DFI NF4 SLI-D + OCZ PC3500 Plat LE BH-5 @260mhz 2-2-2-6 !

**eva2000** · 03-20-2005, 12:22 AM

I got a nice shock when i put my old OCZ BH-5 in DFI NF4 SLI-D for the first time.

3500+ CAA2C 0505 APMW water cooled
DFI NF4 SLI-D 310 official bios
1x eVGA 6800U PCI-E SLI
2x 512MB OCZ PC3500 Plat LE BH-5
Pioneer A05 Burner
80GB WD 800JB
580W Topower 686P6 35A/12v adjustable rail psu.

10x 260mhz 2-2-2-6 1T 3.6v
chipset = 1.6v
LDT = 1.3v
LDT 3x

1.55 + 4% = 1.55-1.57v

CPC = Enabled
CAS Latency(CL) = 2
RAS to CAS(Trcd) = 2
Min RAS# Active time(Tras) = 6
Row Precharge Time(Trp) = 2
Row Cycle Time(Trc) = 8
Row Refresh Cycle Time(Trfc) = 15
Row to Row Delay(Trrd) = 2
Write Recovery Time(Twr) = 2
Write to Read Delay(Twtr) = 1
Read to Write Delay(Trwt) = 2
Refresh Period (Tref) = 3120
Write CAS Latency(Twcl) = Auto
DRAM Bank Interleave = Disabled
Skew = Increase
Skew Value = 128
DRAM Drive Strength = Level 7
DRAM Data Drive Strength = Level 4
Max Async Latency = 7ns
Read Preamble Time = Auto
Idle Cycle Limit = Auto
Dynamic Counter = Auto
R/W Queue Bypass = 8x
Bypass Max = 4x
32 byte Granularity = Disable(8burst)

@260mhz 2-2-2-6 1T 8-15-2212

I would of thought unbuffered would be a bit higher ???

**bachus_anonym** · 03-20-2005, 12:29 AM

Why, oh why did I sell my 2x512MB Mushkin LvL II when I couldn't get it to work on Neo2

Very well done bud :thumbsup:

Also, good to see some more peeps running 32M while testing. I was tired of Sandra and 1M runs only

**eva2000** · 03-20-2005, 12:38 AM

thanks I always run the full suite of benchies

**~~Onepagebook~~** · 03-20-2005, 12:43 AM

Originally Posted by eva2000

thanks I always run the full suite of benchies

I learned.

Nice job!

**eva2000** · 03-20-2005, 12:49 AM

thanks OPB...

a few tidbits i learnt while trying to reach 260mhz with above ram

1. cpuz error message you get 'cpuz already loaded previous version' can be fixed by either upping LDT voltage or lowering LDT multiplier http://i4memory.com/showthread.php?p=2929#post2929

2. DFI NF4 SLI-D drive strength level 7 was needed to be able to pass memtest86+ test #8 loops with level 6 1000s of errors in test #8 first pass. drive strength level 7 + LDT Voltage increase from 1.3 to 1.4 was needed for memtest test #8 to pass @ 266mhz 2-2-2-6 3.6v heh and up chipset voltage from 1.6 to 1.7 to pass super pi 4m so far @266mhz 2-2-2-6 1T 3.6v

**bachus_anonym** · 03-20-2005, 12:53 AM

so, 266x10 was not stable in Windows anymore? only in memtest ?

**eva2000** · 03-20-2005, 12:54 AM

no currently running superpi 8m at 266mhz 2-2-2-6 just haven't finished the full suite yet to post screenies.. if it doesn't pass i only sometimes post screenies most of times i won't post anything lol

edit: just passed superpi 8m in 5min 40s running 16M test right now

**bachus_anonym** · 03-20-2005, 12:58 AM

That is REALLY good news for those

that still

have 2x512MB BH-5 kits

I'll be watching this thread for more updates...

Great stuff, once again

EDIT: Can I cry here ???

**eva2000** · 03-20-2005, 01:02 AM

Yup my gut was telling me to hold on to my 2x 512MB BH-5 pile of ram hence i never sold a single stick hehe. You never know when a new motherboard/cpu combo might take advantage of the legendary BH-5 memory

will keep you folks updated on how i go currently in 8th iteration of superpi 16m looking good

edit: heh i bought these on Dec 25, 2003 from mushkin.com directly heh

sold one pair and kept one for myself

**eva2000** · 03-20-2005, 01:16 AM

couldn't wait to show my excitement and shock at this

10x 266mhz 2-2-2-6 1T 3.6v
chipset = 1.7v
LDT = 1.4v
LDT 3x

1.525 + 10% = 1.6-1.62v

CPC = Enabled
CAS Latency(CL) = 2
RAS to CAS(Trcd) = 2
Min RAS# Active time(Tras) = 6
Row Precharge Time(Trp) = 2
Row Cycle Time(Trc) = 8
Row Refresh Cycle Time(Trfc) = 15
Row to Row Delay(Trrd) = 2
Write Recovery Time(Twr) = 2
Write to Read Delay(Twtr) = 1
Read to Write Delay(Trwt) = 2
Refresh Period (Tref) = 3120
Write CAS Latency(Twcl) = Auto
DRAM Bank Interleave = Disabled
Skew = Increase
Skew Value = 255
DRAM Drive Strength = Level 7
DRAM Data Drive Strength = Level 4
Max Async Latency = 7ns
Read Preamble Time = Auto
Idle Cycle Limit = Auto
Dynamic Counter = Auto
R/W Queue Bypass = 8x
Bypass Max = 4x
32 byte Granularity = Disable(8burst)

@266mhz 2-2-2-6 1T 8-15-2212

running 32m right now

**dfx** · 03-20-2005, 01:26 AM

eva2000,
nice scores u had
here's mine (same CPU stepping as well)

Gskill LA

Corsair BH5

**eva2000** · 03-20-2005, 01:34 AM

what about Gskill at same cpu speed 10x 306 ?

I'm aiming for 270mhz 2-2-2-6 hehe

**andL64** · 03-20-2005, 01:36 AM

but the problem is spi gets not as fast with the old good stuff bh-5/bh-6 like the tccd, about 5mhz more and relaxed timings beat my 2-2-2- everytime.... in spi

i didnt check this out whats better for 3D benching but i think there the bh-5 increase performance through 2-2-2

seEn

PS: nice job eva2000

**dfx** · 03-20-2005, 01:49 AM

Originally Posted by eva2000

what about Gskill at same cpu speed 10x 306 ?

I'm aiming for 270mhz 2-2-2-6 hehe

Compared TCCD n' BH5
I personally favour BH5 over TCCD anytime
But i dont run BH5 @ 3.2v ~ 3.6v (24 x 7) coz the mobo mosfets get real hot
Run them (BH5) only for benchmarks
While TCCD 24 x 7

**d@rkn1ko** · 03-20-2005, 02:04 AM

my ocz ltd bh5

3.5v bios dfi nf4 u bios 310

**eva2000** · 03-20-2005, 02:11 AM

darknico... woah 273mhz! i guess you confirmed my results hehe you removed heatspreaders ? as you list the PCB details ? i still have heatspreaders on

darknico care to share all bios settings/voltages and timings/a64tweaker as well

**bachus_anonym** · 03-20-2005, 02:12 AM

Originally Posted by d@rkn1ko

my ocz ltd bh5

3.5v bios dfi nf4 u bios 310
http://membres.lycos.fr/m4inaction/dfinf4ultrad/273%20bh5%20prime%20(Medium).JPG

hmm... nothing personal, man

but for shots like that, the OC community should come up with some sort of verification.

apart from that, Prime95 SmallFFT is not too good for test like that... almost no memory is tested during this test

**eva2000** · 03-20-2005, 02:19 AM

yeah prime smallFFT doesn't test any memory nearly heh

btw superpi 32m passed in 28min 25s @266mhz 2-2-2-6 at 3.6v

sandra buffered @7550ish

doing more benchies before uploading screenies

**d@rkn1ko** · 03-20-2005, 02:30 AM

it s blend test lot of memory

timming full auto

just cac ras to cas tras trc ras are not in auto in the bios

**eva2000** · 03-20-2005, 02:33 AM

Ah see that's pretty loose can you do superpi 1-32M run for us or at least 1M, 16M and 32M and pifast run just to compare times with what i get

**bachus_anonym** · 03-20-2005, 02:40 AM

Eva, I just dug out old SuperPI 32M run @ 260x10 with my Neo2/Winnie/2x256MB Mushkin...

You need to work on that SPi32M - I got 27:47 - over 1 min faster

It's on Server2003, though

**eva2000** · 03-20-2005, 02:45 AM

yeah win2k3 seems faster and Neo2 gives me faster bandwidth/times than DFI NF4 hehe

**TEDY** · 03-20-2005, 02:57 AM

why sisosftsandra showing 2CMD on me while i have CPC ENABLED ?

**eva2000** · 03-20-2005, 03:03 AM

Originally Posted by TEDY

why sisosftsandra showing 2CMD on me while i have CPC ENABLED ?

explained at http://i4memory.com/showthread.php?t=327 under Windows Stability Testing heading

A few notes and issues before we start:

1. Sandra 2003, 2004, 2005 all incorrectly report 1T/2T or 1CMD/2CMD in sandra bandwidth bandmark even if you have CPC enabled in bios (1T). Reason seems to be Sandra is looking for tWTR and tRTW timings instead. If tWTR or tRTW are set to 1 (very very aggressive) then Sandra reports 1CMD instead of 2CMD. Not alot of memory brands/modules will handle both set at 1. Recommended values are 2. So ignore Sandra's 1CMD/2CMD 1T/2T report.

**TEDY** · 03-20-2005, 03:03 AM

thank you eva

Thread: DFI NF4 SLI-D + OCZ PC3500 Plat LE BH-5 @260mhz 2-2-2-6 !

Thread Tools

Search Thread

Rate This Thread

Display

DFI NF4 SLI-D + OCZ PC3500 Plat LE BH-5 @260mhz 2-2-2-6 !

Bookmarks

Bookmarks

Posting Permissions