Preliminary results with New OCZ DDR500 Gold 2x1gig kit !

[ozzimark]

Just a question, how much performance is gained through using a brainpower 808 PCB over the standard DDR reference design?

look carefully at the first picture major gave us. it's neither BP nor JEDEC reference pcb.. but a micron pcb. check out the similarities between my crucial value stick and the bottom edge of the ocz ddr500. if it's a micron pcb, this also means the same 5b D chips that crucial is using in the ballistix, so overclocking should be very similar between the two.
(i hope you don't mind me messing with your pic major )

[Major]

look carefully at the first picture major gave us. it's neither BP nor JEDEC reference pcb.. but a micron pcb. check out the similarities between my crucial value stick and the bottom edge of the ocz ddr500. if it's a micron pcb, this also means the same 5b D chips that crucial is using in the ballistix, so overclocking should be very similar between the two.
(i hope you don't mind me messing with your pic major )

Definetly not Micron chips

I thought they would be Micron chips when I purchased them based on the timmings. I PMed Tony "bigtoe" and he told me they were samsung chips. I also looked under the HS to check if I could see the small half cirlce that the Micorn have on the edge of the chips and its not there.

Also the PCB is a brain power BP808, I'll take a pic later on tonight of the pcb.

[perkam]

1T...Nice.

Perkam

[ozzimark]

hahaha, i'm retarted

restate the above post.. my crucial value is BP 808


*hides in shame*


but anyhow.. if he says it's samsung, that must mean high density TCCC, unless they have another high density chip i'm not aware of besides UCCC, which is basically the same thing

[xgman]

Samsung UCCD chips.

[dnottis]

Not yet posted.... but here is a sneak peek!

http://www.3dxtreme.net/index.php?id...rpc4000-2048p1

[Major]

Good work as usual Dean !!

[Tony]

2.85V max on these, as I posted over at the OCZ forums its best to tweak in tiny steps on the vdimm, using a modded 3.3v line on your psu or a powerstream's 3.3v pot to fine tune will help.

They do get hot so remember a fan on them.

[Major]

2.85V max on these, as I posted over at the OCZ forums its best to tweak in tiny steps on the vdimm, using a modded 3.3v line on your psu or a powerstream's 3.3v pot to fine tune will help.

They do get hot so remember a fan on them.

I didn't gain anything from going from 2.7v to 2.8v, in fact I got errors @ 2.8v.

I'm running them fine at 2.6v and IMO they would do up to 270mhz like that.

[EnJoY]

Major, get me a cpu-z valid for your OCZ top.

[Major]

OK, give me an hour or so

[Major]

Here are Pi32 and cpuz cert

I ran it at 275x10, If I could get 278x10 or more I would run it 24/7 at that but at 2750 I'm giving up 75mhz on my processor OC and thats more than I'd like to sacrifice. But it sure sounds nice 280x10 would be ideal but I don't think my X2 4400+ can handle that, it couldn't with the ballistix.

[Major]

And here are some memory benchmarks at 275x10 3-3-3-4 1t

[HiJon89]

Not yet posted.... but here is a sneak peek!

http://www.3dxtreme.net/index.php?id...rpc4000-2048p1

Great review, but I'm still looking for a 2GB vs. 1GB comparison that compares the minimum FPS when playing a game, not the average. I think this is where 2GB would have its biggest advantage especially in RAM heavy games such as BF2.

[dnottis]

Great review, but I'm still looking for a 2GB vs. 1GB comparison that compares the minimum FPS when playing a game, not the average. I think this is where 2GB would have its biggest advantage especially in RAM heavy games such as BF2.

I will try to add some more with FRAPS and timedemos and see what the results are. I dont know if the minimums will actually jump or if gameplay is just smoother having more memory to avoid HD caching. It will be interesting, gimme a few days and I'll update the review with some numbers that might help.

[Orker]

These sticks are good but the ballistix tracer seem to be the more complete package.. lower volts, tighter timings, higher bandwidth and even cheaper from where ill be getting them, which will be sometime next week

[.sentinel]

Where Orker

[rodman]

The problem with those Ballistix is they don't come in a matched set, instead your left buying 2 sep stix....I don't like that idea. The OCZ comes in matched pairs as true duel channel ram should. Correct me if I'm wronge

[Vapor]

The problem with those Ballistix is they don't come in a matched set, instead your left buying 2 sep stix....I don't like that idea. The OCZ comes in matched pairs as true duel channel ram should. Correct me if I'm wronge

You're wrong. The Ballistix aren't 'matched' per se, but all modules should work together anyway, they are the exact same type of memory after all. No doubt OCZ just tests the two together for an extremely short amount of time just for truth in advertising laws, the sticks will run together and they know that without testing.

[rodman]

The ozc works better with the ASUS A8N SLI Deluxe so I'm choosing this ram hands down. It's too bad the Ballistix is no good for my MOBO I guess thats why OCZ is the best

[Tony]

I know you guys seem to prefer the Crucial but these OCZ modules seem to have something quite sweet about them. This was a quick and dirty test as i have been playing with a new motherboard and decided to test a 3000+ and 2gigs of ram.

1000mhz overclock at 311fsb 1.6V aircooled with a FX55 heatsink(OEM) and 2gigs of 4000Gold in 166 mode for around 253fsb or so 3-4-3 at 2.65Vdimm.
Now i know its not prime but please take a note of the bandwidth, this is pretty healthy for such slack timings.

[Major]

Tony, I think everyone should remember that obtaining high HTT with any 2x1gig configuration is totally dependent on the memory controller.

Even though I was able to run the ballistix at a high HTT with a low multi, it is far from stable using the higher multis of 11x or 12x at the same memory clock. I've tested it with my San Diego 3700+, my 4600+, and my 4400+ and all of them had to be run at 2t with high memory clocks of over 270mhz or so.

I guess what I'm trying to get accross, if you look at my sig you will see I'm running the OCZ and the Ballistix at a much lower Mhz with a high multi. When you take the A64 controller into account the OCZ is on par with the Ballistix. I think the only expections would be if you have an outstanding controller (FX55, FX57 for example) and can run 2x1gig at over 270mhz without issues.

I'm thinking about getting a second set of OCZ PC4000 for my work rig to give me 4gigs total.

[dnottis]

I know you guys seem to prefer the Crucial but these OCZ modules seem to have something quite sweet about them. This was a quick and dirty test as i have been playing with a new motherboard and decided to test a 3000+ and 2gigs of ram.

1000mhz overclock at 311fsb 1.6V aircooled with a FX55 heatsink(OEM) and 2gigs of 4000Gold in 166 mode for around 253fsb or so 3-4-3 at 2.65Vdimm.
Now i know its not prime but please take a note of the bandwidth, this is pretty healthy for such slack timings.

Would that have anything to do with the "new" motherboard you are testing, or was the bandwidth that high with the DFI nf4 too?

[nugzo]

I know you guys seem to prefer the Crucial but these OCZ modules seem to have something quite sweet about them. This was a quick and dirty test as i have been playing with a new motherboard and decided to test a 3000+ and 2gigs of ram.

1000mhz overclock at 311fsb 1.6V aircooled with a FX55 heatsink(OEM) and 2gigs of 4000Gold in 166 mode for around 253fsb or so 3-4-3 at 2.65Vdimm.
Now i know its not prime but please take a note of the bandwidth, this is pretty healthy for such slack timings.

looks like something funny goin on with your memtest. i wouldnt trust that bandwidth. Its also missing information. It doesnt show the memory speed or what chipset like my memtest86 1.60 does. Ive tried about every config i can think of, including yours, and i cant get close to 3500 bandwidth. Only thing i can think of, if that is correct, is that new unnamed mobo your usin. How about some windows screenies to verify that bandwidth.

[nugzo]

Would that have anything to do with the "new" motherboard you are testing, or was the bandwidth that high with the DFI nf4 too?

heh