

[Angry_Games]

Are we going to see this thread get updated, AG? I'm very interested... as are many others, I'm sure.

 

ya one of these days I'm gonna get around to it, I promise