Is there any real world performance difference going from this ram to this ram

[DaMulta]

Just buy the cheap crap when it comes to memory for real world performance difference in all most ALL CASES. More GB the better is the only real thing to look at.

If you want REAL world performance increase....buy a fast SSD.....

 

[trt740]

checked my box has a blue motherboard with blue and orange video card slots and green IDE but my board is completely blue and has no other colors but is the same motherboard just a different color scheme. Makes me wonder if it not a different board flashed with a UDP4 bios and bios is the only difference.

Per my box I have a MA790XTUD4P rev1.0

 

[mastrdrver]

Well if it was then it would say what it really was on the board itself but you're giving me hope for using it with a X6.Here is some AMD approved Mushkin

Anything faster than cas8 on that board is a waste for the 4x multiplier.

 

[trt740]

Well if it was then it would say what it really was on the board itself but you're giving me hope for using it with a X6.Here is some AMD approved Mushkin

Anything faster than cas8 on that board is a waste for the 4x multiplier.

my old 1090T did N/B 3000+ at 4.0 ghz but my new one won't . I had a old thread on it but cannot find it with screen shots. I'm on my dual core atom now once my 10 year gets off nancy drew I will let ya know.


it also says the same on the board.

 

[trt740]

Okay here is a quick one for you with my crappy cas 9 ram. Pardon my 10 years olds art work. Here it at ddr3 1800 and that the best this ram will do in just about any board according to the reviews.

 

[mastrdrver]

I know we're getting a little off topic but.....

What kind of voltage did you need for 3000 cpu-nb?

Back on topic......

I'll run a few PCMark Vanatge runs on my MSI board to see if anything shows up as far as timings, etc. I've done this before briefly but can't remember the results for sure.

While I'm not sure I've been under the presumption that on the 4x multiplier cas 8 is the limit, cas 7 for 3.33x, cas 6 for 2.66x and these are limitations not dependent on the cpu but because of the way Gigabyte made the board. No bios, cpu, hard mod, etc will be able to get around these limits.

Nice artwork btw. Looks like your 10 yr old has an artistic mind.

 

[trt740]

I know we're getting a little off topic but.....

What kind of voltage did you need for 3000 cpu-nb?

Back on topic......

I'll run a few PCMark Vanatge runs on my MSI board to see if anything shows up as far as timings, etc. I've done this before briefly but can't remember the results for sure.

While I'm not sure I've been under the presumption that on the 4x multiplier cas 8 is the limit, cas 7 for 3.33x, cas 6 for 2.66x and these are limitations not dependent on the cpu but because of the way Gigabyte made the board. No bios, cpu, hard mod, etc will be able to get around these limits.

Nice artwork btw. Looks like your 10 yr old has an artistic mind.

On my old chip it was 1.375v I think.


okay update after a bit of tweaking this chip is turning out to be very good.

After messing around this 1090T isn't as good as my old one it won't prime at those setting above but it will prime at 3.950 with a N/B of 2911 at 1.488v and the cpu to N/B at 1.4v. My old one did 3000+ at 4.0ghz with 1.375v cpu to N/B vid.

This is going to prime and I'm still testing.

 

[trt740]

Well these Redlines just came 5 minutes ago and they are running at the right speed

first try with really loose timing for 2000, will tighten it later

I haven't tried higher yet but I will later on

 

[yogurt_21]

AMD and INtel handle memory different, boys. Far larger impact on AMD than Intel.

and it's been that way since ddr memory came out.

-Intel go for the highest clocked memory you can get don't worry too much about timings but don't expect good performance if you go super loose.

-AMD go for a decent clocked but conservatively timed set, don't head for the extreme in either sense but find a balance.


at least that's been my experience

imo I'd try bumping the volts and seeing if I could drop that tRCD and tRAS down as well. if you can get them to run on 6-6-6-18 you'll see a nice result in load times, boot times, etc. not necessarily gaming though.

 

[cadaveca]

PSC doesn't do 6-6-6.

 

[trt740]

I spoke too soon this ram does make a difference when run at 6 8 6 24 t1 from 9 9 9 24 t2. It is noticeably faster in windows and the more I use it the more I notice it .

 

[yogurt_21]

PSC doesn't do 6-6-6.

bummer

I spoke too soon this ram does make a difference when run at 6 8 6 24 t1 from 9 9 9 24 t2. It is noticeably faster in windows and the more I use it the more I notice it .

well that's nice, and a good thing to know. looking at parts for jan/feb myself.

 

[Tatty_Two]

I build a few systems a week (that's my job) and have a lot of opportunities to compare components side by side. The impact of memory speed and latency on real world performance is one of the things I've tested most extensively, both myself and by doing blind tests with other people trying different setups out and telling me if they detect a difference.

Flat out, no. You would need a stopwatch and an operation that takes a long time to complete in order to detect the difference. To me, that does not qualify as a "real world performance difference."

100% agreement there, I used to have two 6gb kits of DDR3 1600 mhz, one rated at CL7, one at CL8, I have set them both to CL8, ran the one kit at 1500mhz and the other at 1800mhz and can not see the slightest difference in anything.


Originally Posted by trt740
I spoke too soon this ram does make a difference when run at 6 8 6 24 t1 from 9 9 9 24 t2. It is noticeably faster in windows and the more I use it the more I notice it .
Ahhhh right, now your talking about latencies not speed (which i thought you were before), yes if you combine the 2, depending on platform to a certain degree..... you will see some differences.

 

[trt740]

yes this is very good ram. These redlines run very tight.