You are not registered yet. Please click here to register!
Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results - i4memory.com - different look at memory
Tags Register Blogs FAQ Members List Calendar Mark Forums Read

Notices

Reply
 
Thread Tools
Old 20-12-2006, 03:47 PM   #1
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,018
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results

Intel Core 2 Duo E6600
+
Asus 680i Striker Extreme SLI motherboard Results


After showing the Asus 680i Striker Extreme bios and setup photos (bypass photos to Striker board info/links/downloads), it's time to get down to some testing and results



System:
  • Intel Core 2 Duo E6600 QPGE B1 Step 5
  • Scythe Infinity heatsink with mounting mod + Thermaltake 120x25mm 81cfm fan
  • 128MB FX5200 PCI / 256MB BFG 7800GTX / Gainward 8800GTS/SLI
  • Asus 680i Striker Extreme rev1.0 with initial official 302 bios
  • 1x 160GB Seagate 7200.7 SATA NCQ on SATA2 port
  • Pioneer DVR-05 DVD Burner
  • 600W OCZ EvoStream Modular Power supply

Memory to be used:
Room temp: 23.3C to 30.8C average

@3600Mhz - 9x400FSB at 1.39v

As always start of with a known stable overclock on my E6600 which is 3600Mhz at between 1.371v and 1.42v on various motherboards. On Asus 680i Striker Extreme it seems 3600Mhz handles fine at 1.39v vcore under load so far. I used 2x512MB OCZ PC5400 Platinum XTC 3-3-3-12 memory for the tests.

Settings used:
AI Tuning: Manual
Nvidia GPU Ex: Disabled
LinkBoost: Disabled

FSB - Memory Clock Mode: Unlinked
x FSB - Memory Raio: N/A
FSB (QDR), Mhz: 1600 (means 400FSB)
Actual FSB (QDR), Mhz: 1600
MEM (DDR), Mhz: 800 (means 400mhz speeds)
Actual MEM (DDR), Mhz: 880.2

Vcore Voltage: 1.525 set (1.42v bios/1.41v windows idle/1.39v windows load)
Memory Voltage: 2.100v
1.2v HT Voltage: 1.3v
NB Core Voltage: 1.4v
SB Core Voltage: 1.55v
CPU VTT Voltage: 1.4v
DDRII Controller Ref Voltage: AUTO
DDRII Channel A Ref Voltage: AUTO
DDRII Channel B Ref Voltage: AUTO

tCL (CAS Latency): 4
tRCD: 4
tRP: 4
tRAS: 12
Command Per Clock (CMD): 2 clock (2T)

Advance Memory Settings
tRRD: AUTO
tRC: AUTO
tWR: AUTO
tWTR: AUTO
tREF: AUTO
Async Latency: AUTO

CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
MCP PCIE Spread Spectrum: Disabled
SATA Spread Spectrum: Disabled
LDT Spread Spectrum: Disabled

CPU Internal Thermal Control: Disabled
Limit CPUID MaxVal: Disabled
Enhanced C1 (C1E): Disabled
Execute Disable Bit: Disabled
Virtualization Technology: Disabled
Enhanced Intel Speedstep Tech: Disabled
LDT Frequency: 5x

PCIEX16_1 Frequency (Mhz): 100
PCIEX16_2 Frequency (Mhz): 100
PCIEX16_3 Frequency (Mhz): 100
SPP<->MCP Ref Clock, Mhz: 200


Prime95 v25.1 multi-threaded dual small FFTs priming for ~1hr 18mins before I stopped it. Interesting that even coretemp reported cpu temps fluctuated alot more jumping between 59C to 68C while priming.


Post Prime idle temps:


Asus PC Probe2 reported voltages vs Digitech QM1538 Digital Multimeter measurements while under load @2400mhz at 1.3875v bios set (1.28v vcore idle / 1.26v vcore load).

+3.3v rail = 3.26v vs 3.387v real
+5v rail = 4.78-4.89v vs 5.02v real
+12v rail = 11.78v vs 12.23v real

Last edited by eva2000; 28-12-2006 at 05:06 AM.
eva2000 is offline   Reply With Quote
Old 20-12-2006, 03:48 PM   #2
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,018
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Re: Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results

Now that we have a base reference overclock to fall back on @3600Mhz at 1.39v, it's time to mess around with the memory timings within bios and within Memset v3.1 beta. Still using 2x512MB OCZ PC5400 Platinum XTC 3-3-3-12 memory for the tests initially then will try other memory as well.



Memory timings
Being the first time using a 680i platform, there's so many options in bios to play with. So I spent a couple of hours using 2x512MB OCZ PC5400 Platinum XTC 3-3-3-12 and just playing in memtest86+ v1.65 to get a feel of what the main and advance timings could or could not do. I started with 9x333FSB with unlinked memory clocked at 333mhz 3-3-3-12 with 2T and 2.1v as the base reference setting. Then proceeded to tweak each memory timing/advance timing and see how it changed performance.

I'll use the following to quickly test each setting and graph the average result after 3 runs of each test:
Format for testing:
Super Pi v1.50
1M run #1:
1M run #2:
1M run #3:
1M average:

2M run #1:
2M run #2:
2M run #3:
2M average:

Mbench
Access latency run #1:
Access latency run #2:
Access latency run #3:
Access latency average:

Read datarate (INT) run #1:
Read datarate (INT) run #2:
Read datarate (INT) run #3:
Read datarate (INT) average:
Write datarate (INT) run #1:
Write datarate (INT) run #2:
Write datarate (INT) run #3:
Write datarate (INT) average:

Read datarate (MMX) run #1:
Read datarate (MMX) run #2:
Read datarate (MMX) run #3:
Read datarate (MMX) average:
Write datarate (MMX) run #1:
Write datarate (MMX) run #2:
Write datarate (MMX) run #3:
Write datarate (MMX) average:

Read datarate (SSE) run #1:
Read datarate (SSE) run #2:
Read datarate (SSE) run #3:
Read datarate (SSE) average:
Write datarate (SSE) run #1:
Write datarate (SSE) run #2:
Write datarate (SSE) run #3:
Write datarate (SSE) average:
Results @9x333FSB = 3000Mhz with unlinked
2x512MB OCZ PC5400 Platinum XTC 3-3-3-12@333mhz 3-3-3-12


Bios settings Used:
Will highlight bios options in bold for any subsequent tests below which have changed from the initial settings outlined directly below.
AI Tuning: Manual
Nvidia GPU Ex: Disabled
LinkBoost: Disabled

FSB - Memory Clock Mode: Unlinked
x FSB - Memory Raio: N/A
FSB (QDR), Mhz: 1333 (means 333FSB)
Actual FSB (QDR), Mhz: 1332
MEM (DDR), Mhz: 667 (means 333mhz speeds)
Actual MEM (DDR), Mhz: 666.8

Vcore Voltage: 1.3875v
Memory Voltage: 2.100v for 1T/2T stock tests and 2.200v for 1T tweaked
1.2v HT Voltage: 1.3v
NB Core Voltage: 1.4v
SB Core Voltage: 1.55v
CPU VTT Voltage: 1.4v
DDRII Controller Ref Voltage: AUTO
DDRII Channel A Ref Voltage: AUTO
DDRII Channel B Ref Voltage: AUTO

tCL (CAS Latency): 3
tRCD: 3
tRP: 3
tRAS: 12
Command Per Clock (CMD): 2 clock (2T)

Advance Memory Settings
tRRD: AUTO
tRC: AUTO
tWR: AUTO
tWTR: AUTO
tREF: AUTO
Async Latency: AUTO

CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
MCP PCIE Spread Spectrum: Disabled
SATA Spread Spectrum: Disabled
LDT Spread Spectrum: Disabled

CPU Internal Thermal Control: Disabled
Limit CPUID MaxVal: Disabled
Enhanced C1 (C1E): Disabled
Execute Disable Bit: Disabled
Virtualization Technology: Disabled
Enhanced Intel Speedstep Tech: Disabled
LDT Frequency: 5x

PCIEX16_1 Frequency (Mhz): 100
PCIEX16_2 Frequency (Mhz): 100
PCIEX16_3 Frequency (Mhz): 100
SPP<->MCP Ref Clock, Mhz: 200
2T vs 1T vs 1T Tweaked

Graphed results with actual data further below. The graph represents the average of 3 runs for each test.





tCL (CAS Latency) - tRCD - tRP - tRAS - Command Per Clock (CMD)
tRRD - tRC - tWR - tWTR - tREF - Async Latency

3-3-3-12-2T
AUTO-AUTO-AUTO-AUTO-AUTO-AUTO




Super Pi v1.50
1M run #1: 17.188s
1M run #2: 17.187s
1M run #3: 17.187s
1M average: 17.1873s

2M run #1: 44.344s
2M run #2: 44.297s
2M run #3: 44.313s
2M average: 44.318s

Mbench
Access latency run #1: 60.0 ns (180 clocks)
Access latency run #2: 60.0 ns (180 clocks)
Access latency run #3: 60.0 ns (180 clocks)
Access latency average: 60.0 ns (180 clocks)

Read datarate (INT) run #1: 6876 MB/s
Read datarate (INT) run #2: 6873
Read datarate (INT) run #3: 6840
Read datarate (INT) average: 6863 MB/s
Write datarate (INT) run #1: 2246 MB/s
Write datarate (INT) run #2: 2291
Write datarate (INT) run #3: 2123
Write datarate (INT) average: 2220 MB/s

Read datarate (MMX) run #1: 7344 MB/s
Read datarate (MMX) run #2: 7294
Read datarate (MMX) run #3: 7335
Read datarate (MMX) average: 7324.33 MB/s
Write datarate (MMX) run #1: 2226 MB/s
Write datarate (MMX) run #2: 2278
Write datarate (MMX) run #3: 2035
Write datarate (MMX) average: 2179.67 MB/s

Read datarate (SSE) run #1: 7816 MB/s
Read datarate (SSE) run #2: 7756
Read datarate (SSE) run #3: 7764
Read datarate (SSE) average: 7778.67 MB/s
Write datarate (SSE) run #1: 6005 MB/s
Write datarate (SSE) run #2: 6012
Write datarate (SSE) run #3: 5998
Write datarate (SSE) average: 6005 MB/s

Reboot system and set 1T in bios
3-3-3-12-1T
AUTO-AUTO-AUTO-AUTO-AUTO-AUTO




Super Pi v1.50
1M run #1: 17.109s
1M run #2: 17.141s
1M run #3: 17.094s
1M average: 17.1157s

2M run #1: 43.078s
2M run #2: 43.015s
2M run #3: 43.047s
2M average: 43.0467s

Mbench
Access latency run #1: 55.7 ns (167 clocks)
Access latency run #2: 55.7 ns (167 clocks)
Access latency run #3: 55.7 ns (167 clocks)
Access latency average: 55.7 ns (167 clocks)

Read datarate (INT) run #1: 7114 MB/s
Read datarate (INT) run #2: 7145
Read datarate (INT) run #3: 7138
Read datarate (INT) average: 7132.33 MB/s
Write datarate (INT) run #1: 2476 MB/s
Write datarate (INT) run #2: 2482
Write datarate (INT) run #3: 2373
Write datarate (INT) average: 2443.67 MB/s

Read datarate (MMX) run #1: 7644 MB/s
Read datarate (MMX) run #2: 7656
Read datarate (MMX) run #3: 7664
Read datarate (MMX) average: 7654.67 MB/s
Write datarate (MMX) run #1: 2473 MB/s
Write datarate (MMX) run #2: 2478
Write datarate (MMX) run #3: 2371
Write datarate (MMX) average: 2440.67 MB/s

Read datarate (SSE) run #1: 8096 MB/s
Read datarate (SSE) run #2: 8165
Read datarate (SSE) run #3: 8168
Read datarate (SSE) average: 8143 MB/s
Write datarate (SSE) run #1: 6018 MB/s
Write datarate (SSE) run #2: 5998
Write datarate (SSE) run #3: 5982
Write datarate (SSE) average: 5999.33 MB/s
Reboot system and set 1T in bios and tightened advance memory timings in bios and memset.
3-3-3-12-1T
4-12-4-5-AUTO-AUTO




Super Pi v1.50
1M run #1: 17.109s
1M run #2: 17.109s
1M run #3: 17.110s
1M average: 17.1093s

2M run #1: 42.469s
2M run #2: 42.469s
2M run #3: 42.469s
2M average: 42.469s

Mbench
Access latency run #1: 55.7 ns (167 clocks)
Access latency run #2: 55.7 ns (167 clocks)
Access latency run #3: 55.7 ns (167 clocks)
Access latency average: 55.7 ns (167 clocks)

Read datarate (INT) run #1: 7170 MB/s
Read datarate (INT) run #2: 7150
Read datarate (INT) run #3: 7095
Read datarate (INT) average: 7138.33 MB/s
Write datarate (INT) run #1: 2651 MB/s
Write datarate (INT) run #2: 2643
Write datarate (INT) run #3: 2578
Write datarate (INT) average: 2624 MB/s

Read datarate (MMX) run #1: 7676 MB/s
Read datarate (MMX) run #2: 7659
Read datarate (MMX) run #3: 7618
Read datarate (MMX) average: 7651 MB/s
Write datarate (MMX) run #1: 2643 MB/s
Write datarate (MMX) run #2: 2638
Write datarate (MMX) run #3: 2564
Write datarate (MMX) average: 2615 MB/s

Read datarate (SSE) run #1: 8178 MB/s
Read datarate (SSE) run #2: 8175
Read datarate (SSE) run #3: 8179
Read datarate (SSE) average: 8177.33 MB/s
Write datarate (SSE) run #1: 5997 MB/s
Write datarate (SSE) run #2: 6000
Write datarate (SSE) run #3: 5999
Write datarate (SSE) average: 5998.67 MB/s

Last edited by eva2000; 20-12-2006 at 04:23 PM.
eva2000 is offline   Reply With Quote
Old 20-12-2006, 03:48 PM   #3
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,018
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Re: Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results

Before I continue any further, I decided to make use of Asus 680i Striker Extreme's 3 include onboard thermal sensors. They're 2pin header thermal sensors you can attach to the motherboard to allow another 3 thermal monitoring readings from bios as well as Asus PC Probe2.

Here you can see the additional thermal sensors labeled as follows

OPT1 Temp = Mosfet heatpipe heatsink temps (near I/O panel)
OPT2 Temp = NB heatpipe heatsink temps
OPT3 Temp = SB heatpipe heatsink temps (shoved sensor inside the tiny hole in SB heatsink so probably alot higher readings than OPT1 and OPT2 which are stuck to heatsinks.

Idle


Prime95 Load


Room temp: ~29.5-30.8C
Cooling used: Scythe Infinity with mount mod + 1x Thermaltake 120x25mm 81cfm push-pull
Bios idle: 30-31C
Windows idle: 40-41C
Windows load: 53-54C
Rise above ambient delta = 22.2-24.5C
Rise above windows idle delta = 12-14C

As you can see today's a hot day here with 29.5C room temps. But the delta temps falls inline with Scythe Infinity air testing on E6600 with Intel D975XBX done here http://i4memory.com/showthread.php?t=3218 although there's 2x 120x25mm fans instead of one for Asus 680i Striker Extreme test above.

E6600 @2400Mhz at 1.254v + Intel D975XBX Bad Axe
Scythe Infinity + 2x Thermaltake 120x25mm 81cfm push-pull

Room temp: ~27-29C
Cooling used: Scythe Infinity + 2x Thermaltake 120x25mm 81cfm push-pull
Bios idle: 40-43C
Windows idle: 35-36C
Windows load: 50-51C
Rise above ambient delta = 22-24C
Rise above windows idle delta = 15-16C
So I guess the mounting mod works fine

Last edited by eva2000; 20-12-2006 at 04:31 PM.
eva2000 is offline   Reply With Quote
Old 20-12-2006, 03:49 PM   #4
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,018
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Re: Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results

Asus 680i Striker Extreme & other downloads
Well I just updated again using USB flash drive in Asus bios' EZ Flash from 0402 to 0505 official bios and it has fixed my heavy vcore undervolting from bios set value issue!

1.3875v vcore on:
0302 official bios = 1.28v bios and 1.26v under load in windows
0402 beta bios = 1.28v bios and 1.26v under load in windows
0505 official bios = 1.36 bios and 1.34v under load in windows

Lastest Official Bios:
http://support.asus.com/download/dow...iker%20Extreme

1002 bios
ftp://ftp.asus.com/pub/ASUS/mb/socke...treme/1002.zip
1.Fix incorrect display of Vcore value when above 1.80V.
2.Fixed CPU Vcore cannot be set by user when Overclocking at some frequency.
3.Support new CPUs, please refer to our website at: http://support.asus.com/cpusupport/c...iker%20Extreme
Indeed it seems from user reports that vcore issue is fixed http://www.xtremesystems.org/forums/...d.php?t=134117

0701 bios
http://dlsvr01.asus.com/pub/ASUS/mb/...eme/SE0701.zip
Striker Extreme Release BIOS 0701
1. Fixed NB Core voltage setting error when CMOS fail
2. Enable support for Conroe 1333MHz processor
3. Fixed VID incorrect after OC to 1201Mhz(FSB)
4. Added New Micro Code to support new CPU
5. Finetune CPU temperature calculation algorithm
6. Displays "Press TAB to display BIOS POST Message" string even when full screen logo is enabled.
7. Correct incorrect info under Summary screen.
8. Correct "SLI-Ready Memory Detected" display problem at POST
9. Fixed CPU not restore back to original frequncy after CPR
10. Add help string for HPET item.
-----------------------------------------------------------------
0505 bios
http://dlsvr01.asus.com/pub/ASUS/mb/...eme/SE0505.zip
Striker Extreme release BIOS version 0505
1. Support Max LDT Frequency from 5X to 8X
2. Fine tune CPU Temperature detection code
3. Fixed some default value can't save when using BIOS Profile function.
4. Enable adjustments for tRD and TrFC item
5. Enhance DDR667 memory compatibility
6. Enhance performance for Corsair CM2X1024-9136C5D
7. Fixed CPU Voltage max option drops to 1.60V when set Multi-Language option to 'Tri Chinese' , 'Sim Chinese' or 'Japanese'.
Also reportedly from Gary Key (bingo13) at Anandtech, official 0505 bios improves memory clocking for certain modules. G.skill 6400PHU2 2GBHZ Micron D9GMH modules are one of these modules apparently. I also noticed CPU and LDT spread spectrum are greyed out as DISABLED on 0505 official bios by default.

-----------------------------------------------------------------

Comments and opinions:
Have to put this somewhere so thought why not this post. I'll keep updating this as I go along
  1. #1 outstanding feature of the Asus 680i Striker Extreme is it's ease of recovering from a failed overclock or overclocking too far! It's nice that Asus included a Clear CMOS button/switch on the board, but I have not yet once during overclocking tests needed to Clear CMOS when I clocked too far or tightened ram timings too much. Didn't even have to power down the board if it doesn't boot after exiting bios. All I did was hit the power reset switch on the motherboard and hit DEL key and the board loads back up from the bad settings. Then just boot into bios and change the settings to more suitable config and off you go. This comes in handy especially with the wide variety of bios settings and ram timings you can tweak or overtweak
  2. With so many options in bios I'd recommend you use memtest86+ v1.65 to test your memory even before you install windows to ensure your memory timings and settings all are decent enough to not error out.
  3. Don't over tighten your cpu heatsink mounting as I found out, too tight and the motherboard doesn't boot up until you loosen the mount.
  4. In bios there's two settings I would disable immediately and that is GPU Ex and Linkboost which allow the board to automatically overclock your nVidia GPU cards and PCI-E and MCP HT Bus respectively. You don't need to enable this as you'd want to do the overclocking manually
  5. Vcore is severely undervolted, bios set 1.3875v vcore = 1.28v idle and 1.26v load. Bios set 1.675v vcore = 1.57v idle and 1.55v load. Update: looks like later bioses sort of fix that as reports of 402/405 bios showing vcore closer to what's set. Haven't tried any bioses other than 302 that the board came with so far.
  6. Vdimm is slightly overvolted, i.e. 2.1v set = 2.14v vdimm, 2.15v vdimm set = 2.19v, 2.2v set = 2.25v vdimm.
  7. Asus PC Probe2 under reports the psu rails as you can see from my digital multimeter tests here.
  8. Asus Probe2 occasionally has a bug dropping the displayed voltage reading down to 0.00v http://www.xtremesystems.org/forums/...d.php?t=126949 It could be bios related. I'm still on 302 bios while 402, 405, 505, 508 bioses are available.


Bios template for 680i Striker users to report their settings:
AI Tuning: Manual
Nvidia GPU Ex: Disabled
LinkBoost: Disabled

FSB - Memory Clock Mode: Unlinked
x FSB - Memory Raio: N/A
FSB (QDR), Mhz: 1066 (means 266FSB)
Actual FSB (QDR), Mhz: 1066.7
MEM (DDR), Mhz: 667 (means 333mhz speeds)
Actual MEM (DDR), Mhz: 666.7

Vcore Voltage:
Memory Voltage:
1.2v HT Voltage:
NB Core Voltage:
SB Core Voltage:
CPU VTT Voltage:
DDRII Controller Ref Voltage:
DDRII Channel A Ref Voltage:
DDRII Channel B Ref Voltage:

tCL (CAS Latency): 4
tRCD: 4
tRP: 4
tRAS: 12
Command Per Clock (CMD): 2 clock (2T)

Advance Memory Settings
tRRD: AUTO
tRC: AUTO
tWR: AUTO
tWTR: AUTO
tREF: AUTO
Async Latency: AUTO

CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
MCP PCIE Spread Spectrum: Disabled
SATA Spread Spectrum: Disabled
LDT Spread Spectrum: Disabled

CPU Internal Thermal Control: Disabled
Limit CPUID MaxVal: Disabled
Enhanced C1 (C1E): Disabled
Execute Disable Bit: Disabled
Virtualization Technology: Disabled
Enhanced Intel Speedstep Tech: Disabled
LDT Frequency: 5x

PCIEX16_1 Frequency (Mhz): 100
PCIEX16_2 Frequency (Mhz): 100
PCIEX16_3 Frequency (Mhz): 100
SPP<->MCP Ref Clock, Mhz: 200
Relevant user links and info:
Just compiling useful info to help me and others get started. While P5N32-E SLI is a different board from Striker Extreme, it's useful to see what bios settings/voltages folks use to achieve what overclocks they manage to pull off
  1. OCER.net Striker review http://www.ocer.net/article/a_show.php?id=8742 (will need google language translator to read as it's in Chinese)
  2. Asus 680i, P5N32-E Sli Results http://www.xtremesystems.org/forums/...d.php?t=123227
  3. 680i Tips, Glitchs, Things to Know and More http://www.xtremesystems.org/forums/...d.php?t=124363
  4. ASUS P5N32-E/Striker SLI Tips and Tricks Thread http://www.xtremesystems.org/forums/...d.php?t=125790
  5. Asus Striker Extreme, E6600, small review http://www.xtremesystems.org/forums/...d.php?t=124795
  6. Asus P5N32-E SLI 680i Bios Settings http://www.xtremesystems.org/forums/...d.php?t=125718
  7. Official ASUS ROG EXTREME STRIKER 680i Result http://www.xtremesystems.org/forums/...d.php?t=123321
  8. 680i Overclocking thread http://www.xtremesystems.org/forums/...d.php?t=122683

Reviews:
Some of the reviews online from around the world. Please pay attention to the date of the review and bioses used in these reviews as there's usually a natural progression of improvement and bug fixes as the bios matures
Regarding SLI/Sound:
Still testing ram/settings etc and have yet to try SLI so no idea on sound issue yet. But Asus Striker is different from reference 680i in that it moved the sound off the motherboard to a separate daughter board card which is x1 PCI-E based and slots into the PCI-E slot. From what a few folks online have experienced they don't have sound issues when SLI is enabled due to this difference. I can't confirm it until I run SLI later though.

The separate sound card and it's array microphone



Other notes/links:
Nvidia statement on December 18, 2006 regarding 680i SATA corruption issues here
Drew Henry Interview

Drew Henry, General Manager of MCP Business at NVIDIA, took time to answer some pointed questions that will hopefully give you the solid answers you need in reference to the nForce 680i problems.

Q: Concerning the 680i issues, what EXACTLY is the problem? BIOS, driver, MCP/chipset, or other hardware? Or combination thereof?

A: The problem that some users were experiencing related to signal timings on the motherboard. The new BIOS (P32) corrects this.

Q: Is the fix a workaround that compromises other technology on the motherboard?

A: Absolutely not. We should have caught this in our internal QA process. Unfortunately, we didn't, and we have adjusted internal processes to make sure that this does not happen again. As the manufacturer of these products, we need to be absolutely certain that the quality of the product that is brought to market is top-notch and that the experience that consumers will be receiving is nothing but positive.

Q: Are there “bad” 680i MCPs/chipsets out there?

A: No. This is not a MCP issue. It's really a board issue, and in this case, only with the motherboards that were designed and manufactured by NVIDIA. In fact, boards built by other manufacturers that use the same MCPs, including the ASUS Striker Extreme, ASUS P5N32-E SLI, and the ASUS P5N-E SLI motherboards, are NOT affected. When you hear of customer issues, your first thought is "how widespread is this," but it was pretty clear early on that it related only to the boards that we designed and manufactured. Talk about embarrassing!

Q: Why would the issue impact one board and not another on the same exact hardware/software install?

A: There are statistical variations in electrical characteristics between boards, which in some cases, can mask the underlying issue. The BIOS adjusts timing in a way that avoids this condition.

Q: Is this P23 BIOS going to cure all 680i users' SATA problems?

A: We believe so. We are releasing a BETA BIOS version so that customers can use it. We are completing our testing and will release a final version soon.

Q: So what do you have to say customers that have had to endure this frustrating set of events?

A: We are sorry they had to go through it. We've spent all of this time, money and engineering resources to make the best possible platform for Intel CPUs, and then just as our customers and partners are experiencing awesome success with the new NVIDIA nForce 680i SLI motherboard, this happens. The bottom line is that our QA process missed this condition. Once we identified the problem it became our top priority to solve it. The information provided by the forum members proved to be very valuable. They were really instrumental in helping us identify this issue and provide a solution.

Last edited by eva2000; 18-02-2007 at 02:23 AM.
eva2000 is offline   Reply With Quote
Old 20-12-2006, 03:49 PM   #5
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,018
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Re: Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results

@3811Mhz - 9x423FSB 1:1 unlinked at 1.55v


Time to push the system further @3811Mhz - 9x423FSB 1:1 unlinked at 1.55v under load on air.
AI Tuning: Manual
Nvidia GPU Ex: Disabled
LinkBoost: Disabled

FSB - Memory Clock Mode: Unlinked
x FSB - Memory Raio: N/A
FSB (QDR), Mhz: 1694 (means 423FSB)
Actual FSB (QDR), Mhz: 1694
MEM (DDR), Mhz: 847 (means 423mhz speeds)
Actual MEM (DDR), Mhz: 847

Vcore Voltage: 1.675v = 1.57v windows idle/1.55v load
Memory Voltage: 2.100v for 2T stock tests and 2.200v for 1T tweaked
1.2v HT Voltage: 1.4v
NB Core Voltage: 1.4v
SB Core Voltage: 1.55v
CPU VTT Voltage: 1.4v
DDRII Controller Ref Voltage: AUTO
DDRII Channel A Ref Voltage: AUTO
DDRII Channel B Ref Voltage: AUTO

tCL (CAS Latency): 4
tRCD: 4
tRP: 4
tRAS: 12
Command Per Clock (CMD): 2 clock (2T) vs 1 clock (1T)

Advance Memory Settings
tRRD: AUTO
tRC: AUTO
tWR: AUTO
tWTR: AUTO
tREF: AUTO
Async Latency: AUTO

CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
MCP PCIE Spread Spectrum: Disabled
SATA Spread Spectrum: Disabled
LDT Spread Spectrum: Disabled

CPU Internal Thermal Control: Disabled
Limit CPUID MaxVal: Disabled
Enhanced C1 (C1E): Disabled
Execute Disable Bit: Disabled
Virtualization Technology: Disabled
Enhanced Intel Speedstep Tech: Disabled
LDT Frequency: 5x

PCIEX16_1 Frequency (Mhz): 100
PCIEX16_2 Frequency (Mhz): 100
PCIEX16_3 Frequency (Mhz): 100
SPP<->MCP Ref Clock, Mhz: 200
Both 2T and 1T Super Pi 32M were run with 4x cpuz windows open, Asus PC Probe2 running, memset v3.1 beta open as I was also monitoring the PC Probe2 reported temps while 32M ran. 2T managed to screen capture half way mark but 1T I forgot to do the half way mark screen capture

Super Pi 32M
@3811Mhz - 9x423FSB 1:1


2T = 16m 01.093s
1T = 15m 20.203s
1T tweaked 4-24-6-8 = 15m 12.359s
1T tweaked 4-15-6-8 tRFC 35 = 14m 19.844s

2T half way mark


2T completed


1T completed
- extra vdimm wasn't necessary just forgot to set it back to 2.1v


Notice the advance timings when set to AUTO result in different settings for 2T vs 1T:

2T Bios set = Memset reported
tRRD: AUTO = 4
tRC: AUTO = 24
tWR: AUTO = 6
tWTR: AUTO = 8

vs

1T Bios set = Memset reported
tRRD: AUTO = 5 - loosened from 4 in 2T
tRC: AUTO = 25 - loosened from 24 in 2T
tWR: AUTO = 6
tWTR: AUTO = 9 - loosened from 8 in 2T

tRFC which isn't available as an option in bios for both 1T and 2T were set AUTO = 45.

Note, that with 2x512MB modules it will be ~20 seconds slower than same clocked/timing 2x1GB modules for Super Pi 32M.

So what happens when I set it to 1T and use memset to tighten the advance timings so the match the 2T AUTO set timings of 4-24-6-8 ?

1T tweaked half way


1T tweaked completed


Now to use memset to tighten the advance timings to 4-15-6-8 with tRFC = 35 and look at this nearly 102 seconds faster than stock 2T @423mhz 4-4-4-12 at AUTO advance timings! But still slower than 975x chipset at same clocks by around 25-40 seconds. If you take into account ~20s slower with 2x512MB modules, means the Striker was around 5-20 seconds slower than 975x in Super Pi 32M at those particular timings below.

1T memset tweaked


Now tightened tRAS to 4 and advance timings to 4-15-5-7 and tRFC to 32.

1T @423Mhz 4-4-4-4 4-15-5-7


Interesting for the 2x512MB OCZ PC5400 Platinum XTC 3-3-3-12 memory, if advance timings are set to AUTO values in bios, tRAS of 12 is fastest followed by tRAS for 4 and 1. Other tRAS values below 12 or above are slower, until you manually set the advance timings, only then will you see improvement in tRAS below 12.

Last edited by eva2000; 20-12-2006 at 11:56 PM.
eva2000 is offline   Reply With Quote
Old 20-12-2006, 03:49 PM   #6
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,018
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Re: Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results

Flashed to 402 bios
2x1GB G.Skill F2-6400PHU2 2GBHZ
Micron D9GMH 629 batch

Since everyone seems to be doing well with 402 beta bios, I decided to flash from 302 to 402 bios using Asus Ez Flash from within bios using my 4GB Corsair Voyager USB flash drive. Worked flawlessly. 402 bios brought 2 new bios memory timing options to the table, tRD and tRFC. tRFC can be seen in Memset v3.1v beta but tRD can't be seen in Memset right now.

Now using 2x1GB G.Skill F2-6400PHU2 2GBHZ Micron D9GMH 629 batch memory, I decided to play with the unlinked memory option finding my footing using the 5:9 divider below at stock 266FSB and 480mhz clocked memory

As always I used memtest86+ v1.65 to test the clocks before even booting into windows. I'd highly advise doing this all the time

AI Tuning: Manual
Nvidia GPU Ex: Disabled
LinkBoost: Disabled

FSB - Memory Clock Mode: Unlinked
x FSB - Memory Raio: N/A
FSB (QDR), Mhz: 1067 (means 266FSB)
Actual FSB (QDR), Mhz: 1067
MEM (DDR), Mhz: 960 (means 480mhz speeds)
Actual MEM (DDR), Mhz: 960

Vcore Voltage: 1.4v
Memory Voltage: 2.200v
1.2v HT Voltage: 1.4v
NB Core Voltage: 1.4v
SB Core Voltage: 1.50v
CPU VTT Voltage: 1.4v
DDRII Controller Ref Voltage: AUTO
DDRII Channel A Ref Voltage: AUTO
DDRII Channel B Ref Voltage: AUTO

tCL (CAS Latency): 4
tRCD: 4 then 3
tRP: 4 then 3
tRAS: 12 then 4
Command Per Clock (CMD): 2 clock (2T)

Advance Memory Settings
tRRD: AUTO
tRC: AUTO
tWR: AUTO
tWTR: AUTO
tREF: AUTO
tRD: AUTO < new in 402 bios
tRFC: AUTO < new in 402 bios
Async Latency: AUTO

CPU Spread Spectrum: Disabled
PCIE Spread Spectrum: Disabled
MCP PCIE Spread Spectrum: Disabled
SATA Spread Spectrum: Disabled
LDT Spread Spectrum: Disabled

CPU Internal Thermal Control: Disabled
Limit CPUID MaxVal: Disabled
Enhanced C1 (C1E): Disabled
Execute Disable Bit: Disabled
Virtualization Technology: Disabled
Enhanced Intel Speedstep Tech: Disabled
LDT Frequency: 5x

PCIEX16_1 Frequency (Mhz): 100
PCIEX16_2 Frequency (Mhz): 100
PCIEX16_3 Frequency (Mhz): 100
SPP<->MCP Ref Clock, Mhz: 200
E6600@2400Mhz 9x266FSB 480Mhz 4-4-4-12 2T at 2.2v


Advance Memory Settings
The advance memory settings set as AUTO in bios @5:9 unlinked 480mhz memory clocks showed up in Memset v3.1 beta in windows as:



tRRD: AUTO = 3
tRC: AUTO = 26
tWR: AUTO = 6
tWTR: AUTO = 8
tREF: AUTO = ??
tRD: AUTO = ??
tRFC: AUTO = 37
Async Latency: ??

I then ran Super Pi 32M but tightened the timings using Memset v3.1 beta to the following:

tRRD: 3
tRC: 26 tightened to 7
tWR: 6 tightened to 5
tWTR: 8 tightened to 7
tRFC: 37 tightened to 18

Managed 32M time of 20m 24.640s at E6600@2400Mhz 9x266FSB 480Mhz 4-4-4-12 2T at 2.2v



E6600@2400Mhz 9x266FSB 480Mhz 4-3-3-4 2T at 2.2v


Then I tightened both the main and advance timings even more in same windows session and ran Super Pi 32M with the following:

tCL (CAS Latency): 4
tRCD: 4 tightened to 3
tRP: 4 tightened to 3
tRAS: 12 tightened to 4
Command Per Clock (CMD): 2 clock (2T)

tRRD: 3
tRC: 7 tightened to 5
tWR: 5
tWTR: 7
tRFC: 18

Managed 32M time of 20m 16.891s at E6600@2400Mhz 9x266FSB 480Mhz 4-3-3-4 2T at 2.2v





Still have alot of tweaking to do. The unlinked memory mode isn't as flexible as I thought it would be - basically unlinked mode gives you more sets of fixed dividers than the linked mode, so you only have set options for each FSB you set. An example just to name a few memory options you have in unlinked mode for 266FSB (1067 QDR) are:

Memory clocks in unlinked mode:
  • 915 actual 914.3Mhz
  • 934 actual 933.4Mhz
  • 949 actual 948.1Mhz
  • 960 actual 960Mhz
  • 978 actual 977.8Mhz
  • 991 actual 990.5Mhz
  • 1000 actual 1000Mhz
  • 1067 actual 1066.7Mhz
Those are just a few of the examples there's many more.
  • So if you typed in 960, you'll get 960 = 480Mhz memory clocks.
  • If you typed in 970, you'll still get 960 = 480Mhz memory clocks.
  • If you typed in 985, you'll get 977.8Mhz = 488.9Mhz memory clocks.

Last edited by eva2000; 23-12-2006 at 03:34 AM.
eva2000 is offline   Reply With Quote
Old 20-12-2006, 03:49 PM   #7
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,018
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Re: Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results

Advance Memory Timings - What does AUTO option assign ?

More memory tests... This time to investigate what the bios sets the advance memory timings to when set to AUTO for each memory clock frequency. Again ALWAYS use memtest86+ v1.65 to test settings before booting windows.

Quick Memtest86+ v1.65 Procedure

I use a quick test first consisting of:
  • 2 full memtest86+ v1.65 loops of test #1 to #8
  • Then continuous test #5 loops for at least 10 loops.
  • If there's errors in test #1-4 then try raising voltages i.e. NB, CPU VTT or 1.2v HT volts
  • If test #5 errors try raising vdimm as well as NB, CPU VTT or 1.2v HT volts.
  • Settings and clocks must also be able to reboot properly on exiting bios and memtest testing. Since unstable clocks or lack of NB and/or vDimm volts can result in problems rebooting on bios/memtest testing exit.
That isn't to say it's stable but at least stable enough to be able to boot into windows without BSOD'ing

System:
  • Intel Core 2 Duo E6600 QPGE B1 Step 5
  • Scythe Infinity heatsink with mounting mod + Thermaltake 120x25mm 81cfm fan
  • 128MB FX5200 PCI / 256MB BFG 7800GTX / Gainward 8800GTS/SLI
  • Asus 680i Striker Extreme rev1.0 0701 beta bios
  • 1x 160GB Seagate 7200.7 SATA NCQ on SATA2 port
  • Pioneer DVR-05 DVD Burner
  • 600W OCZ EvoStream Modular Power supply

9x266FSB 4-4-4-12 2T fixed

I'll be only changing the unlinked mode divider keeping FSB fixed at 266FSB and timings fixed at 4-4-4-12 2T for initial tests. These screenshots below are just to find out what AUTO responds to for advance timings in memset v3.1 beta in windows.

Asus PC Probe2 reported PSU rail voltages:
Asus PC Probe2 reported voltages vs Digitech QM1538 Digital Multimeter measurements while under load @2400mhz at 1.3875v bios set (1.28v vcore idle / 1.26v vcore load).

+3.3v rail = 3.26v vs 3.387v real
+5v rail = 4.78-4.89v vs 5.02v real
+12v rail = 11.78v vs 12.23v real

Bios template for 680i Striker users to report their settings:
AI Tuning: Manual
Nvidia GPU Ex: Disabled
LinkBoost: Disabled

FSB - Memory Clock Mode: Unlinked
x FSB - Memory Raio: N/A
FSB (QDR), Mhz: 1066 (means 266FSB)
Actual FSB (QDR), Mhz: 1066.7
MEM (DDR), Mhz:
Actual MEM (DDR), Mhz:

Vcore Voltage: 1.325
Memory Voltage: 2.1 to 2.5v
1.2v HT Voltage: 1.4
NB Core Voltage: 1.4 to 1.6v
SB Core Voltage: 1.5
CPU VTT Voltage: 1.4
DDRII Controller Ref Voltage: AUTO
DDRII Channel A Ref Voltage: AUTO
DDRII Channel B Ref Voltage: AUTO

tCL (CAS Latency): 4
tRCD: 4
tRP: 4
tRAS: 12
Command Per Clock (CMD): 2 clock (2T)

Advance Memory Settings
tRRD: AUTO
tRC: AUTO
tWR: AUTO
tWTR: AUTO
tREF: AUTO
Async Latency: AUTO
4:5 - 333Mhz



2:3 - 400Mhz



8:15 - 500Mhz

Raised vdimm to 2.25v and NB to 1.5v


1:2 - 533Mhz

Raised vdimm to 2.4v and NB to 1.6v


9x266FSB 5-4-4-12 2T fixed


Changed only to CAS = 5. Findings:
  • CAS5 automatically loosens TWTR from 9 to 10 it seems when TWTR is set to AUTO.

1:2 - 533Mhz

Raised vdimm to 2.4v and NB to 1.6v


7:15 - 571Mhz

Raised vdimm to 2.45v and NB to 1.6v


6:13 - 577Mhz

Raised vdimm to 2.45v and NB to 1.65v


5:11- 586Mhz

Raised vdimm to 2.55v and NB to 1.65v


4:9- 600Mhz

Raised vdimm to 2.575v and NB to 1.65v




Raising NB from 1.65v to 1.7v allowed me to pass single 32M run @600Mhz 5-4-4-12 at 2T at 2.575v vdimm. Memset seems to have not reported the right tRC value compared to what CPUZ is reporting tRC = 33.

Last edited by eva2000; 29-12-2006 at 07:05 PM.
eva2000 is offline   Reply With Quote
Old 20-12-2006, 03:49 PM   #8
eva2000
Administrator
eva2000's PC Specs
 
Join Date: Jul 22 2004
Location: Brisbane, Australia
Posts: 23,018
twitter.com/i4memorycom Facebook Page livestream.com/i4memorycom
Blog Entries: 42
Re: Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results

The REAL reason for choosing an Asus 680i Striker Extreme SLI motherboard is of course for SLI

2x 640MB Gainward 8800GTS SLI twins with 97.44 drivers







System:
  • Intel Core 2 Duo E6600 QPGE B1 Step 5
  • Scythe Infinity heatsink with mounting mod + Thermaltake 120x25mm 81cfm fan
  • 2x 640MB Gainward 8800GTS SLI
  • Asus 680i Striker Extreme rev1.0 0402 bios
  • 1x 160GB Seagate 7200.7 SATA NCQ on SATA2 port
  • Pioneer DVR-05 DVD Burner
  • 600W OCZ EvoStream Modular Power supply

3dmark06 stock @513/1584



3dmark06 oc'd @648/2052



Yup cpu limited for 3dmark06 with E6600@2400Mhz stock

Asus PC Probe2 reported PSU rail voltages:
Asus PC Probe2 reported voltages vs Digitech QM1538 Digital Multimeter measurements while under load @2400mhz at 1.3875v bios set (1.28v vcore idle / 1.26v vcore load).

+3.3v rail = 3.26v vs 3.387v real
+5v rail = 4.78-4.89v vs 5.02v real
+12v rail = 11.78v vs 12.23v real

Now let's check out nVidia's LinkBoost technology which is available on 680i motherboards when couplied with 79xx/88xx nVidia video cards. Straight from nVidia.com web site pdf tech documents on LinkBoost:
NVIDIA LinkBoost technology is a technological advancement that ensures blazing graphics bus speeds, bolstering overall system performance. It features hardware and software innovations within select NVIDIA GeForce graphics processing units and media and communications processor (MCP) products. NVIDIA nForce 680i SLI MCP and NVIDIA nForce 590 SLI AMD MCP are the only core logic platforms that support LinkBoost.

Most PC platforms are designed to provide maximum performance under industry standard specifications. Speed of interlink connections between the various PC components is an influential aspect of performance. Depending on the application, PCI Express, as well as the interchip connection between MCP chips, can significantly affect performance.


LinkBoost, a feature of the NVIDIA nForce 680i SLI and NVIDIA nForce 590 SLI AMD MCPs, becomes active once the system detects a GPU capable of supporting LinkBoost technology. NVIDIA GeForce 8800 and 7900 GTX are the only GPU add-in cards that support LinkBoost technology. When the supported GPU graphics cards are present, the system automatically recognizes their presence and dynamically boosts PCI Express and MCP HyperTransport bus speeds by 25 percent.


LinkBoost does not require an NVIDIA SLI–enabled system configuration. The increased PCI Express and MCP HyperTransport bandwidth makes higher data transfer rates across the links possible, allowing more data packets to be transmitted at any time.
Gainward Bliss 8800GTS SLI @621/2052

3dmark06
Linkboost disabled in bios = 15,331
Linkboost enabled in bios = 15,361

Linkboost disabled



Linkboost enabled




SM2.0 Score = dropped 0.0688%
HDR/SM3.0 Score = dropped 0.412%
CPU Score = increased 1.28%

Looks like Linkboost didn't do much for the actual GPU component tests ? Might need to test a full range of graphic benchmarks etc to find out.

Notice SPP/MCP PCI-E bandwidth went from 2500Mhz to 3125Mhz exactly 25% increase

LinkBoost made 3dmark benchmarks a bit unstable it seems so disabled it in bios again.

Gainward Bliss 8800GTS SLI @621/2052

3dmark05


3dmark2003


There's some improvement for 3dmark06 for higher mem clocks 400mhz 4-4-4-12 2T auto vs 533mhz 4-4-4-12 2T auto - 533mhz gained 15490 - 15361 = 129pts @621/2052 clocked 8800GTS SLI









Something weird i noticed with linkboost disabled at 1.55v NB volts, ntune shows MCP/SPP PCI-E at 2500mhz.. but raise NB volts to 1.6v with linkboost still disabled, ntune shows MCP PCI-E at 3125mhz ... MCP PCI-E auto scales with NB volts ?

Looks like 426Mhz 4-4-4-12 1T is faster than 533mhz 4-4-4-12 2T even for 3dmark06

426mhz 4-4-4-12 1T = 15,506
533mhz 4-4-4-12 2T = 15,490


Last edited by eva2000; 31-12-2006 at 06:08 PM.
eva2000 is offline   Reply With Quote
Reply

Bookmarks

Tags
680i, asus, core, duo, e6600, extreme, intel, motherboard, results, sli, striker


Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)
 
Thread Tools


Similar Threads for: Intel Core 2 Duo E6600 + Asus 680i Striker Extreme SLI motherboard Results
Thread Thread Starter Forum Replies Last Post
[hwbot] Super Pi v1.50 - 1M rankings pt2 hwbot HWBOT Official Rankings 25 22-12-2009 07:11 AM
[hwbot] Super Pi v1.50 32M rankings - i4memory.com hwbot HWBOT Official Rankings 437 14-04-2008 12:34 PM
Intel Core 2 Duo E6700 + Asus 680i Striker Extreme SLI eva2000 Asus Intel motherboards / CPU 13 21-02-2007 02:35 AM
Asus 680i Striker Extreme SLI + Sound issues [LinkBoost/GPU EX] eva2000 Asus Intel motherboards / CPU 0 03-01-2007 01:07 PM
Intel Core 2 Duo E6600 Conroe + Intel D975XBX Installed eva2000 General Intel Motherboards / CPU 128 23-07-2006 02:56 AM


All times are GMT +11. The time now is 12:05 AM.

no new posts