• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

New FOA, i7 Low PPD?

GhosT15

Weaksauce
Joined
Jul 9, 2010
Messages
81
Started putting my little super computer to work while I'm away about a week ago now. It's a Core i7 Running at 4.06 GHz (Water Cooled) with a GTX 280 OC.

It seems like I might have something mis-configured, as the Core i7 has a very low PPD.

Some of the results:

Project : 6701
Core : Unknown
Frames : 100
Credit : 921

Min. Time / Frame : 7mn 47s - 1703.95 ppd
Avg. Time / Frame : 8mn 55s - 1487.37 ppd
Cur. Time / Frame : 9mn 26s - 1405.91 ppd
R3F. Time / Frame : 9mn 07s - 1454.74 ppd
Eff. Time / Frame : 11mn 43s - 1131.93 ppd

Project : 6020
Core : Unknown
Frames : 100
Credit : 467

Min. Time / Frame : 3mn 43s - 1809.36 ppd
Avg. Time / Frame : 3mn 58s - 1695.33 ppd
Cur. Time / Frame : 4mn 42s - 1430.81 ppd
R3F. Time / Frame : 4mn 35s - 1467.23 ppd
Eff. Time / Frame : 4mn 05s - 1646.89 ppd

Project : 6012
Core : Unknown
Frames : 100
Credit : 470

Min. Time / Frame : 3mn 07s - 2171.55 ppd
Avg. Time / Frame : 3mn 07s - 2171.55 ppd


The Configuration on the SMP is:

[settings]
username=Allen_Waldrop
team=33
passkey=*
asknet=no
machineid=2
bigpackets=big
extra_parms=-smp -verbosity 9
local=18

[http]
active=no
host=localhost
port=8080
usereg=no

[core]
priority=0
cpuusage=100
disableassembly=no
checkpoint=5
ignoredeadlines=no
nocpulock=0
addr=

[power]
battery=yes

[clienttype]
memory=2048
type=3


If this is normal, just let me know! It seems crazy the GPU is that much faster...

GPU sample:

Project : 10515
Core : GPUv2 Gromacs
Frames : 100
Credit : 587

-- MainBoxGPU --

Min. Time / Frame : 1mn 04s - 7924.50 ppd
Avg. Time / Frame : 1mn 05s - 7802.58 ppd
Cur. Time / Frame : 1mn 07s - 7569.67 ppd
R3F. Time / Frame : 1mn 07s - 7569.67 ppd
Eff. Time / Frame : 1mn 08s - 7458.35 ppd

 
First of all - WELCOME TO THE TEAM! :D

6701's are horrible WU's, we're all currently grumbling about them.
They're absolutely killing anything including i7's....they've turned my q9450 and e8400 into practically useless f@h boxen. :( And that's all that seems to be available these days, even though I've pulled the -advmethods flag from all of my clients. :eek:
Basically there are quite a few low PPD/low point total, high processing time WU's out right now; no one has any idea how long they'll be around or if they are a sign of [worse] things to come.
 
I think that there's something going on here besides the crappy 6701 ppd that everyone is experiencing. Your ppd for 6012 and 6020 should be much higher. Assuming that your pass key was not mistyped, your config looks right. Is it possible that you have HT disabled or that your CPU is throttling due to temperature?
 
I had setup another machine the other day with no advmethods and it still had client type set to 3. I changed it to 0 in the config file and no 670x WU's since then.

Also is HFM.net or whatever you're using set to calculate the bonuses?
 
In the options in HFM there is an option to calculate bonus points - do you have that checked as your time frames indicate at least 9k PPD.

Also make sure your smp and GPU client machine ID's are different
 
I had setup another machine the other day with no advmethods and it still had client type set to 3. I changed it to 0 in the config file and no 670x WU's since then.

Also is HFM.net or whatever you're using set to calculate the bonuses?

Now that is good info... ill see if that works on my i7 rig.
 
Thanks for all the replies,

HT is enabled, I'm getting all 8 Cores as expected.

Looks like I found the problem, originally I thought it wasnt lowering the Multiplier on me, during my inital OC it didnt; Looks like the summer temps are causing it to lower the multiplier at full load once it heats up! I'm supprised that the water cooling can't keep up with the 4x120mm radiator. Might be time to re-apply thermal paste and check everything.

The Machine ID is 2 and the GPU Machine ID is 1

I'm using FahMon 2.3.99.1 (it seemed to be one of the recommended programs when I was reading over the reviews)

Should I use a different program to check?


 
Last edited:
Use HFM.NET. You get it here.

I'm glad that we were able to figure out the problem. Check all of your connections and thermal paste, because your water cooling system should be more than capable of keeping up (there are many i7 users here and most are just using air). Once you get it working, you'll love the ppd.

 
Use HFM.NET. You get it here.

I'm glad that we were able to figure out the problem. Check all of your connections and thermal paste, because your water cooling system should be more than capable of keeping up (there are many i7 users here and most are just using air). Once you get it working, you'll love the ppd.


I did a little digging, and I might have found it to be an Intel Limit on 130w thermal.

"Now for the good news. Switch off the CPU TM Function in ASUS’ BIOS (or the CPU VR Current Limit Override in Intel’s), and the capability stops interfering—I looped Cinebench runs without ever encountering a power-induced throttling state."

Might be worth looking at too!


Thanks for the Program link, will take a look at that
 
Yeah FahMon used to be the one to use before the bonus plan rolled out; now it's the only one that doesn't calculate the bonus plan.... HFM.NET and FahSpy do though (and the majority of us use HFM.NET because it's a lot better supported; the developer even regularly posts updates here too :)).

Once you do your 10 regular units with your passkey in place, you'll start getting bonus points for all the units you complete from that point on that are eligible for bonus points (which are most of the cpu WU's these days). ;) Again, welcome to the team!
 
Add moar fans!!!!

moar1.jpg


BTW, Welcome to the team!

 
So i've done some experimenting, and apparently its not a thermal issue but a watt issue. The CPU is going over 130w when under full load and OC'ed, and its lowering my multiplier from 30 to 27... Its not thermal, as
1) its only running at 65-66c per Real Temp
2) I disabled TM (Thermal Managment) in the BIOS.

Does anyone have any Ideas on what I can do to fix this multiplier problem?

Computer Details:

Multplier coming from "i7 Turbo 6.95"
CPU: Intel Core i7 Exteme 965
MB: P6T Deluxe , BIOS Revision 2101 (Just flashed to lastest) Dated 2/24/10
Memory: 3x2GB DDR3 @ 804 MHz Corair
 
Welcome to the team! We can use all the help we can get keeping eVGA's team at bay! I'm glad we were able to get the issue worked out..



 
Done some more googling,

seems my problem is : Overspeed Protection , which (from what google has provided so far) cannot be turned off on my motherboard (P6T Deluxe) ... I really really hope this is wrong, or I'm going to be pissed at Asus
 
My board is an Asus and the only time my multiplier ever dropped was due to high temps.
 
My board is an Asus and the only time my multiplier ever dropped was due to high temps.

I've been trying to research this, but I've seen plenty of reports of 70c+ and not being thermally blocked. I've also Disabled the Thermal Protection on the Motherboard, and it is still lowering the clock.

From what i've read, its because I have my CPU Voltage above 1.3 that it is trying to pull more than 130w which causes this problem. Other motherboards have the ability to raise or turn off this protection "Overspeed Protection" which is different from the thermal protection.

"Overspeed protection is the feature that caps the maximum power usage of the Core i7 at 130W by monitoring the core voltage and amperage."


When I start up Prime95, the Watts go from 30-40 to: 144ish, clearly its lowering my muliplier because of this:

power.png
 
Last edited:
Gotta be on your mobo then, my volts are above 1.3 on all my nehalems and I've never heard of something like that
Posted via [H] Mobile Device
 
I disabled the TM function on my RIIE and set the multiplier to 21. No problems at all, even with running my CPU past 90C. If you really can't get the 21x multiplier to stay though, I suggest you consider lowering it to 20 and just raising your base clock more.
 
When I start up Prime95, the Watts go from 30-40 to: 144ish, clearly its lowering my muliplier because of this:
Not really. Mine does that too, and it's still at 21x multi.
 
I've been trying to research this, but I've seen plenty of reports of 70c+ and not being thermally blocked. I've also Disabled the Thermal Protection on the Motherboard, and it is still lowering the clock.

From what i've read, its because I have my CPU Voltage above 1.3 that it is trying to pull more than 130w which causes this problem. Other motherboards have the ability to raise or turn off this protection "Overspeed Protection" which is different from the thermal protection.

"Overspeed protection is the feature that caps the maximum power usage of the Core i7 at 130W by monitoring the core voltage and amperage."


When I start up Prime95, the Watts go from 30-40 to: 144ish, clearly its lowering my muliplier because of this:

why are you running such a high multiplyer? 200X20 would be much faster in everyway.
 
I disabled the TM function on my RIIE and set the multiplier to 21. No problems at all, even with running my CPU past 90C. If you really can't get the 21x multiplier to stay though, I suggest you consider lowering it to 20 and just raising your base clock more.

Thanks, you've confirmed that the problem is not thermal like most have been thinking, as mine never go above 65c, TM enabled or disabled.

Keep in mind I'm running on a 30x multiplier not 20x-21x. I have the unlocked i7. The default Multiplier is 25x

Not really. Mine does that too, and it's still at 21x multi.

Is your's running with a higher cpu voltage than default?

why are you running such a high multiplyer? 200X20 would be much faster in everyway.

Why not run a higher mulitplier? This way less components are overclocked on the bclkc. This is my First i7 overclock, but do you have any information on why the higher bclck and lower multiplier would be better? The default Multiplier is 25x on this i7, so I'd be lowering the multiplier from base in your suggestion. Or should I Just leave the default Mutliplier and only up the bclck?

My Settings:

vCore: 1.3625
QPIv: 1.325
DDRv: 1.64 (stupid mb wont do 1.65, next step is 1.66)
DDR Frequency: 1624
BClck: 135
Multiplier: 30
Speed: 4.03GHz
 
Thanks, you've confirmed that the problem is not thermal like most have been thinking, as mine never go above 65c, TM enabled or disabled.

Keep in mind I'm running on a 30x multiplier not 20x-21x. I have the unlocked i7. The default Multiplier is 25x
Ah, you never said that. I assumed that you were using a normal i7 and taking advantage of Turbo Mode to raise the multiplier by one, and that feature wasn't working properly. If you've got an Extreme Edition CPU and the multiplier isn't staying, then that's a much bigger problem.
 
Ah, you never said that. I assumed that you were using a normal i7 and taking advantage of Turbo Mode to raise the multiplier by one, and that feature wasn't working properly. If you've got an Extreme Edition CPU and the multiplier isn't staying, then that's a much bigger problem.

Yeah, its very concerning that the EE Processor is not keeping the multiplier. Now that I'm home I tried out a new OC (so far stable)...

Using 25 Multiplier (The base) and using 162 Bclck, getting 4.05GHz , Temps are so far the same, 63-64c and the Mutliplier is not dropping.

It almost sounds like I have a bad EE/MB causing the Multiplier to drop down; I should be able to set it at whatever I want and it should not go down under load.

Anyone have any ideas?
 
higher blck clock vs higher multiplyer. given the same finall speed the one with the higher blck clock will always be faster. more memory bandwidth and faster internal data movment. i run 210X19. you posted doing a 6012 in 3:07 which is 16,169ppd i do those in 2:43 which is 19,869ppd. Not trying to give you a hard time. just wanna help you get the most out your hardware.
 
Last edited:
Yeah, its very concerning that the EE Processor is not keeping the multiplier. Now that I'm home I tried out a new OC (so far stable)...

Using 25 Multiplier (The base) and using 162 Bclck, getting 4.05GHz , Temps are so far the same, 63-64c and the Mutliplier is not dropping.

It almost sounds like I have a bad EE/MB causing the Multiplier to drop down; I should be able to set it at whatever I want and it should not go down under load.

Anyone have any ideas?
Sounds like a motherboard issue to me honestly. The CPU itself doesn't have any control over what the multiplier is set to.

Have you disabled all the power-saving options in the BIOS?
 
Thanks for all the feedback everyone!

higher blck clock vs higher multiplyer. given the same finall speed the one with the higher blck clock will always be faster. more memory bandwidth and faster internal data movment. i run 210X19. you posted doing a 6012 in 3:07 which is 16,169ppd i do those in 2:43 which is 19,869ppd. Not trying to give you a hard time. just wanna help you get the most out your hardware.

No problem, and I didn't take any offence (if you thought I did) but the memory was running at the same speed either way, I just adjusted the DDR:BCLCK Ratio, both setups the memory was running at ~1824 MHz Your advice has however temporarily netted me a fix to my Frequency problem, until I can (hopefully) get it resolved with ASUS. Otherwise what’s the point of the unlocked multiplier if I can’t set it myself and have it stay!



I'll be trying to get in touch with Asus next weekend, although I've been able to steadly climb the ranks!


 
Back
Top