Can't boot when I change my latency?

[tvfreak]

I followed a tutorial from DFI-street and to have my latency case change to 2-2-5-2. When I do my comp won't boot. Any ideas why? So, I kept it to the factory setting which is auto 2.5-3-5-2. So which is better to have? And what can I do to make it better? Or are the settings that i have know is fine? :twisted:

 

[TabrisDarkPeace]

I followed a tutorial from DFI-street and to have my latency case change to 2-2-5-2. When I do my comp won't boot. Any ideas why? So, I kept it to the factory setting which is auto 2.5-3-5-2. So which is better to have? And what can I do to make it better? Or are the settings that i have know is fine? :twisted:

Ask yourself "Why am I doing this in the first place ?"

If it won't even boot at 2-2-5-2 then even at 2.5-2-5-2 or 2-3-5-2 sure it might boot fine, but chances are it would fail a Prime95 'memory' test.

Sure the odd corruption here and there won't matter, after a few defrags you'll eventually corrupt most your files, and registry, and even over 6+ months of normal use you'll notice 'issues' from varioues files (eg: ones with executable content) changing. Aswell as the odd document going corrupt.

"I read it on the Internet by someone I've never met, so it must be true" , Sure with decent enough memory perhaps, but you'll gain 3 fps, if that, in most games going to those settings from 2.5-3-5-2 and risk far more than you'll gain.

If you want to overclock slow the timings down to 3-3-8-3, overclock, then reduce timings bit by bit until Prime95 tests fail.

================================================

You can run Prime95 from two different folders (or more, up to 32 should be OK) at once btw, so it hits both CPUs and memory at once.

It will cut your testing time in half if you have a dual-core system.... if you are not already doing it. Even on single-core with HyperThreading you'll still gain a fair bit of time. On quad-core it helps heaps

Tabris - Recommended Prime95 settings.txt:

Min FFT size (in K): 128
Max FFT size (in K): 4096 (or larger if supported)
Run FFTs in place: Disabled / Off
Memory to use (in MB): 1/n *
Time to run each FFT size (in minutes): As low as 15 and as high as 60 minutes to personal taste. (15 min recommended though, even for todays systems)

* - Where n is the number of cores/threads the system can run in parallel, and how many Prime95 instances are being run, each from their own folder at the same time.

eg: 4 threads at once = 1/4 memory = eg: 4 GB total, so 1024 MB each

Prime95 can detect errors in calculations (usually caused by faulty, or misconfigured, memory, or overclocking CPU to far), I don't know if SuperPI can though.

================================================

Remember a 60+ minute Prime95 stable system is pretty damn stable.

You may be risking more than you understand, try it on a system with no data your care about where data corruption won't 'hurt' while learning, then move from there after 3 months or so. That's just a suggestion though.

 

[-Feller-]

"You can run Prime95 from two different folders (or more, up to 32 should be OK) at once btw, so it hits both CPUs and memory at once. "

Check out this link

http://www.dfi-street.com/forum/showthread.php?t=16446

Very nice guide on how to use Prime95 on duel cores

 

[cisco]

That ram runs at 2.3.5.2 1T without a problem but trying to make it run any tighter will make it not boot or very unstable at best. I am using the same ram @2.3.5.2 1T without any problems.

 

[tvfreak]

:twisted:

I followed a tutorial from DFI-street and to have my latency case change to 2-2-5-2. When I do my comp won't boot. Any ideas why? So, I kept it to the factory setting which is auto 2.5-3-5-2. So which is better to have? And what can I do to make it better? Or are the settings that i have know is fine? :twisted:

Ask yourself "Why am I doing this in the first place ?"

If it won't even boot at 2-2-5-2 then even at 2.5-2-5-2 or 2-3-5-2 sure it might boot fine, but chances are it would fail a Prime95 'memory' test.

Sure the odd corruption here and there won't matter, after a few defrags you'll eventually corrupt most your files, and registry, and even over 6+ months of normal use you'll notice 'issues' from varioues files (eg: ones with executable content) changing. Aswell as the odd document going corrupt.

"I read it on the Internet by someone I've never met, so it must be true" , Sure with decent enough memory perhaps, but you'll gain 3 fps, if that, in most games going to those settings from 2.5-3-5-2 and risk far more than you'll gain.

If you want to overclock slow the timings down to 3-3-8-3, overclock, then reduce timings bit by bit until Prime95 tests fail.

================================================

You can run Prime95 from two different folders (or more, up to 32 should be OK) at once btw, so it hits both CPUs and memory at once.

It will cut your testing time in half if you have a dual-core system.... if you are not already doing it. Even on single-core with HyperThreading you'll still gain a fair bit of time. On quad-core it helps heaps

Tabris - Recommended Prime95 settings.txt:

Min FFT size (in K): 128
Max FFT size (in K): 4096 (or larger if supported)
Run FFTs in place: Disabled / Off
Memory to use (in MB): 1/n *
Time to run each FFT size (in minutes): As low as 15 and as high as 60 minutes to personal taste. (15 min recommended though, even for todays systems)

* - Where n is the number of cores/threads the system can run in parallel, and how many Prime95 instances are being run, each from their own folder at the same time.

eg: 4 threads at once = 1/4 memory = eg: 4 GB total, so 1024 MB each

Prime95 can detect errors in calculations (usually caused by faulty, or misconfigured, memory, or overclocking CPU to far), I don't know if SuperPI can though.

================================================

Remember a 60+ minute Prime95 stable system is pretty damn stable.

You may be risking more than you understand, try it on a system with no data your care about where data corruption won't 'hurt' while learning, then move from there after 3 months or so. That's just a suggestion though.
I was just following the tutorial. But I did change it back to stock settings which is 2.5-3-5-2. I ran memtest through the genie bios for 7hr + and no errors. But i'm going to run prime95. And I'll let you guys know what happen. Thanks for the tip. And yea i'm a nub. lol.

 

[tvfreak]

I followed a tutorial from DFI-street and to have my latency case change to 2-2-5-2. When I do my comp won't boot. Any ideas why? So, I kept it to the factory setting which is auto 2.5-3-5-2. So which is better to have? And what can I do to make it better? Or are the settings that i have know is fine? :twisted:

Ask yourself "Why am I doing this in the first place ?"

If it won't even boot at 2-2-5-2 then even at 2.5-2-5-2 or 2-3-5-2 sure it might boot fine, but chances are it would fail a Prime95 'memory' test.

Sure the odd corruption here and there won't matter, after a few defrags you'll eventually corrupt most your files, and registry, and even over 6+ months of normal use you'll notice 'issues' from varioues files (eg: ones with executable content) changing. Aswell as the odd document going corrupt.

"I read it on the Internet by someone I've never met, so it must be true" , Sure with decent enough memory perhaps, but you'll gain 3 fps, if that, in most games going to those settings from 2.5-3-5-2 and risk far more than you'll gain.

If you want to overclock slow the timings down to 3-3-8-3, overclock, then reduce timings bit by bit until Prime95 tests fail.

================================================

You can run Prime95 from two different folders (or more, up to 32 should be OK) at once btw, so it hits both CPUs and memory at once.

It will cut your testing time in half if you have a dual-core system.... if you are not already doing it. Even on single-core with HyperThreading you'll still gain a fair bit of time. On quad-core it helps heaps

Tabris - Recommended Prime95 settings.txt:

Min FFT size (in K): 128
Max FFT size (in K): 4096 (or larger if supported)
Run FFTs in place: Disabled / Off
Memory to use (in MB): 1/n *
Time to run each FFT size (in minutes): As low as 15 and as high as 60 minutes to personal taste. (15 min recommended though, even for todays systems)

* - Where n is the number of cores/threads the system can run in parallel, and how many Prime95 instances are being run, each from their own folder at the same time.

eg: 4 threads at once = 1/4 memory = eg: 4 GB total, so 1024 MB each

Prime95 can detect errors in calculations (usually caused by faulty, or misconfigured, memory, or overclocking CPU to far), I don't know if SuperPI can though.

================================================

Remember a 60+ minute Prime95 stable system is pretty damn stable.

You may be risking more than you understand, try it on a system with no data your care about where data corruption won't 'hurt' while learning, then move from there after 3 months or so. That's just a suggestion though.
Could this damage my system running prime95 into 2 folders for my dual core? 8O

 

[tvfreak]

That ram runs at 2.3.5.2 1T without a problem but trying to make it run any tighter will make it not boot or very unstable at best. I am using the same ram @2.3.5.2 1T without any problems.

What is your sig? Have you ran prime95?

 

[tvfreak]

I followed a tutorial from DFI-street and to have my latency case change to 2-2-5-2. When I do my comp won't boot. Any ideas why? So, I kept it to the factory setting which is auto 2.5-3-5-2. So which is better to have? And what can I do to make it better? Or are the settings that i have know is fine? :twisted:

Ask yourself "Why am I doing this in the first place ?"

If it won't even boot at 2-2-5-2 then even at 2.5-2-5-2 or 2-3-5-2 sure it might boot fine, but chances are it would fail a Prime95 'memory' test.

Sure the odd corruption here and there won't matter, after a few defrags you'll eventually corrupt most your files, and registry, and even over 6+ months of normal use you'll notice 'issues' from varioues files (eg: ones with executable content) changing. Aswell as the odd document going corrupt.

"I read it on the Internet by someone I've never met, so it must be true" , Sure with decent enough memory perhaps, but you'll gain 3 fps, if that, in most games going to those settings from 2.5-3-5-2 and risk far more than you'll gain.

If you want to overclock slow the timings down to 3-3-8-3, overclock, then reduce timings bit by bit until Prime95 tests fail.

================================================

You can run Prime95 from two different folders (or more, up to 32 should be OK) at once btw, so it hits both CPUs and memory at once.

It will cut your testing time in half if you have a dual-core system.... if you are not already doing it. Even on single-core with HyperThreading you'll still gain a fair bit of time. On quad-core it helps heaps

Tabris - Recommended Prime95 settings.txt:

Min FFT size (in K): 128
Max FFT size (in K): 4096 (or larger if supported)
Run FFTs in place: Disabled / Off
Memory to use (in MB): 1/n *
Time to run each FFT size (in minutes): As low as 15 and as high as 60 minutes to personal taste. (15 min recommended though, even for todays systems)

* - Where n is the number of cores/threads the system can run in parallel, and how many Prime95 instances are being run, each from their own folder at the same time.

eg: 4 threads at once = 1/4 memory = eg: 4 GB total, so 1024 MB each

Prime95 can detect errors in calculations (usually caused by faulty, or misconfigured, memory, or overclocking CPU to far), I don't know if SuperPI can though.

================================================

Remember a 60+ minute Prime95 stable system is pretty damn stable.

You may be risking more than you understand, try it on a system with no data your care about where data corruption won't 'hurt' while learning, then move from there after 3 months or so. That's just a suggestion though.

Is this test for only OCer? My stuff is only stock at the moment.

 

[TabrisDarkPeace]

Could this damage my system running prime95 into 2 folders for my dual core? 8O

Only if the hardware was unstable and massively misconfigured to begin with. It is highly unlikely but I wont lie by saying there is no chance.

Bearing in mind you can take the heatsink off a CPU these days and not see smoke from the CPU. 8) - It would be difficult to damage a decent system using just Prime95 + common sense. Of course if it fails at stock setting then diagnose that first, especially before overclocking.

If the system is unstable running Prime95 at 100% load, regardless of settings, even at stock, then something is wrong with it.

I'd rather Prime95 tell me, than the Windows Registry (by going corrupt), or an all important file, or a game installation not working (eg: files 'corrupted' while sitting in write back cache [RAM] then flushed to disk = wrong contents, but maybe only an inverted bit here and there... can cause screwy, hard to diagnose problems).

Prime95 isn't only for the overclocker, it is a more effective, and faster to run, way of testing the memory subsystem and CPU. The settings above will hit the RAM and CPU fairly hard, harder than most games. Unlike games Prime95 can detect miscalculations and 'handle' them gracefully, a game, the registry, or the OS kernel on the other hand...

 

[tvfreak]

Could this damage my system running prime95 into 2 folders for my dual core? 8O

Only if the hardware was unstable and massively misconfigured to begin with. It is highly unlikely but I wont lie by saying there is no chance.

Bearing in mind you can take the heatsink off a CPU these days and not see smoke from the CPU. 8) - It would be difficult to damage a decent system using just Prime95 + common sense. Of course if it fails at stock setting then diagnose that first, especially before overclocking.

If the system is unstable running Prime95 at 100% load, regardless of settings, even at stock, then something is wrong with it.

I'd rather Prime95 tell me, than the Windows Registry (by going corrupt), or an all important file, or a game installation not working (eg: files 'corrupted' while sitting in write back cache [RAM] then flushed to disk = wrong contents, but maybe only an inverted bit here and there... can cause screwy, hard to diagnose problems).

Prime95 isn't only for the overclocker, it is a more effective, and faster to run, way of testing the memory subsystem and CPU. The settings above will hit the RAM and CPU fairly hard, harder than most games. Unlike games Prime95 can detect miscalculations and 'handle' them gracefully, a game, the registry, or the OS kernel on the other hand...

Since I have a dual core opty do I need to run prime95 in 2 files to get a accurate reading, or can I just do it the regular way??? I was reading the forum in dfi-street and telling how to test it in 2 files, but it just seems way to confusing for a newbie like me. :twisted:

 

[TabrisDarkPeace]

It'll only be a test on 1 core otherwise.

Use my instructions, affinity only helps if you want to test and submit Prime numbers.

Not using affinity will place the system under slightly higher load.

Just unzip it twice, eg:
C:\TESTING\PRIME95.1
C:\TESTING\PRIME95.2

Run both at once, just the EXEs, don't bother with the shortcuts, etc
Click Just Testing
Options, Torture Test, Custom, [Use Above settings].

Boom easy, takes < 3 min to setup the first time.

8) Also practical enough to store all 4 on on of those 'useless' small / free USB sticks everyone has.

 

[cisco]

Haven't ran Prime, haven't felt the need. I am running the ram within its specs and have not had any issues that would make me think I had a stability problem. This is my second set of this ram both ran fine at those timings. I also owned the 2x512mb same timings for about 2 years.



3500+ Winchester
MSI K8N Neo Platimun
2x1gig OCZ Plat EL
BFG 7800GT OC 470/1150
74gig Raptor
36gig Raptor
Ultra 500w PSU

 

[tvfreak]

It'll only be a test on 1 core otherwise.

Use my instructions, affinity only helps if you want to test and submit Prime numbers.

Not using affinity will place the system under slightly higher load.

Just unzip it twice, eg:
C:TESTINGPRIME95.1
C:TESTINGPRIME95.2

Run both at once, just the EXEs, don't bother with the shortcuts, etc
Click Just Testing
Options, Torture Test, Custom, [Use Above settings].

Boom easy, takes < 3 min to setup the first time.

8) Also practical enough to store all 4 on on of those 'useless' small / free USB sticks everyone has.

Well I did get it to work. I did follow the instructions and for some reason a newbie like me did it lol. So in the dfi-street forum the guy "angry games" stated that if your using 2 gig stick of memory make sure to use around 800mb in the options/torture test/custom and change the "memory to use". Cause the default settings is more than that. :twisted: