Who wants DDR2-6000 (750MHZ) memory?

[V I P E R]

Look here guys: http://www.legitreviews.com/article.php?aid=153

[Funny_S]

to bad every mainboard would melt at these speeds, the asus P5GD2-E or the Abit Fatal1ty are the only mainboards imo that will hold

[agenda2005]

Nice speed and hopefully we should be seeing DDR2-800 support soon. However those Sandra Scores are awful at that speed. 7216MB/s for DDR866.

[crotale]

Here's some 800 DDR action:
http://www.anandtech.com/memory/showdoc.aspx?i=2306&p=9

[SPL15]

I'll stick with my 1MB 36 Pin Fast Page Mode RAM in my 486 DX2 with turbo enabled. This thing Halls Butt. I got the FSB OC'd to 50Mhz and the CPU Multiplier set at 3 so I get 150 MHZ With turbo enabled!! WOO HOO.

I find it funny how it takes DDRII 866 to get 7.2 GB/s when My DDR550 @ DDR564 gets 7GB/s with 3448 timings...

So far I'm not impress by the new stuff. Hopefully that changes soon with dual core....

[agenda2005]

[QUOTE=SPL15]I'll stick with my 1MB 36 Pin Fast Page Mode RAM in my 486 DX2 with turbo enabled. This thing Halls Butt. I got the FSB OC'd to 50Mhz and the CPU Multiplier set at 3 so I get 150 MHZ With turbo enabled!! WOO HOO.

[QUOTE]

I can understand your dilemma. DDR2 have a lot of upside and it will take time to develop. The fact that DDR2 is a new technology does not mean that everyone will jump into the wagon like sheep with no Sheperd. We will adopt DDR2 when it's worth it.

[quicksilverXP]

I get the same problem... I'm near DDR800 and my bandwidth is only slightly above 7,000 in Sandra... what gives? These are even at low timings of 3-2-2-5 on Centon DDR2.

[Xassius]

I'll stick with my 1MB 36 Pin Fast Page Mode RAM in my 486 DX2 with turbo enabled. This thing Halls Butt. I got the FSB OC'd to 50Mhz and the CPU Multiplier set at 3 so I get 150 MHZ With turbo enabled!! WOO HOO.

I find it funny how it takes DDRII 866 to get 7.2 GB/s when My DDR550 @ DDR564 gets 7GB/s with 3448 timings...

So far I'm not impress by the new stuff. Hopefully that changes soon with dual core....

That's the problem. They run at 3:4 divider. For example, if the FSB is the same, and memory was running at the same latencies, running at 3:4 would only yield very little gain over 1:1.

[SPL15]

I'm just kinda perplexed that it takes DDRII that much of a clock frequency boost to get performance that regular DDR can do.

Is Intel expecting to play the clock frequency game with the FSB and memory now, like they did on CPU's for close to a decade?

I think they should focus on maximizing efficiency of current technology instead of building new stuff to boost clock frequencies to gain performance marginal performance.

My memory bandwidth is 77% to 78% efficient according to sandra 2005. I wonder what the efficiency is with the new stuff?? From the looks of the results thus far, efficiency is hurting badly just to be able to clock higher and get marginally better performance. AND the performance is either horrible or great depending on how you look at it. It's a very argueable topic.

I'm hoping all my ranting is proved to be false as the newer stuff matures and starts giving REAL definite unarguable performance benefits... but i dont see it happening too soon. Hopefully I'm wrong

[Xassius]

I'm just kinda perplexed that it takes DDRII that much of a clock frequency boost to get performance that regular DDR can do.

Is Intel expecting to play the clock frequency game with the FSB and memory now, like they did on CPU's for close to a decade?

I think they should focus on maximizing efficiency of current technology instead of building new stuff to boost clock frequencies to gain performance marginal performance.

My memory bandwidth is 77% to 78% efficient according to sandra 2005. I wonder what the efficiency is with the new stuff?? From the looks of the results thus far, efficiency is hurting badly just to be able to clock higher and get marginally better performance. AND the performance is either horrible or great depending on how you look at it. It's a very argueable topic.

I'm hoping all my ranting is proved to be false as the newer stuff matures and starts giving REAL definite unarguable performance benefits... but i dont see it happening too soon. Hopefully I'm wrong

That's the problem. They ran the memory asynchronously. I get around 77-78% efficiency with DDRII and I get about 7300mb/s with just DDR2-675 (3-2-2-2 timings).

[SPL15]

That's the problem. They ran the memory asynchronously. I get around 77-78% efficiency with DDRII and I get about 7300mb/s with just DDR2-675 (3-2-2-2 timings).


wow that's pretty good!! I didnt get what you were sayin before I get it now.

I've had no experiance with DDRII yet, so I'm just goin by articles. 7.3GB/s at DDR2-675 sounds a whole ship ton lot better than 7.2GB/s at DDR2-866!!

Maybe I'll upgrade sooner than I thought.

[Xassius]

wow that's pretty good!! I didnt get what you were sayin before I get it now.

I've had no experiance with DDRII yet, so I'm just goin by articles. 7.3GB/s at DDR2-675 sounds a whole ship ton lot better than 7.2GB/s at DDR2-866!!

Maybe I'll upgrade sooner than I thought.

And that's on a Northwood. Prescotts will yield higher sandra scores :p:
Right now, there's no point running DDR2 with the 3:4 divider.

My article :cool: :http://www.ocprices.com/index.php?rev_id=258&page=5&action=reviews

EDIT: My current DDR2 can run all the way up to DDR2-720 at 3-2-2-4 timings. Therefore, I don't need any better memory until I break the 360mhz FSB barrier (I can run 361mhz FSB, but only on x8 multi, which isn't good)


See attachment :)