Gigabyte Super OC GTX 480's - VERY LOW PRICES!

Cleeecooo · 7 May 2011 at 11:03

DTWD said:
Do you think I should avoid calling back then?

no

. You may end up waiting for much longer

DTWD · 7 May 2011 at 11:12

Hmm what to do.

I could wait and GB send a 580 to OCuk and then someone "liberates" it and sends me a 480 from stock. I would be none the wiser, but time would have passed.

GB could send OCuk a 480 anyway, and I have just wasted time.

Or I just call and get a replacement 480...

Decisions decisions...

DTWD · 7 May 2011 at 11:14

All I am thinking is I have an Xbox in the house so I am not at a loss for games and I still have my old 9800 so I am not without the use of the PC...

DTWD · 7 May 2011 at 11:16

DTWD said:
Hmm what to do.

I could wait and GB send a 580 to OCuk and then someone "liberates" it and sends me a 480 from stock. I would be none the wiser, but time would have passed.

GB could send OCuk a 480 anyway, and I have just wasted time.

Or I just call and get a replacement 480...

Decisions decisions...

Of course another option is that OCuk recieve a 580 and send it my way, I forgot the honest option

ToucHDowN · 7 May 2011 at 11:42

Ok, news update. I got my replacement gtx480soc. It's the same revision (1.1) and batch number (s/n: 1045xxxxxxxx) as the first, faulty one. Plugged it in, PC boots up ok. GPU-Z correctly identifies clock, memory, shaders. Idle temp: 52oC. Drivers: 270.61. Stock voltage according to MSI Afterburner: 1.038.

(Writing as I test)

1st Test: Stock voltage (1.038v)
Running Unigine Heaven maxxed out: Artifacted

Running OCCT v3.1.0: 250.000 errors in the first 25sec.

Running OCCT v4.0.0.b5: 204.000 errors in the first 25sec.

I feel angry, but I'd like to increase the voltage, just in case...

2nd Test: Voltage (1.050v)
Running Unigine Heaven maxxed out: Artifacted.

Running OCCT v3.1.0: 42.000 errors in the first 25sec. :confused:

Running OCCT v4.0.0.b5: 44.000 errors in the first 25sec. :confused:

????

3rd Test: Voltage (1.075v)
Running Unigine Heaven maxxed out: Not artifacted!

Running OCCT v3.1.0: 3 (!!!) errors in the first 25sec. :rolleyes:

Running OCCT v4.0.0.b5: 0 (!!!) errors in the first 25sec. :rolleyes:

4th Test: Voltage (1.088v)
Running Unigine Heaven maxxed out: Not artifacted!

Running OCCT v3.1.0: 0 (!!!) errors in the first 120sec.

Running OCCT v4.0.0.b5: 0 (!!!) errors in the first 120sec.

Well...

I ran the above set of tests 5 times in various configurations. All results reproducible, meaning that my cure is 1.088v.

Feeling happy, don't know though if i must RMA or not in this case.

DTWD · 7 May 2011 at 11:47

What temps are you getting over those 120 seconds in OCCT?

Run it for 10 minutes + and see what happens. 120 seconds is barely enough time to get the card warmed up. My problems came in quite quickly in OCCT but only really noticable after playing a game for 30 minutes, then they tended to go a bit mad.

Cleeecooo · 7 May 2011 at 12:12

ToucHDowN said:
Ok, news update. I got my replacement gtx480soc. It's the same revision (1.1) and batch number (s/n: 1045xxxxxxxx) as the first, faulty one. Plugged it in, PC boots up ok. GPU-Z correctly identifies clock, memory, shaders. Idle temp: 52oC. Drivers: 270.61. Stock voltage according to MSI Afterburner: 1.038.

(Writing as I test)

1st Test: Stock voltage (1.038v)
Running Unigine Heaven maxxed out: Artifacted
Running OCCT v3.1.0: 250.000 errors in the first 25sec.
Running OCCT v4.0.0.b5: 204.000 errors in the first 25sec.

I feel angry, but I'd like to increase the voltage, just in case...

2nd Test: Voltage (1.050v)
Running Unigine Heaven maxxed out: Artifacted.
Running OCCT v3.1.0: 42.000 errors in the first 25sec.
Running OCCT v4.0.0.b5: 44.000 errors in the first 25sec.

????

3rd Test: Voltage (1.075v)
Running Unigine Heaven maxxed out: Not artifacted!
Running OCCT v3.1.0: 3 (!!!) errors in the first 25sec.
Running OCCT v4.0.0.b5: 0 (!!!) errors in the first 25sec.

4th Test: Voltage (1.088v)
Running Unigine Heaven maxxed out: Not artifacted!
Running OCCT v3.1.0: 0 (!!!) errors in the first 120sec.
Running OCCT v4.0.0.b5: 0 (!!!) errors in the first 120sec.

Well... I ran the above set of tests 5 times in various configurations. All results reproducible, meaning that my cure is 1.088v.

Feeling happy, don't know though if i must RMA or not in this case.

If you are happy with the fact you have a stable card after a bit of playing around, then keep it. otherwise, just RMA

whipgeez · 7 May 2011 at 12:20

20 mins of occt will give a good impression, 120 secs isn't very long at all

20mins to be on the safe side i'd say

Cleeecooo · 7 May 2011 at 12:26

whipgeez said:
20 mins of occt will give a good impression, 120 secs isn't very long at all

20mins to be on the safe side i'd say

10 mins is enough but 1hr would be the best

Buckster · 7 May 2011 at 15:01

re-looked through the 20 screenshots GB sent mewhen I RMA'd card

interested to see they did tests in LN bios -as shown by their GPU-Z screenshot !

still-nomore stable for me

Cleeecooo · 7 May 2011 at 15:05

Buckster said:
re-looked through the 20 screenshots GB sent mewhen I RMA'd card

interested to see they did tests in LN bios -as shown by their GPU-Z screenshot !

still-nomore stable for me

is this with a new PSU?

ToucHDowN · 7 May 2011 at 17:21

whipgeez said:
20 mins of occt will give a good impression, 120 secs isn't very long at all

20mins to be on the safe side i'd say

Well, I don't benchmarks for living nor I'm a professional reviewer. This is the first time in one month time and after using 2 gtx480 soc cards that I get 120sec completely free of errors in OCCT whereas I was getting 300.000 of them in the first 20 seconds. So 120sec is enough for me.

Slow-cooking my graphics card for 30min or 60min at >95oC is something I do not consider to do, in any case. I will do some gaming though tonight to see if I indeed solved my problems and see how it goes.

ToucHDowN · 7 May 2011 at 17:25

DTWD said:
What temps are you getting over those 120 seconds in OCCT?

Run it for 10 minutes + and see what happens. 120 seconds is barely enough time to get the card warmed up. My problems came in quite quickly in OCCT but only really noticable after playing a game for 30 minutes, then they tended to go a bit mad.

At 120sec or so, the card has already exceeded 90oC and rising. Don't intend to stress it more. Fan speed reaches 99% rpm. With my previous rma'ed gtx480 soc, when OCCT was giving errors in the first seconds, I knew I'd had artifacts and freezes soon enough within Crysis 2 or Battlefield Bad Company 2. As I said, I'll do some gaming tonight before I make any final conclusions about RMA #2 or refund.

whipgeez · 7 May 2011 at 17:27

i wouldn't worry, occt won't cook your cards. it's no different from playing crysis for example where your gpu gets 99% usage,

whipgeez · 7 May 2011 at 17:29

ToucHDowN said:
At 120sec or so, the card has already exceeded 90oC and rising. Don't intend to stress it more. Fan speed reaches 99% rpm. With my previous rma'ed gtx480 soc, when OCCT was giving errors in the first seconds, I knew I'd had artifacts and freezes soon enough within Crysis 2 or Battlefield Bad Company 2. As I said, I'll do some gaming tonight before I make any final conclusions about RMA #2 or refund.

thats hot for 99% fan speed occt is just like a graphically intense game being put on the gpu so it's pretty colse to what temps your get in games, unlike furmark which somehow raises temps even higher

Buckster · 7 May 2011 at 17:40

right an update-I solved my connector issue but was difficult (no matter what I did I could not get one of the power connectors off off the card from my old PSU- in the end with a few tools (non destructive thankfully) and help from a colleague in our labs we managed to remove it yesterday

then ordered a HX850 - which has now been installed today and connected up

I'd love to now say a big improvement - but no - practically zip difference - possibly minorly less artefacts than before at stock, but still loads in Heaven at stock volts, and OCCT plastered with them - Furmark too

so now down another 100+ to no avail

I categorically will state I personally will NEVER EVER buy another Gigabyte product now - fingers well and truly burnt -and in fact I'll prob avoid all pre-clocked cards from all manuf. in the future

how they tested my card stable - I have no idea - perhaps theres some weird interaction between Gigabyte mobos and Gigabyte GPUs (perhaps volts on the PCI-E are higher or something?)

I'm going to be taking this further as its just ridiculous

note my cardpassed all of GBs tests - so I just don't understand it

my mobo is PCI-E 1.1 but that shouldn't matter as they are backwardly compatible

I think number of people with problems on this thread (ok peoplewith problems are more vocal than those with no problems) - and the fact that GB have accepted at least 2 RMAs from people on this thread is testament to some fundamental problem somewhere

I'm going to try and talk with someone as senior as I can within GB RMA early next week as not happy at all

With PSU out of way - onlything remaining is my mobo - but since that was stable with my 4890 - and highlyoverclocked GTX570 (30 mins OCCT stable) - I am at a loss

conversely the Corsair HX850 is a lovely product - quiet as a mouse - fan noise wise and no "whine" under load at all - highly highly recommended

Cleeecooo · 7 May 2011 at 17:42

Buckster said:
right an update-I solved my connector issue but was difficult (no matter what I did I could not get one of the power connectors off off the card from my old PSU- in the end with a few tools (non destructive thankfully) and help from a colleague in our labs we managed to remove it yesterday

then ordered a HX850 - which has now been installed today and connected up

I'd love to now say a big improvement - but no - practically zip difference - possibly minorly less artefacts than before at stock, but still loads in Heaven at stock volts, and OCCT plastered with them - Furmark too

so now down another 100+ to no avail

I categorically will state I will NEVER EVER buy another Gigabyte product now - fingers well and truly burnt

how they tested my card stable - I have no idea - perhaps theres some weird interaction between Gigabyte mobos and Gigabyte GPUs (perhaps volts on the PCI-E are higher or something?)

I'm going to be taking this further as its just ridiculous

note my cardpassed all of GBs tests - so I just don't understand it

my mobo is PCI-E 1.1 but that shouldn't matter as they are backwardly compatible

I think number of people with problems on this thread (ok peoplewith problems are more vocal than those with no problems) - and the fact that GB have accepted at least 2 RMAs from people on this thread is testament to some fundamental problem somewhere

I'm going to try and talk with someone as senior as I can within GB RMA early next week as not happy at all

conversely the Corsair HX850 is a lovely product - quiet as a mouse - fan noise wise and no "whine" under load at all - highly highly recommended

why not ask to RMA to them again? It's not like they're going to decline?

whipgeez · 7 May 2011 at 17:44

Cleeecooo said:
why not ask to RMA to them again? It's not like they're going to decline?

sounds like this was a rejected rma and sent back

Buckster · 7 May 2011 at 17:44

I will but I'll be charged another 30-40 odd if it passes their tests

and why would it be any different this time when itpassed last time ?

tempted to pay for it to be independantly tested and then take it forward formally somehow

Cleeecooo · 7 May 2011 at 17:46

Buckster said:
I will but I'll be charged another 30-40 odd if it passes their tests

and why would it be any different this time when itpassed last time ?

tempted to pay for it to be independantly tested and then take it forward formally somehow

How bout RMA'ing to OcUK? only £20 including delivery back to you if it passes

Competitor rules

**Gigabyte Super OC GTX 480's - VERY LOW PRICES!**

Gigabyte Super OC GTX 480's - VERY LOW PRICES!