• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Asus P5K Deluxe - Important tip = Black dimm slots > Yellow dimm slots

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

eva2000

Member
Joined
Aug 5, 2002
Location
Brisbane, Australia
2x1GB Twin2X2048-10000C5DF Dominator
Asus P5K Deluxe 0311 bios

I decided to revisit my 2x1GB Twin2X2048-10000C5D Dominator memory rated at DDR-1250Mhz 5-5-5-18 at 2.4v (kindly supplied by Corsair :) ).

Originally, even with 0311 official bios the 10000C5D Dominators were a bit more than flaky on P5K Deluxe - understandably so as 10000C5D Dominators are validated only on eVGA 680i motherboards. But it seems now that I am retesting them in black dimm slots instead of yellow dimm slots, they're alot more stable particularly at 4-4-4-x timings needing less vdimm and clocking a bit tighter on memory timings in black dimm slots than in yellow dimm slots!

Note, prior to these tests all my 2x1GB memory testing to date on P5K Deluxe has been in the yellow dimm slots.

System

I managed to pull off some quick tests for low clock Super Pi Challenge @3600Mhz and managed my fastest Super Pi 1M and 32M times to date for a quad core Kentsfield Q6600 cpu
CPUZ Validation
cpuz_validated.png


Super Pi 1M
superpi-1m_13s906ms.png


Super Pi 32M
superpi-32m_single_13m10s829ms.png


From my experience on Asus P5B Deluxe, quad core kentsfield B3 stepping is around 10-20 seconds slower clock for clock than core 2 duo B2 dual core cpus. And even my personal best E6700 B2 dual core time on P5B Deluxe @3600Mhz is only 13min 08ish.

Will have to do some more tests with black dimm slots it seems.

Edit: okay confirmation from bingo13/Gary from Anandtech here, black dimm slots are designed to do this according to Asus!

The black slots are required for high memory clocks or stable low latency settings on either P5K board, in fact on the P5K3, the only way to get above 1800 in dual channel mode is to use the black slots. ASUS has the official explanation coming shortly...

Edit 2: Yup can confirm now black dimm slots rule for 10000C5D Dominators on P5K Deluxe

Black dimm slot = 625mhz 5-5-5-15 3-30-3-3-8 at 2.4v memtest able
Yellow dimm slot = 590mhz 5-5-5-18 10-42-10-10-11 at 2.45v semi stable in memtest

Huge difference!
 
According to Gary/bingo13's work with with Asus, it's be design from Asus.. official announcement of the reason is due soon apparently
 
well that sucks for me because I am running 4x1gb with all the slots populated. I am going to pull two sticks and see if i can duplicate eva2000 results.
 
Hmm I will give this a shot to night .
How far will 6400 D9's go I had it @ ddr1100 last night and that seamed to be as far as I could get .
 
That's interesting...but if it's a matter of the black slots being 'better' and the others being 'the same' as other mobos it's really just a bonus not a negative. For more standard oc'ing and not max FSB/Ram and pure benchmarking runs it shouldn't be an issue though, 450+ FSB ought to be enough for any current CPU. It could matter for Penryns though...:shrug:

Have you tried this with any other memory?
 
Last edited:
Good find! My Ballistix at 4:5 wont pass memtest, so I backed down to 5:6 then eventually went back to 1:1. I'll try this when I get home :)
 
I don't see it as problem with 4x1GB configs, i had 4x1GB micron d9gmh at 490mhz 4-4-4-9 2-25-2-2-4 tight sub timings on P5K Deluxe fine.

Haven't tried this with other memory yet focusing on 10000C5D first.

2x1GB Twin2X2048-10000C5DF Dominator
Asus P5K Deluxe 0311 bios

More tests to confirm that indeed black dimm slots on Asus P5K Deluxe seem to clock these 10000C5D Dominators alot higher with tighter timings and lower vdimm than yellow dimm slots.

System

Corsair 10000C5D Dominators:

Using memtest86+ v1.70:

Yellow slots = 590mhz 5-5-5-18 10-42-10-10-11 at 2.45v memtest semi stable
Black slots = 625mhz 5-5-5-15 3-30-3-3-8 at 2.4v memtest able

In windows had to loosen the subtimings up and drop the clocks a bit for 2.4v vdimm but still @620Mhz 5-5-5-15 8-30-8-8-8 at 2.4v is much better than Yellow slots @590mhz 5-5-5-18 10-42-10-10-11 at 2.45v which was only semi stable in memtest.

JumperFree Configuration Settings
AI Overclocking: Manual
CPU Ratio Control: Manual
- Ratio CMOS Setting: 9
FSB Frequency: 388
PCI-E Frequency: 100
DRAM Frequency: DDR2-1242
DRAM Command Rate: 2T
DRAM Timing Control: Manual
CAS# Latency: 5
RAS# to CAS# Delay: 5
RAS# Precharge: 5
RAS# Activate to Precharge: 15
TWR: 8
TRFC: 30
TWTR: 8
TRRD: 8
TRTP: 8
DRAM Static Read Control: AUTO

Transaction Booster: AUTO
Boost Level: N/A
Clock Over-Charging Mode: AUTO

CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled

CPU Voltage: 1.475v
CPU Voltage Reference: AUTO
CPU Voltage Damper: Enabled
CPU PLL Voltage: 1.50
DRAM Voltage: 2.4
FSB Termination Voltage: 1.40
North Bridge Voltage: 1.55
North Bridge Voltage Reference: AUTO
South Bridge Voltage: 1.20

Advance CPU Settings
CPU Ratio Control: Manual
- Ratio CMOS Setting: 9
C1E Suppport: Disabled
Max CPUID Value Limit: Disabled
Vanderpool Technology: Disabled
CPU TM Function: Disabled
Execute Disable Bit: Disabled
PECI: Disabled

USB Configuration
USB Functions: Enabled
Legacy USB Support: Disabled (need to enable it to detect USB flash drives)

CPUZ Validation​
cpuz_validated.png


Everest Ultimate​
everest_bandwidth.png


Short Prime95 v25.2 blend test​
prime95_blend_load.png


Super Pi 32M Single & Dual​

Single 32M
superpi-32m_single_13m48s891ms.png


Dual 32M

half way mark

superpi-32m_halfway.png


Complete

superpi-32m_dual.png


Sandra XI SP2​
Buffered



Unbuffered



Latency
 
Interesting. Thanks for posting this. I'll be putting together my P5K-Deluxe / E2160 system pretty soon, and this tip will be very handy.
 
Back