Since I have a stable build I decided yesterday just to see what this 6600
could do with just default voltages on CPU and ram. Ram stayed set on
'Linked' and Auto and SLI disabled. I eased the FSB up from 1066 (default)
and finally up to 1333.3 for 3.0ghz easily. I was a bit confused with some
of the nomenclature of this bios as it doesn't seem to be the same as Phil's
on his 680i board. Mine has "FSB and Memory Settings" listed and the FSB
settings are listed as the actual ratings. I still may be doing this wrong,
but it does accomplish the OC. I ran Orthos for 16-18hrs overnight and all
seems well. I forget who reccomended it, but it sure will push the two cores
more than any other program I have seen. It actually pushed both cores up
into the 50c range which even video rendering could not get past 45c or so
after several hours. I didn't do any memory tweaking as I was just
interested in what I could get the CPU to for right now. To be perfectly
honest, I am a bit confused with all the memory settings except for the
timings and dividers. 5-4 and 3-2 will not boot to XP with the default
timings so I will have to do some experimenting there after I run awhile at
these settings just to be sure of good stability. I think that relaxing the
default timings might help a lot. Default is 3-3-3-4 IIRC so I have some
room to work with. The Overclocking section of the Asus P5N32-E SLI manual
is not very helpful when it comes to actually doing anything. The Asus forum
is not much better. Oh well, I will find what I need to know somewhere after
a bit of digging. That is where everything is now.
'Ed Medlin' wrote, in part:
| Since I have a stable build I decided yesterday just to see what this 6600
| could do with just default voltages on CPU and ram. Ram stayed set on
| 'Linked' and Auto and SLI disabled. I eased the FSB up from 1066 (default)
| and finally up to 1333.3 for 3.0ghz easily. I was a bit confused with some
| of the nomenclature of this bios as it doesn't seem to be the same as
Phil's
| on his 680i board. Mine has "FSB and Memory Settings" listed and the FSB
| settings are listed as the actual ratings. I still may be doing this
wrong,
| but it does accomplish the OC.
_____
It seems you have a good solid start on your system.
It is too bad that a translator is needed to compare BIOS settings even for
motherboards with the same chipset but different manufacturers.
So here is my contribution, using nTune reports for the actual values.
(Caveat - nTune is still a bit squirrelly; as an example sometimes I can get
nTune to report the SPD information for memory modules, sometimes not.)
Here below are the nTune reports and SiSoft Sandra Lite ver. 2007.4.11.22
Benchmarks for my E4300 / EVGA 680i system, CPU speed 2.7 GHz for two
different Memory Bus speeds, 600 MHz and 1200 MHz.
***** 1200 MHz FSB / 600 MHz Memory Bus nTune Information
System Clocks:
Reference clock (FSB): 300 MHz
CPU multiplier: 9 X
CPU core frequency: 2,700 MHz
FSB Frequency: 1,200 (QDR) MHz
Memory bus frequency: 600 (DDR2) MHz
SPP Reference clock: 2,500 MHz
MCP Reference clock: 2,500 MHz
System Voltages:
CPU: 1.250 v
Memory: 2.300 v
FSB: 1.400 v
HT nForce SPP <-> MCP: 1.200 v
nForce SPP: 1.400 v
nForce MCP: 1.500 v
Auxililary: 1.500 v
Memory Timings:
Row Address Strobe (tRAS): 5
Write Recovery Time (tWR): 4
RAS to CAS access (tRCD): 3
RAS to CAS Delay (tRRD): 3
Memory bank switch (tRP): 3
R to W Turnaround (tRWT): 5
R to R Timing (tRDRD): 3
Row Cycle Time (tRC): 17
W to W Timing (tWRWR): 2
CAS Latency (tCL): 3
Command Per Clock (tCPC): 0
Row Cycle Time (tRC): 17
~~~~~SiSoft Sandra Lite ver. 2007.4.11.22 Benchmarks with above nTune
reported values
***** 1200 MHz FSB / 1200 MHz Memory Bus nTune Information
System Clocks:
Reference clock (FSB): 300 MHz
CPU multiplier: 9 X
CPU core frequency: 2,700 MHz
FSB Frequency: 1,200 (QDR) MHz
Memory bus frequency: 1200 (DDR2) MHz
SPP Reference clock: 2,500 MHz
MCP Reference clock: 2,500 MHz
System Voltages:
CPU: 1.250 v
Memory: 2.300 v
FSB: 1.400 v
HT nForce SPP <-> MCP: 1.200 v
nForce SPP: 1.450 v
nForce MCP: 1.500 v
Auxililary: 1.500 v
Memory Timings:
Row Address Strobe (tRAS): 10
Write Recovery Time (tWR): 6
RAS to CAS access (tRCD): 6
RAS to CAS Delay (tRRD): 5
Memory bank switch (tRP): 6
R to W Turnaround (tRWT): 6
R to R Timing (tRDRD): 3
Row Cycle Time (tRC): 33
W to R Command Delay (tWTR): 11
W to W Timing (tWRWR): 2
CAS Latency (tCL) 5
Command Per CLock (tCPC): 0
~~~~~SiSoft Sandra Lite ver. 2007.4.11.22 Benchmarks with above nTune
reported values
Hope this helps somebody, ghod knows I don't understand it all yet B^(
Phil Weldon
"Ed Medlin" <edmedlin@yahoo.com> wrote in message
news:kXF_h.16593$YL5.9354@newssvr29.news.prodigy.n et...
| Since I have a stable build I decided yesterday just to see what this 6600
| could do with just default voltages on CPU and ram. Ram stayed set on
| 'Linked' and Auto and SLI disabled. I eased the FSB up from 1066 (default)
| and finally up to 1333.3 for 3.0ghz easily. I was a bit confused with some
| of the nomenclature of this bios as it doesn't seem to be the same as
Phil's
| on his 680i board. Mine has "FSB and Memory Settings" listed and the FSB
| settings are listed as the actual ratings. I still may be doing this
wrong,
| but it does accomplish the OC. I ran Orthos for 16-18hrs overnight and all
| seems well. I forget who reccomended it, but it sure will push the two
cores
| more than any other program I have seen. It actually pushed both cores up
| into the 50c range which even video rendering could not get past 45c or so
| after several hours. I didn't do any memory tweaking as I was just
| interested in what I could get the CPU to for right now. To be perfectly
| honest, I am a bit confused with all the memory settings except for the
| timings and dividers. 5-4 and 3-2 will not boot to XP with the default
| timings so I will have to do some experimenting there after I run awhile
at
| these settings just to be sure of good stability. I think that relaxing
the
| default timings might help a lot. Default is 3-3-3-4 IIRC so I have some
| room to work with. The Overclocking section of the Asus P5N32-E SLI manual
| is not very helpful when it comes to actually doing anything. The Asus
forum
| is not much better. Oh well, I will find what I need to know somewhere
after
| a bit of digging. That is where everything is now.
|
| Ed
|
|
Ed Medlin wrote:
> Since I have a stable build I decided yesterday just to see what this
> 6600 could do with just default voltages on CPU and ram. Ram stayed
> set on 'Linked' and Auto and SLI disabled. I eased the FSB up from
> 1066 (default) and finally up to 1333.3 for 3.0ghz easily. I was a
> bit confused with some of the nomenclature of this bios as it doesn't
> seem to be the same as Phil's on his 680i board. Mine has "FSB and
> Memory Settings" listed and the FSB settings are listed as the actual
> ratings. I still may be doing this wrong, but it does accomplish the
> OC. I ran Orthos for 16-18hrs overnight and all seems well. I forget
> who reccomended it, but it sure will push the two cores more than any
> other program I have seen. It actually pushed both cores up into the
> 50c range which even video rendering could not get past 45c or so
> after several hours. I didn't do any memory tweaking as I was just
> interested in what I could get the CPU to for right now. To be
> perfectly honest, I am a bit confused with all the memory settings
> except for the timings and dividers. 5-4 and 3-2 will not boot to XP
> with the default timings so I will have to do some experimenting
> there after I run awhile at these settings just to be sure of good
> stability. I think that relaxing the default timings might help a
> lot. Default is 3-3-3-4 IIRC so I have some room to work with. The
> Overclocking section of the Asus P5N32-E SLI manual is not very
> helpful when it comes to actually doing anything. The Asus forum is
> not much better. Oh well, I will find what I need to know somewhere
> after a bit of digging. That is where everything is now.
"Thomas" <ThomasH@lycosmail.nl> wrote in message
news:463c350b$0$16936$9a622dc7@news.kpnplanet.nl.. .
> Ed Medlin wrote:
>> Since I have a stable build I decided yesterday just to see what this
>> 6600 could do with just default voltages on CPU and ram. Ram stayed
>> set on 'Linked' and Auto and SLI disabled. I eased the FSB up from
>> 1066 (default) and finally up to 1333.3 for 3.0ghz easily. I was a
>> bit confused with some of the nomenclature of this bios as it doesn't
>> seem to be the same as Phil's on his 680i board. Mine has "FSB and
>> Memory Settings" listed and the FSB settings are listed as the actual
>> ratings. I still may be doing this wrong, but it does accomplish the
>> OC. I ran Orthos for 16-18hrs overnight and all seems well. I forget
>> who reccomended it, but it sure will push the two cores more than any
>> other program I have seen. It actually pushed both cores up into the
>> 50c range which even video rendering could not get past 45c or so
>> after several hours. I didn't do any memory tweaking as I was just
>> interested in what I could get the CPU to for right now. To be
>> perfectly honest, I am a bit confused with all the memory settings
>> except for the timings and dividers. 5-4 and 3-2 will not boot to XP
>> with the default timings so I will have to do some experimenting
>> there after I run awhile at these settings just to be sure of good
>> stability. I think that relaxing the default timings might help a
>> lot. Default is 3-3-3-4 IIRC so I have some room to work with. The
>> Overclocking section of the Asus P5N32-E SLI manual is not very
>> helpful when it comes to actually doing anything. The Asus forum is
>> not much better. Oh well, I will find what I need to know somewhere
>> after a bit of digging. That is where everything is now.
>
> Ed, did you read up on this document yet? I downloaded it from a link
> supplied in this newsgroup. I cannot find it anymore, but here's the link
> to my FTP:
> http://www.eastwestjewels.nl/temp/nf...erclocking.pdf
>
> --
> Met vriendelijke groeten, Thomas van der Horst
Thanks Thomas. It looks like a good read and will help. There is a lot more
there than in my Asus manual.
'Phil Weldon' wrote, in part:
So here is my contribution, using nTune reports for the actual values.
| (Caveat - nTune is still a bit squirrelly; as an example sometimes I can
get
| nTune to report the SPD information for memory modules, sometimes not.)
_____
The solution (at least for my EVGA 680i motherboard) for reading the SPD
information is to use 'nVidia Control Panel' and select 'View System
Information' (this can also be done by starting 'nTune'.
THEN
Double-click on one of the processors in 'Processor information'; this will
display the processor information.
THEN
Double-click on one of the DIMMs in 'Memory information'; this will display
the SPD information, including any extended SPD information (SLI-ready
information.)
After double-clicking on one of the processors in 'Processor information' in
'View System Information' clicking on the 'View SPD' button in 'Adjust
Motherboard Settings', then double-clicking on a DIMM in the 'Memory
Information' window WILL display the SPD information, including any extended
SPD information (SLI-ready information.)
If I do not double-click on a processor in 'View System Information' FIRST,
double-clicking on a DIMM in 'Memory information' does nothing AND
double-clicking on a DIMM in 'Adjust Motherboard Settings' in the 'Memory
Information' window [brought up by the 'View SPD information'] does nothing.
Using the 'Create a tech support log file:' 'Save' button ALWAYS works and
creates a text log file that includes the SPD and extended SPD information.
The text log file SPD and extended SPD (NVIDIA Enhanced Performance Profile
SPD information)
for my Patriot PDC22G8500ELK DDR2-667/PC8500 1 GByte modules (which, by the
way, have gone from the ~ $200 US for a pair I paid in late March to ~ $270
US in early May) contain the following information
~~~~~~~~~~~~~~
Memory Dimm 0 (0x50)
Memory type: DDR2 SDRAM
Error correction: NONE
Voltage interface level: SSTL (1.8V)
Data width: 64 bits
Number of physical banks: 2 (double sided)
Number of logical banks: 4
Number of row addresses: 14 14
Number of col addresses: 10 10
Bank density: 512 512 MB
Primary RAM Width: 8 8 bits
Error checking RAM Width: 0 0 bits
Module density: 1024 MB
Refresh rate: 7812.500 us
Self-refresh: Yes
Burst lengths supported: 4 8
CAS 3.0:
Cycle time (tCYC) = 5.000 ns fMax = 200.000 Mhz
Access time (tAC) = 0.600 ns fMax = 1666.667 MHz
CAS 4.0:
Cycle time (tCYC) = 4.300 ns fMax = 232.558 Mhz
Access time (tAC) = 0.500 ns fMax = 2000.000 MHz
CAS 5.0:
Cycle time (tCYC) = 2.500 ns fMax = 400.000 Mhz
Access time (tAC) = 0.400 ns fMax = 2500.000 MHz
Not reportedNot reportedMin delay, Bk/Bk random cols: 0 clocks
--- NVIDIA Enhanced Performance Profile SPD information ---
SPD version: B1
Profile for Optimal Performance: 01
Profile 0 invalid
Voltage Level: 4.9V
Address Command Rate: 2T
Address Drive Strength: 2.0x
Chip Select Drive Strength: 2.0x
Clock Drive Strength: 1.5x
Data Drive Strength: 1.5x
DQS Drive Strength: 1.5x
Address Fine Delay: 31/64 MEMCLK delay
Address Setup Time: 1 MEMCLK
Chip Select Fine Delay: 31/64 MEMCLK delay
Chip Select Setup Time: 1 MEMCLK
Profile 1 valid
Voltage Level: 2.3V
Address Command Rate: 2T
Address Drive Strength: 1.0x
Chip Select Drive Strength: 1.0x
Clock Drive Strength: 1.0x
Data Drive Strength: 1.0x
DQS Drive Strength: 1.0x
Address Fine Delay: 1.0
Address Setup Time: 1 MEMCLK
Chip Select Fine Delay: No delay
Chip Select Setup Time: 1 MEMCLK
Sorry to be so wordy, but I wanted to be precise. Evidently the user
interface for nVidia Control Panel and nTune were not thoroughly trested; I
can't see any reason that the motherboard implementation could cause this
problem.
"Phil Weldon" <not.disclosed@example.com> wrote in message
news:YrV_h.7081$Ut6.4714@newsread1.news.pas.earthl ink.net...
| 'Ed Medlin' wrote, in part:
|| Since I have a stable build I decided yesterday just to see what this
6600
|| could do with just default voltages on CPU and ram. Ram stayed set on
|| 'Linked' and Auto and SLI disabled. I eased the FSB up from 1066
(default)
|| and finally up to 1333.3 for 3.0ghz easily. I was a bit confused with
some
|| of the nomenclature of this bios as it doesn't seem to be the same as
| Phil's
|| on his 680i board. Mine has "FSB and Memory Settings" listed and the FSB
|| settings are listed as the actual ratings. I still may be doing this
| wrong,
|| but it does accomplish the OC.
| _____
|
| It seems you have a good solid start on your system.
| It is too bad that a translator is needed to compare BIOS settings even
for
| motherboards with the same chipset but different manufacturers.
|
| So here is my contribution, using nTune reports for the actual values.
| (Caveat - nTune is still a bit squirrelly; as an example sometimes I can
get
| nTune to report the SPD information for memory modules, sometimes not.)
|
| Here below are the nTune reports and SiSoft Sandra Lite ver. 2007.4.11.22
| Benchmarks for my E4300 / EVGA 680i system, CPU speed 2.7 GHz for two
| different Memory Bus speeds, 600 MHz and 1200 MHz.
|
|
| ***** 1200 MHz FSB / 600 MHz Memory Bus nTune Information
| System Clocks:
| Reference clock (FSB): 300 MHz
| CPU multiplier: 9 X
| CPU core frequency: 2,700 MHz
| FSB Frequency: 1,200 (QDR) MHz
| Memory bus frequency: 600 (DDR2) MHz
| SPP Reference clock: 2,500 MHz
| MCP Reference clock: 2,500 MHz
|
| System Voltages:
| CPU: 1.250 v
| Memory: 2.300 v
| FSB: 1.400 v
| HT nForce SPP <-> MCP: 1.200 v
| nForce SPP: 1.400 v
| nForce MCP: 1.500 v
| Auxililary: 1.500 v
|
| Memory Timings:
| Row Address Strobe (tRAS): 5
| Write Recovery Time (tWR): 4
| RAS to CAS access (tRCD): 3
| RAS to CAS Delay (tRRD): 3
| Memory bank switch (tRP): 3
| R to W Turnaround (tRWT): 5
| R to R Timing (tRDRD): 3
| Row Cycle Time (tRC): 17
| W to W Timing (tWRWR): 2
| CAS Latency (tCL): 3
| Command Per Clock (tCPC): 0
| Row Cycle Time (tRC): 17
|
| ~~~~~SiSoft Sandra Lite ver. 2007.4.11.22 Benchmarks with above nTune
| reported values
|
| Processor Arithmetic:
| Dhrystone ALU: 24787 MIPS
| Whetstone ISSE3: 16853 MFLOPS
| Memory Bandwidth:
| Int Buffered iSSE2: 6051
| Float Buffered iSSE2: 6082
| Bandwidth Efficiency: 63% estimated
| Maximum Bus Bandwidth: 9600 MB/s (estimated)
| Cache and Memory:
| Combined Index: 17082 MB/s
| Speed Factor: 51.4
| Processor Multi-Media:
| Integer x8 iS-SSE3: 148529 it/s
| Float x4 iSSE2: 80918 it/s
| Memory Latency:
| Random 16 MByte: 71.8 ns (193.8 clocks)
| Linear 16 MByte: 11.1 ns (30.0 clocks)
|
| ***** 1200 MHz FSB / 1200 MHz Memory Bus nTune Information
| System Clocks:
| Reference clock (FSB): 300 MHz
| CPU multiplier: 9 X
| CPU core frequency: 2,700 MHz
| FSB Frequency: 1,200 (QDR) MHz
| Memory bus frequency: 1200 (DDR2) MHz
| SPP Reference clock: 2,500 MHz
| MCP Reference clock: 2,500 MHz
|
| System Voltages:
| CPU: 1.250 v
| Memory: 2.300 v
| FSB: 1.400 v
| HT nForce SPP <-> MCP: 1.200 v
| nForce SPP: 1.450 v
| nForce MCP: 1.500 v
| Auxililary: 1.500 v
|
| Memory Timings:
| Row Address Strobe (tRAS): 10
| Write Recovery Time (tWR): 6
| RAS to CAS access (tRCD): 6
| RAS to CAS Delay (tRRD): 5
| Memory bank switch (tRP): 6
| R to W Turnaround (tRWT): 6
| R to R Timing (tRDRD): 3
| Row Cycle Time (tRC): 33
| W to R Command Delay (tWTR): 11
| W to W Timing (tWRWR): 2
| CAS Latency (tCL) 5
| Command Per CLock (tCPC): 0
|
| ~~~~~SiSoft Sandra Lite ver. 2007.4.11.22 Benchmarks with above nTune
| reported values
|
| Processor Arithmetic:
| Dhrystone ALU: 24789 MIPS
| Whetstone ISSE3: 16866 MFLOPS
| Memory Bandwidth:
| Int Buffered iSSE2: 6318
| Float Buffered iSSE2: 6320
| Bandwidth Efficiency: 66% (estimated)
| Maximum Bus Bandwidth: 9600 MB/s (estimated)
| Cache and Memory:
| Combined Index: 17815 MB/s
| Speed Factor: 46.3 (estimated)
| Processor Multi-Media:
| Integer x8 iS-SSE3: 148471 it/s
| Float x4 iSSE2: 80910 it/s
| Memory Latency:
| Random 16 MByte: 63.7 ns (171.9 clocks)
| Linear 16 MByte: 9.8 ns (26.5 clocks)
|
| Hope this helps somebody, ghod knows I don't understand it all yet B^(
|
| Phil Weldon
|
|
|
| "Ed Medlin" <edmedlin@yahoo.com> wrote in message
| news:kXF_h.16593$YL5.9354@newssvr29.news.prodigy.n et...
|| Since I have a stable build I decided yesterday just to see what this
6600
|| could do with just default voltages on CPU and ram. Ram stayed set on
|| 'Linked' and Auto and SLI disabled. I eased the FSB up from 1066
(default)
|| and finally up to 1333.3 for 3.0ghz easily. I was a bit confused with
some
|| of the nomenclature of this bios as it doesn't seem to be the same as
| Phil's
|| on his 680i board. Mine has "FSB and Memory Settings" listed and the FSB
|| settings are listed as the actual ratings. I still may be doing this
| wrong,
|| but it does accomplish the OC. I ran Orthos for 16-18hrs overnight and
all
|| seems well. I forget who reccomended it, but it sure will push the two
| cores
|| more than any other program I have seen. It actually pushed both cores up
|| into the 50c range which even video rendering could not get past 45c or
so
|| after several hours. I didn't do any memory tweaking as I was just
|| interested in what I could get the CPU to for right now. To be perfectly
|| honest, I am a bit confused with all the memory settings except for the
|| timings and dividers. 5-4 and 3-2 will not boot to XP with the default
|| timings so I will have to do some experimenting there after I run awhile
| at
|| these settings just to be sure of good stability. I think that relaxing
| the
|| default timings might help a lot. Default is 3-3-3-4 IIRC so I have some
|| room to work with. The Overclocking section of the Asus P5N32-E SLI
manual
|| is not very helpful when it comes to actually doing anything. The Asus
| forum
|| is not much better. Oh well, I will find what I need to know somewhere
| after
|| a bit of digging. That is where everything is now.
||
|| Ed
||
||
|
|
"Phil Weldon" <not.disclosed@example.com> wrote in message
news:8B6%h.8332$j63.4198@newsread2.news.pas.earthl ink.net...
> 'Phil Weldon' wrote, in part:
> So here is my contribution, using nTune reports for the actual values.
> | (Caveat - nTune is still a bit squirrelly; as an example sometimes I can
> get
> | nTune to report the SPD information for memory modules, sometimes not.)
> _____
>
> The solution (at least for my EVGA 680i motherboard) for reading the SPD
> information is to use 'nVidia Control Panel' and select 'View System
> Information' (this can also be done by starting 'nTune'.
>
> THEN
>
> Double-click on one of the processors in 'Processor information'; this
> will
> display the processor information.
>
> THEN
>
> Double-click on one of the DIMMs in 'Memory information'; this will
> display
> the SPD information, including any extended SPD information (SLI-ready
> information.)
>
> After double-clicking on one of the processors in 'Processor information'
> in
> 'View System Information' clicking on the 'View SPD' button in 'Adjust
> Motherboard Settings', then double-clicking on a DIMM in the 'Memory
> Information' window WILL display the SPD information, including any
> extended
> SPD information (SLI-ready information.)
>
> If I do not double-click on a processor in 'View System Information'
> FIRST,
> double-clicking on a DIMM in 'Memory information' does nothing AND
> double-clicking on a DIMM in 'Adjust Motherboard Settings' in the 'Memory
> Information' window [brought up by the 'View SPD information'] does
> nothing.
>
> Using the 'Create a tech support log file:' 'Save' button ALWAYS works and
> creates a text log file that includes the SPD and extended SPD
> information.
>
Thanks for the heads-up. The same works on my Asus P5N32-SLI also so you
have hit upon something that, at least, gives NView some use. Before, I
could not get any drop-down adjustment choices at all. I will do a bit of
tweaking there just to see if it can actually work.
Core 2 Duo and NV 680i Overclocking 
Linear Mode
