Really Stupid Fb Ram questions

Working on trying to get whatever i can out of my old D5400XS machinewith 2 Xeon X5450's(8 physical cores), as i can't afford a nice second machine to have with me. Mostly is doing video rendering a stuff, but i have the issue of it performs worse than my core i7 920. My 920 however is set to 4Ghz, with 954 mhz ram(around 19xx at ddr speeds), my D5400XS is at 3.6ghz with 400mhz(800 at ddr) Ram. it is using 2RX4 Ram.

I was curious, would getting 2RX8, 4RX8 or 1RX4 or whatever ram be faster than my current sticks? I believe my Ram to be the limiting factor of my Machine. Run P95 with Ram, CPU's show 100% load but get no hotter than 60C. Run Small FTT's and it skips ram, it jumps upto 70C on my CPU's. Run Linpack(Oh boy) and the temps rocket to 80C and the Cooling hits max speed. I look at it and it keeps on saying 2 things. Memory, or BUS. CPU's with Linpack appear to be outputting around 96 Gflops, but my i7 only gets 48. My D540xs beats it in linpack and P95 smal FTT's. So to test it, tried gaming on both, the D5400xs performs worse with an interesting note, the CPU core the game is in is almost at 100% usage and the GPU never hits 100%. It sits around 50% usage, but when i swap GPU's(they both use GTX 570's for this test, normally the D5400XS has a Quadro 6000) in the i7, they hit 100% usage In the D5400XS i can get 100% usage when running Heaven or Valley. So I tried rendering in Sony Vegas with cpu only rendering. i7 still outpaced the D5400XS by almost 2x the speed. There's only a couple things i can think of. The BUS, The RAM, The HDD(it's on a rotating drive in the D5400XS, the i7 uses an SSD), or possibly even the CPU's. I can't figure it out anymore and would like some to please help me on this, but I am believing my memory to be the culprit.

On the bright-side, I made our team get the #1 slot for memory speeds on HW bot for FB DRAM >.> not like it's impressive with 5 other results but yeah

http://hwbot.org/submission/2585558_arkade_maxxmem_fb_dimm_dram_219.9_marks

The stress testing you are doing, and the temperatures, I wouldn't beleive to have anything to do with the memory or performance of memory. What you are seeing is just how it works. If I tested my system or any other for that matter, you would see similar temperature differences and my ram is fine.

There is something up with that system, but your measuring method is flawed. Try Maxmemm or AIDA64 and see what the bandwidth is.

Did you look at the hwbot link? It is a maxmemm result x.x

5438 - Copy (MByte/sec.)
3404 - Read (MByte/sec.)
8097 - Write (MByte/sec.)
76.9 - Latency (ns)

my goal with measuring was to try to use real world applications between the two systems. I've found little existent help anywhere about this system and little information on the best configurations. I'm lost at what to do here. I never remembered this system being fast, but i don't remember it being this damn slow. My method with P95 taking the ram out of the test, none of my other machines do that. The massive jump in heat makes it sound like it's the RAM cause the CPU's appear to be getting more done.

I did not, I saw "memory speeds" and though it was a MHz link for a contest.

So... how does that compare with similar speed sticks would be my next question...are there any similar sticks to compare to?

I COULD Rip sticks out of a Dell Poweredge 2950. They are the only sticks that i have left that are FB's. But they i think are 2RX4 as well. and 2GB instead of 4. which gives me 8GB of ram vs 16

Confirmed, they are 2RX4's as well. same speed, 667.

I meant to compare with at Hwbot.

Why would capacity matter when you are looking at speed?

Oh. And capacity matters because Video editing = Ram rape. >.> When you have alot of 1920x1080 videos which make up 200GB of HDD space, setting dynamic video preview ram usage to 12GB is helpful. The ram specs seem to be that of my current ram, except my current ram, i lowered the timming and bumped it's freq to force it to go faster.

I get that... but that isn't the point here... you are asking about SPEED not what happens when you stuff the available ram full.

Very true my friend. My thoguth was it shares the same Rank, density(i think) and mhz. They "Should" perform the same, just less size. Was curious of the difference between rank 1, 2, and 4 Ram

Now THERE is where you may see a difference in performance... though not sure how that translates to old school DDR2 FB ram...

Thats what I'm thinking, but idk which to buy and test with >.>