A64 MEM Which should I get?

[OBLIVIONLORD]

SO anyone for benches on an A64 with loose and tight timmings?

 

[Osirus]

sure.. here you go
2.6ghz 1:1 with 3-4-8-4 1T
http://www.sevenlabs.net/a64/2.6-sisoft05_1_1.jpg

 

[Mysti_Rayne]

Osirus, why are the timings on your overclock so loose? Are they not able to reach tighter timings at that FSB?

 

[OBLIVIONLORD]

What about tight timmings? How did you get 3-4-8-4? I cant find any mobo that will allow that.

 

[Comp. B]

Not everyone posts their timings in the same manner, I'm sure that he means what we most commonly see: 3-4-4-8.

 

[cuddles]

If you're on A64 and only looking for 230-240 FSB guaranteed, all the OCZ VX modules are rated for 240 @ 2-2-2 (3.4v). Some people get 260+ with 2-2-2 timings (3.3v+), others 'only' get 240.

 

[cuddles]

I made a thread asking for people to do ingame benching using tight timmings vs loose timmings like mine and there was only a 3 fps diffrence on a P4 just like mine.

Is thise going to be the same small affect if I put what I have already on an A64 board? Just a tiny tiny diffrence of 3 fps ingame. I really dont care alot about synthetic benchmarking since it means squat compared to what happens in the game itself.

If you feel that way, then it really doesn't matter what ram you get. Running 240 FSB at CAS2 vs. CAS2.5 will cost you a pretty penny in terms of your ram purchases, and net you next to nothing in real world performance gain. Most high quality ram like this is for the benchers of the world, not gamers.

 

[bobmanfoo]

yeah tighter timings aren't really noticable, but are nice to have

 

[cuddles]

Of course they are, but if he's only interested in real world FPS gains, spending this kind of money on ram will just be a waste.

 

[Osirus]

I showed that screenshot of the 'normal' timings I use on my system to achive 289mhz 1:1.
2.5-4-8-4 1T gave me the same performance, but with MAYBE 100MB difference in bandwidth, plus I was getting error's when playing games, 3-4-8-4 1T gives me no error's and nets me the above screenshot, which is my everyday speed, Real world gaming wise, I dont notice one bit of difference in the two. CS Source at 1284x1024 on max everything gets me 90-100+ fps in gaming. As soon as I can find a way to view the FPS in farcry and HL2, Ill take down those numbers too.

 

[DestructionX]

I showed that screenshot of the 'normal' timings I use on my system to achive 289mhz 1:1.
2.5-4-8-4 1T gave me the same performance, but with MAYBE 100MB difference in bandwidth, plus I was getting error's when playing games, 3-4-8-4 1T gives me no error's and nets me the above screenshot, which is my everyday speed, Real world gaming wise, I dont notice one bit of difference in the two. CS Source at 1284x1024 on max everything gets me 90-100+ fps in gaming. As soon as I can find a way to view the FPS in farcry and HL2, Ill take down those numbers too.

HL2 FPS = cl_showfps 1

I don't know far cry though

 

[cuddles]

For HL2 use cl_showfps 2, it's a much smoother way to check FPS. Ideally you should be using Fraps to benchmark though. Press the benchmark button before a firefight and again after, it'll show you min, max, and average FPS.

 

[Osirus]

I might give it a try, Im didnt build my system to have braggin rights or anything like that, I could honestly careless about 2-3 fps if my timings are a little loser then the die hard guy who has to have 100+ fps...
Ive used netgraph 3 in counter-strike and it shows anywhere between 75-90 fps.

 

[cuddles]

I'll run some benchmarks with different timings if the original poster is still interested when I get home (Wed. night). I love running benchmarks

 

[OBLIVIONLORD]

Well if you could that would be great. I am thinking of getting a MSI neo 3 and A64 3500 mid next year hopfully for under $300 and still use my ram which I know will get to 233fsb no problem. I just dont have the luxury of going any tigher on the timmings since as it is now im stuck at 3-4-4-8 and any tighter even at 200FSB it fails the memtest.

This is why i am hoping you dont get drastic fps diffrences in games. hehehehehe

 

[Odie812]

I just thought of another option as far as your A64 memory choice goes. You might also consider the PQI PC4000 -- based on TCCD -- which is about $60 cheaper than the GSkill PC4400 and is rated for DDR500 at 2.5-3-3-7. Granted, it isn't quite the guarantee of DDR550 that the GSkill has, but, I'm almost certain that you would be able to get it there without much of an issue yet still achieve moderate clocks at 2-2-2-5 if you so desired.

See it here.

Odie

 

[cuddles]

GSKill makes great ram. Tonight (12 AM EST) when I get home from work I'll do a few quick benches for you. I'll use my 53 at stock and use 2-2-2 vs 2.5-2-2 vs. 3-3-3 like you're forced with and it'll show the differences.

Did you want Doom 3 and VST benches? 3DMark 2001?

 

[icedogg9999]

Did you want Doom 3 and VST benches? 3DMark 2001?

That would be awesome. Would you mind some Sisoft Sandra, and 3dmark03 or 3dmark05 benchies as well. I'd like to see the performance differences on this memory.

 

[orion25]

That would be awesome. Would you mind some Sisoft Sandra, and 3dmark03 or 3dmark05 benchies as well. I'd like to see the performance differences on this memory.

Yes, PLEASE!!!

 

[Twister1234]

hm,whats about the maximum tccd will be able to achieve with 6-2-2-2 1T timings?