Bitcoin Forum
November 23, 2017, 06:31:11 AM *
News: Latest stable version of Bitcoin Core: 0.15.1  [Torrent].
 
   Home   Help Search Donate Login Register  
Pages: [1]
  Print  
Author Topic: XFX 5970 BE pulling too many amps?  (Read 3708 times)
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 20, 2013, 09:14:48 AM
 #1

Hey guys, I recently bought an XFX 5970 black edition 2GB and I am somewhat confused about the amps this thing seems to be pulling. I'm running all stock at 725/1000 and stock volts 1.1 When I'm mining litecoin, both gpu-z and trixx are reporting the card VDDC drawing an average of 75.5-76.6 amps while VDDCI sits at 23.1-24.1. I'm using a 750watt Alienware PSU http://www.geeks.com/details.asp?invtid=NPS-750AB-1-R and I'm somewhat confused with the results. The PSU itself has 4x 6pins but no 8pins. I'm currently using a x2 molex to 8pin to power the card Has anyone ever heard of the card pulling this much amperage? I have read that the 5970's usually pull around 50Amps. Any help would be much appreciated!
1511418671
Hero Member
*
Offline Offline

Posts: 1511418671

View Profile Personal Message (Offline)

Ignore
1511418671
Reply with quote  #2

1511418671
Report to moderator
1511418671
Hero Member
*
Offline Offline

Posts: 1511418671

View Profile Personal Message (Offline)

Ignore
1511418671
Reply with quote  #2

1511418671
Report to moderator
Join ICO Now Coinlancer is Disrupting the Freelance marketplace!
Advertised sites are not endorsed by the Bitcoin Forum. They may be unsafe, untrustworthy, or illegal in your jurisdiction. Advertise here.
1511418671
Hero Member
*
Offline Offline

Posts: 1511418671

View Profile Personal Message (Offline)

Ignore
1511418671
Reply with quote  #2

1511418671
Report to moderator
1511418671
Hero Member
*
Offline Offline

Posts: 1511418671

View Profile Personal Message (Offline)

Ignore
1511418671
Reply with quote  #2

1511418671
Report to moderator
1511418671
Hero Member
*
Offline Offline

Posts: 1511418671

View Profile Personal Message (Offline)

Ignore
1511418671
Reply with quote  #2

1511418671
Report to moderator
crazyates
Legendary
*
Offline Offline

Activity: 952



View Profile
January 20, 2013, 04:05:54 PM
 #2

That's actually to be expected. The problem is you're only looking at one number, and it's kinda useless by itself.

For basic purposes, Wattage = Voltage x Amperage. A 5970 often needs a 30A rail on a PSU rather than a 20A, but the PSU is running at 12V. 12V x 20A = 240W is not enough to run a 5970. 12V x 30A = 360W is enough.

When GPU-z is showing you that amp figure for the VDDC, it's running 75A x 1.1V = ~83W. This should be 83W per core, so just your cores are pulling 165W. The rest of your card should be pulling another 100W or so, but I don't remember. 

Note: I'm no electrical engineer, but does that make sense?

Tips? 1crazy8pMqgwJ7tX7ZPZmyPwFbc6xZKM9
Previous Trade History - Sale Thread
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 21, 2013, 12:55:06 AM
 #3

That's actually to be expected. The problem is you're only looking at one number, and it's kinda useless by itself.

For basic purposes, Wattage = Voltage x Amperage. A 5970 often needs a 30A rail on a PSU rather than a 20A, but the PSU is running at 12V. 12V x 20A = 240W is not enough to run a 5970. 12V x 30A = 360W is enough.

When GPU-z is showing you that amp figure for the VDDC, it's running 75A x 1.1V = ~83W. This should be 83W per core, so just your cores are pulling 165W. The rest of your card should be pulling another 100W or so, but I don't remember. 

Note: I'm no electrical engineer, but does that make sense?

That somewhat makes sense, so my PSU pretty much doesn't have the required amps on the rails to properly run the 5970 then?


crazyates
Legendary
*
Offline Offline

Activity: 952



View Profile
January 21, 2013, 03:57:06 AM
 #4

That's actually to be expected. The problem is you're only looking at one number, and it's kinda useless by itself.

For basic purposes, Wattage = Voltage x Amperage. A 5970 often needs a 30A rail on a PSU rather than a 20A, but the PSU is running at 12V. 12V x 20A = 240W is not enough to run a 5970. 12V x 30A = 360W is enough.

When GPU-z is showing you that amp figure for the VDDC, it's running 75A x 1.1V = ~83W. This should be 83W per core, so just your cores are pulling 165W. The rest of your card should be pulling another 100W or so, but I don't remember. 

Note: I'm no electrical engineer, but does that make sense?
That somewhat makes sense, so my PSU pretty much doesn't have the required amps on the rails to properly run the 5970 then?
Like I said, your PSU is powering your GPU off a 12V rail. Most 5970s won't work on a 20A rail, but work fine on a 30A rail. Your PSU looks like it has 4 rails at 18A each. You might be able to load balance them to power the 5970 off 2 of the 18A rails, but you have to be careful which plugs go where. If load-balancing is not an option, you can pick up a decent Single Rail PSU for under $100.

Tips? 1crazy8pMqgwJ7tX7ZPZmyPwFbc6xZKM9
Previous Trade History - Sale Thread
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 21, 2013, 04:18:53 AM
 #5


Like I said, your PSU is powering your GPU off a 12V rail. Most 5970s won't work on a 20A rail, but work fine on a 30A rail. Your PSU looks like it has 4 rails at 18A each. You might be able to load balance them to power the 5970 off 2 of the 18A rails, but you have to be careful which plugs go where. If load-balancing is not an option, you can pick up a decent Single Rail PSU for under $100.

I realize that my PSU is powering my GPU off a 12v rail, but if most 5970's wont work on a 20A rail, how is it mine is working and currently been mining the past few days? By load balancing do you mean to switch up where certain things are plugged in? I currently only have the mobo,gpu,hdd and an optical hooked up. I have the x2 molex to 8pin plugged into the same string of molex connectors, So should I try a different string, or plug each molex into 2 different strings?
crazyates
Legendary
*
Offline Offline

Activity: 952



View Profile
January 21, 2013, 04:53:34 AM
 #6


Like I said, your PSU is powering your GPU off a 12V rail. Most 5970s won't work on a 20A rail, but work fine on a 30A rail. Your PSU looks like it has 4 rails at 18A each. You might be able to load balance them to power the 5970 off 2 of the 18A rails, but you have to be careful which plugs go where. If load-balancing is not an option, you can pick up a decent Single Rail PSU for under $100.

I realize that my PSU is powering my GPU off a 12v rail, but if most 5970's wont work on a 20A rail, how is it mine is working and currently been mining the past few days? By load balancing do you mean to switch up where certain things are plugged in? I currently only have the mobo,gpu,hdd and an optical hooked up. I have the x2 molex to 8pin plugged into the same string of molex connectors, So should I try a different string, or plug each molex into 2 different strings?

Yes, if you're using molex to PCIe 6/8 pin adapters, then you should be putting the adapters on different strings to make sure they're on different rails.

Tips? 1crazy8pMqgwJ7tX7ZPZmyPwFbc6xZKM9
Previous Trade History - Sale Thread
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 21, 2013, 05:29:29 AM
 #7


Yes, if you're using molex to PCIe 6/8 pin adapters, then you should be putting the adapters on different strings to make sure they're on different rails.

Tried using different strings but the same 75amps is being pulled,which means more power is being burned up right? I guess my last option is to buy a single rail PSU.
crazyates
Legendary
*
Offline Offline

Activity: 952



View Profile
January 21, 2013, 05:43:37 AM
 #8


Yes, if you're using molex to PCIe 6/8 pin adapters, then you should be putting the adapters on different strings to make sure they're on different rails.
Tried using different strings but the same 75amps is being pulled,which means more power is being burned up right? I guess my last option is to buy a single rail PSU.
I already explained that 75A at 1.1V is well within tolerances. At stock voltage and 1100MHz, GPUz says my 7970 VDDC current is between 150-160A.

Tips? 1crazy8pMqgwJ7tX7ZPZmyPwFbc6xZKM9
Previous Trade History - Sale Thread
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 21, 2013, 05:55:38 AM
 #9

I already explained that 75A at 1.1V is well within tolerances. At stock voltage and 1100MHz, GPUz says my 7970 VDDC current is between 150-160A.

Well compared to a 5970 powered by an single rail 850w drawing about 50A, wouldn't mine drawing 75A be sucking more power from the wall resulting in a increased power bill?
bcpokey
Hero Member
*****
Offline Offline

Activity: 602



View Profile
January 21, 2013, 08:39:50 AM
 #10

I already explained that 75A at 1.1V is well within tolerances. At stock voltage and 1100MHz, GPUz says my 7970 VDDC current is between 150-160A.

Well compared to a 5970 powered by an single rail 850w drawing about 50A, wouldn't mine drawing 75A be sucking more power from the wall resulting in a increased power bill?


Dude please read what he has been telling you, he was pretty clear.

Your PSU offers 12V power to your GPU. Internally your GPU transforms the voltage and feeds it to various components that would likely explode at 12V or some such.

PSUs offer 3 voltages depending on the line, 12V, 5V and 3.3V, that's it. 5970s require PCI-E connectors, which are 12V (or in your case Molex to PCI-E, which is still 12V). If your card were using 75A of 12V power, it would melt, and your Mobo would melt, basically your computer would be dead.

Again, look at your GPU-Z, look at VDDC, and VDDC Current, they should say something like 1.0V, 75A. As the first response said, Watts = Volts x Amps; or 75A * 1.0V = 75W. I suggest you buy a kill-a-watts meter if this is still bothering you, to confirm you don't have some freak GPU pulling 900W.

Also maybe google the terms, amperage, voltage, and watts, to get a clearer sense of how they relate.
Gomeler
Hero Member
*****
Offline Offline

Activity: 698



View Profile
January 21, 2013, 10:20:38 PM
 #11

I already explained that 75A at 1.1V is well within tolerances. At stock voltage and 1100MHz, GPUz says my 7970 VDDC current is between 150-160A.

Well compared to a 5970 powered by an single rail 850w drawing about 50A, wouldn't mine drawing 75A be sucking more power from the wall resulting in a increased power bill?


Dude please read what he has been telling you, he was pretty clear.

Your PSU offers 12V power to your GPU. Internally your GPU transforms the voltage and feeds it to various components that would likely explode at 12V or some such.

PSUs offer 3 voltages depending on the line, 12V, 5V and 3.3V, that's it. 5970s require PCI-E connectors, which are 12V (or in your case Molex to PCI-E, which is still 12V). If your card were using 75A of 12V power, it would melt, and your Mobo would melt, basically your computer would be dead.

Again, look at your GPU-Z, look at VDDC, and VDDC Current, they should say something like 1.0V, 75A. As the first response said, Watts = Volts x Amps; or 75A * 1.0V = 75W. I suggest you buy a kill-a-watts meter if this is still bothering you, to confirm you don't have some freak GPU pulling 900W.

Also maybe google the terms, amperage, voltage, and watts, to get a clearer sense of how they relate.

This should be clear enough to the OP. If that isn't enough, I'm running 3 rigs with 3 watercooled 5970s @ stock core clocks, 1.0vddc and 500 MHz on the memory. I'm seeing around 700w pulled from the wall per rig for the three cards, fans, pumps, mobo, cpu, etc. You most definitely are not pulling 75 amps of +12vdc from your 5970. If that were the case you wouldn't have a 5970, you'd have a puddle of slag  Grin
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 22, 2013, 12:06:12 AM
 #12

Here is a GPUZ screen I took mining a couple days ago http://imgbox.com/abhSVwhp
crazyates
Legendary
*
Offline Offline

Activity: 952



View Profile
January 22, 2013, 03:47:31 AM
 #13

Here is a GPUZ screen I took mining a couple days ago

I don't see anything out of the ordinary in that picture. 76A @ 1.05V is perfectly normal. Remember, this is all internal, inside the card. The external Amps that the card draws from the PSU (@12v) is going to be different, and both of those numbers are going to be different than the Amps the PSU pulls from the wall (@120v).

The only hint I do see is that the the current drops off 5 times in that pic. Those 5 drops prolly coincide with a quick drop in "GPU Load" farther up on the GPU-z sensor scree, right? That's your mining software pausing for a quick fraction of a second when a new block is detected and new work is loaded. What mining software are you using?

Tips? 1crazy8pMqgwJ7tX7ZPZmyPwFbc6xZKM9
Previous Trade History - Sale Thread
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 22, 2013, 03:55:19 AM
 #14

I am currently using CGminer 2.10.4 and mining litecoin on notroll.in
zvs
Legendary
*
Offline Offline

Activity: 1484


I have some bitcoins. Somewhere.


View Profile WWW
January 22, 2013, 11:11:37 PM
 #15

I was always under the assumption that if you had your amps all jumping around like that, then there was some issue somewhere... I mean, the GPUs (most) of the time will still run unless it's really bad, but..

re:  http://imgbox.com/abhSVwhp


Dacentec, best deals for US dedicated servers. They regularly restock $20-$25 Opterons with 8-16GB RAM & 2x1-2TB HDD's (ofc, usually lots of other good stuff to choose from).  I did a Serverbear benchmark of one of my $20/mo Opteron (June last year), it's here.  Have had about a half dozen different servers with Dacentec, & none have failed to sustain at least 40MB/s (burst higher). My favorite is a 12-month rent-to-own ZT Systems 2XL5520 16GB 2x2TB SATA for $40/month (got lucky with the 'off-brand', haven't seen a RTO 2xL5520 for under $50/mo since -- at least for monthly contracts).  wholesaleinternet.com has some ancient 2-core intel CPUs @ $10/mo sometimes (I got an Intel Core 2 6300 @ 1.86GHz, with a 250GB HDD with 46000 hours on it, LOL. $20 @ Dacentec is much better, if you can grab one). joesdatacenter.com (same location as Wholesale Internet) also occasionally has specials (or if you don't want to wait, it has an AMD Opteron 170 @ $16/mo).
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 22, 2013, 11:29:36 PM
 #16

I was always under the assumption that if you had your amps all jumping around like that, then there was some issue somewhere... I mean, the GPUs (most) of the time will still run unless it's really bad, but..

re:  http://imgbox.com/abhSVwhp

Exactly what I thought, so I made this thread...
zvs
Legendary
*
Offline Offline

Activity: 1484


I have some bitcoins. Somewhere.


View Profile WWW
January 23, 2013, 12:17:36 AM
 #17

I was always under the assumption that if you had your amps all jumping around like that, then there was some issue somewhere... I mean, the GPUs (most) of the time will still run unless it's really bad, but..

re:  http://imgbox.com/abhSVwhp

Exactly what I thought, so I made this thread...
Well, I have one computer that has a Sempron that's so horrible that GPU usage jumps between 95-100 quite a bit, so the amps change too.

here are some of my 5970's

5970 #1  



5970 #2   (a card moderately in need of a new thermal paste job..  but, hey, it's winter and I won't be using it anymore in march or april, right?)



and just for kicks

5870



that one must have some mad caked up thermal paste,  not to mention a few holes wherever gpu 2 and 3 sensors are...  but it CAN get 1040 core at 1.1750,  and is even stable at 1080 @ 1.212.  it's like 98 amps then

well, at night that is.  usually.   these 65o days are bothersome






links fixed.  whee.  hmm.  just noticed GPU load 95% on card #2.  i guess that was probably from switching in to gpu-z with an elite sempron

Dacentec, best deals for US dedicated servers. They regularly restock $20-$25 Opterons with 8-16GB RAM & 2x1-2TB HDD's (ofc, usually lots of other good stuff to choose from).  I did a Serverbear benchmark of one of my $20/mo Opteron (June last year), it's here.  Have had about a half dozen different servers with Dacentec, & none have failed to sustain at least 40MB/s (burst higher). My favorite is a 12-month rent-to-own ZT Systems 2XL5520 16GB 2x2TB SATA for $40/month (got lucky with the 'off-brand', haven't seen a RTO 2xL5520 for under $50/mo since -- at least for monthly contracts).  wholesaleinternet.com has some ancient 2-core intel CPUs @ $10/mo sometimes (I got an Intel Core 2 6300 @ 1.86GHz, with a 250GB HDD with 46000 hours on it, LOL. $20 @ Dacentec is much better, if you can grab one). joesdatacenter.com (same location as Wholesale Internet) also occasionally has specials (or if you don't want to wait, it has an AMD Opteron 170 @ $16/mo).
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 23, 2013, 12:35:40 AM
 #18

I also have no idea why there is such a major temp difference between my VDDC Phase Temps, notice that VDDC Phase 1 is a good 20c cooler than VDDC phase 2&3
zvs
Legendary
*
Offline Offline

Activity: 1484


I have some bitcoins. Somewhere.


View Profile WWW
January 23, 2013, 12:49:03 AM
 #19

I also have no idea why there is such a major temp difference between my VDDC Phase Temps, notice that VDDC Phase 1 is a good 20c cooler than VDDC phase 2&3
i'm not sure about the VDDC's myself.  i think that's probably related to some thermal tape that's not on there very good or going bad or something.  i don't think i'd open it unless you had some to replace it, though.

my template for the GPU cores has always been, diff of less than 5o on all of them, perfect.  diff of less than 10o, you're ok.  diff of >10o, crappy thermal paste job.  diff of >15o, mad caking and flaking going on, diff of >20o, jesus, open that shit and fix it.

usually the case on those is that someone has applied too much thermal paste.  like there'll be flakes and big chunks of that crap everywhere

Dacentec, best deals for US dedicated servers. They regularly restock $20-$25 Opterons with 8-16GB RAM & 2x1-2TB HDD's (ofc, usually lots of other good stuff to choose from).  I did a Serverbear benchmark of one of my $20/mo Opteron (June last year), it's here.  Have had about a half dozen different servers with Dacentec, & none have failed to sustain at least 40MB/s (burst higher). My favorite is a 12-month rent-to-own ZT Systems 2XL5520 16GB 2x2TB SATA for $40/month (got lucky with the 'off-brand', haven't seen a RTO 2xL5520 for under $50/mo since -- at least for monthly contracts).  wholesaleinternet.com has some ancient 2-core intel CPUs @ $10/mo sometimes (I got an Intel Core 2 6300 @ 1.86GHz, with a 250GB HDD with 46000 hours on it, LOL. $20 @ Dacentec is much better, if you can grab one). joesdatacenter.com (same location as Wholesale Internet) also occasionally has specials (or if you don't want to wait, it has an AMD Opteron 170 @ $16/mo).
bcbigbud420
Full Member
***
Offline Offline

Activity: 172



View Profile
January 27, 2013, 08:42:54 AM
 #20

i'm not sure about the VDDC's myself.  i think that's probably related to some thermal tape that's not on there very good or going bad or something.  i don't think i'd open it unless you had some to replace it, though.

my template for the GPU cores has always been, diff of less than 5o on all of them, perfect.  diff of less than 10o, you're ok.  diff of >10o, crappy thermal paste job.  diff of >15o, mad caking and flaking going on, diff of >20o, jesus, open that shit and fix it.

usually the case on those is that someone has applied too much thermal paste.  like there'll be flakes and big chunks of that crap everywhere

I thought the exact same thing, or someone forgot to put some thermal tape on! but I have looked high and low and could not find a single person who has the same temp difference as I do, also I worry about these temps, being in a 20c ambient temp room with the window open, in the winter which results in these temps leaves me to believe I'm gonna have a hell of a time with  temps in +30c to +40c come summer time  Undecided
Pages: [1]
  Print  
 
Jump to:  

Sponsored by , a Bitcoin-accepting VPN.
Powered by MySQL Powered by PHP Powered by SMF 1.1.19 | SMF © 2006-2009, Simple Machines Valid XHTML 1.0! Valid CSS!