Because you wrote about the pcie x1->x16 adapters. There are some with Molex power to get the needed power. But not easy to get them shipped to austria in my case :( In Germany they are available over several shops. I found one too for UK people.

I bought one here.

Thats not a pcie x1 to x16 adapter ;)

---

DSKAG Austria: http://www.dskag.at

Because you wrote about the pcie x1->x16 adapters. There are some with Molex power to get the needed power. But not easy to get them shipped to austria in my case :( In Germany they are available over several shops. I found one too for UK people.

I bought one here.

Thats not a pcie x1 to x16 adapter ;)

Oh dear... This is getting complicated.

Please show me one. Thank you.

Interesting what I read here.
To measure some things I have some tips. I use a power monitor, it's a device what is set between the wall plug and the PC plug. The readings give the entire wattage used. There is a difference when idle and at full load. I see around 300W used with one GPU (550Ti) and 4 CPU at full load. And 358W with a GTX660 and 7 cores from an i7, with also 1 SSD, 1 Blueray, 1 DVD and 2 HD. This PSU gives 880W.

A second handy thing is an infrared thermometer, not expensive anymore bought it at Conrad. With a laser pointer it is possible the pin point at a certain location and then read the temperature. With an open case it is even easier.

Both instruments can be an eye-opener for power draw/consumption and temperature/overheating.

@Tomba, with that PSU of yours, 2 GPU's will work. However I don't now about the riser, as I have no experience with that. But skgive wrote that he had errors. So keep in mind that if it doesn't work in your system, do not directly think that the PSU is the bottle neck.

---

Greetings from TJ

And 358W with a GTX660 and 7 cores from an i7, with also 1 SSD, 1 Blueray, 1 DVD and 2 HD. This PSU gives 880W.

How much is it drawing when your HDDs and optical drives are in use?

You have a lot of headroom with your PSU, but imagine you only had around 600W (at 100% PSU capacity) and two mid-high range GPUs at full load. With HDDs connected I would be concerned for possible damage, meaning data backups would need greater frequency. Overclocking also adds to power stability requirements. Many people may be comfortable with 80%+ load, but I like to keep it under that.

Modern desktop HDDs draw about 3 W more under load than when idle. Add 1 - 2 W for older ones. Idle power draw is in the range of 5 W, up to 8 W for older ones. At startup they'll briefly need 20 - 25 W, but at this point the GPUs do not yet draw power, so for cruncher PCs this doesn't matter at all ;)

Not as sure for optical drives (they're hardly under load anymore, aren't they?), but around 10 W under load should be about right, a bit more for a burner.

MrS

---

Scanning for our furry friends since Jan 2002

Western Digital Black drives pull about 10-11 W at load. I could debate the numbers on optical drives, etc, but it would distract from my main point. The more PSU headroom the better, and going over 80% load may cause stability issues. This is just for caution. If someone is comfortable with pushing their PSU, more power(hah) to them.

Just found and installed SPECCY http://www.piriform.com/speccy a Windows tool which will tell you both (and a whole lot more information about your system). They have a free version and a pay version, I used the free version.

That's a "Bingo!". Thank you.

Right now, before I install the second GPU, the room temperature is 26.5C and Speccy shows my Dell/Intel mobo between 63C and 67C.

When I install the second GPU, what is the max mobo temperature I should be looking for to avoid it bursting into flames??

Thanks, Tom

>60C for your motherboard is VERY hot! In the hot Greek summer I'm going through now, with ambient temps ~35C, my temps are:

coretemp-isa-0000
Adapter: ISA adapter
Core 0:       +72.0°C  (high = +83.0°C, crit = +99.0°C)
Core 1:       +66.0°C  (high = +83.0°C, crit = +99.0°C)
Core 2:       +64.0°C  (high = +83.0°C, crit = +99.0°C)
Core 3:       +64.0°C  (high = +83.0°C, crit = +99.0°C)

atk0110-acpi-0
Adapter: ACPI interface
Vcore Voltage:       +1.22 V  (min =  +0.80 V, max =  +1.60 V)
+3.3V Voltage:       +3.38 V  (min =  +2.97 V, max =  +3.63 V)
+5V Voltage:         +5.14 V  (min =  +4.50 V, max =  +5.50 V)
+12V Voltage:       +12.26 V  (min = +10.20 V, max = +13.80 V)
CPU Fan Speed:      1110 RPM  (min =  600 RPM)
Chassis1 Fan Speed: 1520 RPM  (min =  600 RPM)
Chassis2 Fan Speed: 1052 RPM  (min =  600 RPM)
Power Fan Speed:     896 RPM  (min =    0 RPM)
CPU Temperature:     +71.0°C  (high = +45.0°C, crit = +45.5°C)
MB Temperature:      +38.0°C  (high = +45.0°C, crit = +46.0°C)

    Fan Speed                       : 55 %
        Gpu                         : 67 C

With a 27C ambient temperature, my temps would be <60C (around 55) for CPU and GPU and about 30C for my motherboard.

Most probably, the temperature you're seeing is not your motherboard's.

---

Most probably, the temperature you're seeing is not your motherboard's.

That is hot Tomba.
I have checked my PC's with the same program and see reading between 36-48°C with an ambient temperature of 30.5°C!.
Very strange for a Dell, as I have several types even an XPS420 and that one runs 24/7 and quite cool (40°C of MOBO) but has a smaller PSU than you have. I have still the original Dell.
Your temperature is to high, you should try do to something about it. If you clean all the fans with canned air and you close to case (airflow will be tunneled then, Dell does that smart), and you don't start any applications for a while, what is then the temperature reading?

---

Greetings from TJ

Your temperature is to high, you should try do to something about it. If you clean all the fans with canned air and you close to case (airflow will be tunneled then, Dell does that smart), and you don't start any applications for a while, what is then the temperature reading?

Thanks for responding. As a first step I suspended the active GPUGrid WU. Temp is now 40C/41C.

Now for the blow bit...

First I'd make sure the reading is actually correct. Which sensor measures what is something which can quickly be screwed up any party involved. That the tool calls it "system temperature" doesn't mean anything on its own. Hence my suggestion to feel the air for yourself - then you'll know for sure if you've got some software reading error or if you're cooking your hardware.

And hence the question regarding the case fans.. the less the more likely the reading is true ;)

MrS

---

Scanning for our furry friends since Jan 2002

Going by that window grab 'System Temperature' means the South Bridge Chip, and it's 63°C. It's probably covered by a heatsink. IIRC 63°C is reasonable enough for an early i7.

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Now for the blow bit...

I usually blow out the case every three months so, apart from a small amount of dust, it was clean enough. All the case-side ventilation holes were clear of fluff and spider webs(!)

After reassembling I took temp readings with nothing running - top pic - and with just GPUGrid running - bottom pic:

Your CPU and system temperatures are too high. The CPU, I can attribute to the lousy Intel stock cooler, it is an EPIC fail on its part. I HIGHLY recommend, especially for prolonged crunching, a good aftermarket cooler. You can find many decent ones at very reasonable prices, that will just blow the Intel cooler out of the water!

A high system temperature is unusual. I would attribute it to poor case ventilation or even a failure of some cooling part. Perhaps you should check your motherboard's coolers. Maybe something like the south-bridge cooler has detached from the chip? I would do a visual as well as "touching" test of the various coolers. See if everything is where it must be and make sure it is firmly seated / attached to what it's supposed to cool.

---

I know some of you will be cringing, with finger on fire department emergency button, but I had to do it...

The PCIe 1-16 riser, and the Molex PCIe power adapter, arrived. I connected the 660 to the Molex, and the two PSU PCIe connectors to the 460. Windows spent a long time installing S/W for the 460 but it appeared. See pic below.

My next challenge (Help!) is how to tell BOINC about the 460. I tried several GPUGrid "Updates" but no joy. How do I do that?

Did you try this?

<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
</options>
</cc_config>

Save as cc_config.xml if you have it already, disregard.

cc_config - use all GPU's - FAQ - Best configurations for GPUGRID

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

Did you try this?

<cc_config>
<options>
<use_all_gpus>1</use_all_gpus>
</options>
</cc_config>

Save as cc_config.xml if you have it already, disregard.

Where do I save it in Win 7 ??

C:\ProgramData\BOINC (this folder will likely be hidden, but should open if you type it into Explorer)

From, FAQ - Best configurations for GPUGRID

---

FAQ's

HOW TO:
- Opt out of Beta Tests
- Ask for Help

C:\ProgramData\BOINC (this folder will likely be hidden, but should open if you type it into Explorer

Wow! It's all happening!! Thank you!

First, I got two POEMs, and the GPUGrid WU went to sleep. Then one of the POEMs died and a GPUGrid WU downloaded. Here we are now, which I guess is what I was aiming for!

Many thanks, Tom



Temps are here: