DDR2-1066 performance

[jimbo6846]

If you werent posting incorrect information I wouldnt have needed to contribute further than my original post #5, where I gave a fairly balanced opinion and advice.

Read the results you posted, they cannot be down to clockspeed alone as the difference is sometimes over twice the theoretical max achievable for a 10% clockspeed shift alone.
You also have taken a non real world scenario of playing games such that they are not GPU limited at all.
That test is to verify the maximum possible benchmarkable benefit between 2 different memory modules, not the same module at different speeds and not for everyday usage scenarios either.
You havent provided any worthwhile proof yet.

If you wish to prove that running faster ram on C2D will help the readers of this forum looking for advice on purchasing new hardware, please do.

I already have, you simply don't agree with it, nor have you provided any proof to the contrary. It seems like all you can come up with is "That's not what I consider a real-world usage benchmark."

 

[Nenu]

I already have, you simply don't agree with it, nor have you provided any proof to the contrary. It seems like all you can come up with is "That's not what I consider a real-world usage benchmark."

Its not that I dont agree with the tests, they show the difference between 2 different memory modules and the results in one test are over twice the maximum achievable if the same module is overclocked 10%.
The gaming tests are not representative of real world gaming, you cannot be serious that you think they hold merit.
The tests are not the same as overclocking a module so are not valid for your purpose.
The results are not translatable to real world use and you do yourself no favor by using them for your argument.

 

[jimbo6846]

Its not that I dont agree with the tests, they show the difference between 2 different memory modules and the results in one test are over twice the maximum achievable if the same module is overclocked 10%.
The gaming tests are not representative of real world gaming, you cannot be serious that you think they hold merit.
The tests are not the same as overclocking a module so are not valid for your purpose.
The results are not translatable to real world use and you do yourself no favor by using them for your argument.

You should take a second glance at the article. There's 3 results, but only 2 modules: 1 Mushkin module and 1 G-Skill, the 3rd result is the G-Skill module overclocked to 1100Mhz.

It's funny that you say "The gaming tests are not representative of real world gaming, you cannot be serious that you think they hold merit", because I'm thinking you cannot be serious that you think they don't hold merit. What do you think would have happened had they chose to run the resolution at 1280x1024? Would you expect the FPS to identical across all 3 tests?

And again you don't provide any proof of your own. At this point you're just arguing for the sake of arguing. You're bringing nothing to this discussion but a chip on your shoulder.

 

[Nenu]

That makes the results even worse as the maximum performance difference from a 10% overclocked memory module is shown to be 19.5% which cannot be true.
I believe that there is another difference in their test systems which is skewing their results.

A link has already been given in post #10 showing the real world difference of gaming using different speed ram and latency settings.
Sometimes slower memory performs faster at the same latency settings because it can be set to run with T1 wait states.
The faster ram requiring T2 wait states loses much of its performance benefit and sometimes makes gaming lose performance.

It makes no sense to overclock the ram above a C2D CPU's overclocked FSB (ie 1:1) if there is no worthwhile benefit.

edit:
I just came across a thread where you posted this 2 days ago
http://www.hardforum.com/showpost.php?p=1033117489&postcount=4

Yep, you got it. And there's very little difference between the memory speeds, like you noticed. I believe it's only noticeable in benchmarking (where the difference is presented to you in numerical form) and possibly gaming, though I'm not sure on that one.

lol

 

[jimbo6846]

That makes the results even worse as the maximum performance difference from a 10% overclocked memory module is shown to be 19.5% which cannot be true.
I believe that there is another difference in their test systems which is skewing their results.

A link has already been given in post #10 showing the real world difference of gaming using different speed ram and latency settings.
Sometimes slower memory performs faster at the same latency settings because it can be set to run with T1 wait states.
The faster ram requiring T2 wait states loses much of its performance benefit and sometimes makes gaming lose performance.

You mean the link that shows increase in performance with increase in speed clock speed? I never said there would be a huge difference.

It makes no sense to overclock the ram above a C2D CPU's overclocked FSB (ie 1:1) if there is no worthwhile benefit.

Not everything is worth the same to everyone. Some people may enjoy 3 FPS increase more than others, for example. It's obvious you wouldn't enjoy that though.

edit:
I just came across a thread where you posted this 2 days ago
http://www.hardforum.com/showpost.php?p=1033117489&postcount=4

lol

I wasn't sure then, but I am now. I'm not going to comment on the fact that you don't have anything better to do than to look up my posting history.

 

[Nenu]

You mean the link that shows increase in performance with increase in speed clock speed? I never said there would be a huge difference.

Not everything is worth the same to everyone. Some people may enjoy 3 FPS increase more than others, for example. It's obvious you wouldn't enjoy that though.

I wasn't sure then, but I am now. I'm not going to comment on the fact that you don't have anything better to do than to look up my posting history.

lol you 'might' (most games showed no benefit) get 3 fps if you can clock your DDR 667 / 800 to DDR 1181.
Not something the average overclocker will achieve.

Its hard to miss your post in a previous thread while researching information that might assist you.
It stuck out like a sore thumb.
It seems you cant make your mind up

You didnt understand the results or read the reviews conclusion:

Well there you have it, the results may be a little hard to swallow for those that shelled out big dollars for high speed memory under the impression that is was actually faster. I have to admit, while I did not expect huge gains, I did expect that on a 500MHz FSB DDR2-1181 memory would be faster! Strangely this just is not the case and as it turns out, low latency memory is far more beneficial to Core 2 Duo owners. When you consider that a 1:1 ratio on a 500MHz FSB means the memory will be operating at 1000MHz, it is hard to believe DDR2-667 has a chance, let alone believe it to be faster.

 

[jimbo6846]

EDIT: nvm, I don't care enough to continue this discussion.

 

[Nenu]

I read it, that doesn't apply to alot of current RAM modules though, unfortunately (notice that the article is 1 year old). Nowadays, there's a lot of 2x1GB modules that are rated for DDR2-800 4-4-4-12 that can OC to well over DDR2-1000 and show a substantial bandwidth/latency improvement at those speeds, even with looser timings, which I'm sure you're aware of. There's also a lot of 2x2GB modules rated at higher speeds and looser timings by default, which incidentally show a decrease in performance with tighter timings and 1:1 ratio. This is how the newer Intel chipsets work.

Lots of ifs there but no evidence that overclocking the memory higher than 1:1 is worth doing.
Modern ram will show the same differences while running on a C2D system.
If there is some memory that has problems at lower latency, my advice is avoid it.

Please tell us how do the newer Intel chipsets work in relation to those memory modules that perform slower at tighter timings?

 

[Starrbuck]

EDIT: nvm, I don't care enough to continue this discussion.

Some people just love to argue, even though they're not very good at it.

BTW, does this forum have an ignore feature? It's time mine got one entry put in it (not you).

 

[jimbo6846]

Some people just love to argue, even though they're not very good at it.

BTW, does this forum have an ignore feature? It's time mine got one entry put in it (not you).

I wouldn't ignore him, flat out. Just sort through the crap manually. If someone posts something I don't like and I want to ignore them, I just don't respond to them.

But to answer your question, I'm not sure if this forum has an actual ignore option. I have to admit that sometimes it would be nice. But if anything, those posts can be good for a laugh every now and then.

 

[EVGA-Team]

I am using the GSkill 1066 4Gig Kit, and works like a charm.

I am thinking about picking up a set of the Dominator Ram, as it is $25 at Newegg with the Rebates. I do not know if this is really ANY better than the GSkill.

What do you think?

And yes, I am a power-user.