Stress testing with bottleneck - fair assessment?

[STORMComponents]

I've recently come into a bunch of high-end graphics cards recently, but I need to test them before I can re-sell. I don't have a spare test rig, and don't really want to go through 25+ cards and drivers in my home rig. I'm looking to spend a little money on a test rig, but wanted to know if a furmark and 3dmark test while the GPU is blatantly bottlenecked (single / dual core, DDR2 ram) would give a fair test?

I'm checking to see if the cards are stable or faulty etc, but some are as high as GTX 580s or 7970s and I don't know if using a low-grade testing rig would actually tell me if they're stable or not.

Can anyone advise?

 

[Sangeet Khatri]

Can you tell me how much are you willing to spend on a test rig. It should do just fine when with a i5 CPU. But you would need a good PSU for that.

 

[STORMComponents]

I'm not really sure. If possible I'd like to utilise components that I have, which are AMD dual cores (maybe a quad 9500) and some DDR2 RAM. I just wondered if a furmark test under those conditions would push the card as far as if it were in a full gaming system.

 

[Sangeet Khatri]

Anyways, this is a good 500 dollar test bench for testing and it would not bottleneck any of your cards that you have.

PCPartPicker part list / Price breakdown by merchant / Benchmarks

CPU: Intel Core i5-3470 3.2GHz Quad-Core Processor ($179.99 @ Amazon)
Motherboard: Gigabyte GA-B75M-HD3 Micro ATX LGA1155 Motherboard ($62.99 @ Newegg)
Memory: Patriot Viper 3 8GB (2 x 4GB) DDR3-1600 Memory ($54.99 @ Newegg)
Storage: Western Digital Caviar Blue 1TB 3.5" 7200RPM Internal Hard Drive ($59.98 @ Outlet PC)
Case: DIYPC Alpha-GT3 ATX Test Bench Case ($59.99 @ Newegg)
Power Supply: OCZ Fatal1ty 750W 80 PLUS Bronze Certified ATX12V / EPS12V Power Supply ($55.00 @ Newegg)
Optical Drive: Asus DRW-24B1ST/BLK/B/AS DVD/CD Writer ($17.98 @ Outlet PC)
Total: $490.92
(Prices include shipping, taxes, and discounts when available.)
(Generated by PCPartPicker 2013-08-08 09:03 EDT-0400)

I hope this helps.

 

[Sangeet Khatri]

Forget stressing that GPU while using the DDR2 RAM and the old AMD CPU. There would be a big bottleneck with those components.

 

[Sangeet Khatri]

If you want to go a bit more cheaper, then look at this test bench. It would also not bottleneck those GPU's

PCPartPicker part list / Price breakdown by merchant / Benchmarks

CPU: Intel Core i5-3350P 3.1GHz Quad-Core Processor ($169.99 @ Newegg)
Motherboard: ASRock B75M-DGS R2.0 Micro ATX LGA1155 Motherboard ($54.98 @ SuperBiiz)
Memory: Patriot Viper 3 8GB (2 x 4GB) DDR3-1600 Memory ($54.99 @ Newegg)
Storage: Hitachi 500GB 3.5" 7200RPM Internal Hard Drive ($49.99 @ Newegg)
Case: DIYPC Alpha-DB6 ATX Test Bench Case ($44.98 @ Newegg)
Power Supply: OCZ ModXStream Pro 700W 80 PLUS Certified ATX12V / EPS12V Power Supply ($46.50 @ Newegg)
Total: $421.43
(Prices include shipping, taxes, and discounts when available.)
(Generated by PCPartPicker 2013-08-08 09:12 EDT-0400)

 

[muzzpro]

you can only test if the gpu gives display or not otherwise everything will bottleneck for 3dmark tests and funmark (i don't advise using funmark software)
you NEED an i5 and 8gb ram and a good psu for testing out all the gpus.

the tests will not be fair at all.

 

[Sangeet Khatri]

This is much more ideal test bench for testing :

PCPartPicker part list / Price breakdown by merchant / Benchmarks

CPU: Intel Core i7-3770 3.4GHz Quad-Core Processor ($289.98 @ SuperBiiz)
Motherboard: ASRock B75M-DGS R2.0 Micro ATX LGA1155 Motherboard ($54.98 @ SuperBiiz)
Memory: Patriot Viper 3 8GB (2 x 4GB) DDR3-1600 Memory ($54.99 @ Newegg)
Storage: Hitachi 500GB 3.5" 7200RPM Internal Hard Drive ($49.99 @ Newegg)
Case: DIYPC Alpha-DB6 ATX Test Bench Case ($44.98 @ Newegg)
Power Supply: OCZ ModXStream Pro 700W 80 PLUS Certified ATX12V / EPS12V Power Supply ($46.50 @ Newegg)
Total: $541.42
(Prices include shipping, taxes, and discounts when available.)
(Generated by PCPartPicker 2013-08-08 09:22 EDT-0400)

 

[STORMComponents]

I'm not in the market to spend $500 on a testbench as that may be my entire profit margin on the cards (if they prove to be faulty).

I simply wanted to know if testing them with a bottleneck would show any faults, or if they need a full gaming system to properly test them.

Looks like I'll be taken them home and testing them on my home rig which is plenty fast enough to max the cards.

 

[Sangeet Khatri]

I'm not in the market to spend $500 on a testbench as that may be my entire profit margin on the cards (if they prove to be faulty).

I simply wanted to know if testing them with a bottleneck would show any faults, or if they need a full gaming system to properly test them.

Looks like I'll be taken them home and testing them on my home rig which is plenty fast enough to max the cards.

I am not sure. You should probably wait for someone else to answer that.

 

[muzzpro]

I'm not in the market to spend $500 on a testbench as that may be my entire profit margin on the cards (if they prove to be faulty).

I simply wanted to know if testing them with a bottleneck would show any faults, or if they need a full gaming system to properly test them.

Looks like I'll be taken them home and testing them on my home rig which is plenty fast enough to max the cards.

yes, take them home to test the cards because on the other rigs it will be extremely inadequate.

 

[muzzpro]

Can you tell me how much are you willing to spend on a test rig. It should do just fine when with a i5 CPU. But you would need a good PSU for that.

READ what he has posted, he is not willing to spend at all on a expensive rig

 

[STORMComponents]

I think I'll just take them home so I can guarantee a fair test. I suppose unless the cards are maxed out, there's no way to know if faults, overheating or artifacts come into play at high stress. If they're bottlenecked down to 50% or something, it may never fail when being pushed.

Thanks for your replies anyway! Fingers crossed the cards are in good condition!

 

[Sangeet Khatri]

Can you tell me how much are you willing to spend on a test rig. It should do just fine when with a i5 CPU. But you would need a good PSU for that.

READ what he has posted, he is not willing to spend at all on a expensive rig

I posted this before I was knowing this. Why are you talking about an older post which was posted at the time I was not aware about this?

 

[Sangeet Khatri]

I think I'll just take them home so I can guarantee a fair test. I suppose unless the cards are maxed out, there's no way to know if faults, overheating or artifacts come into play at high stress. If they're bottlenecked down to 50% or something, it may never fail when being pushed.

Thanks for your replies anyway! Fingers crossed the cards are in good condition!

Do you have any friend who has a good rig or someone who is an enthusiast? I mean this is when friends come handy. Isn't it?

And if he is enthusiast then I am sure he would love to help. Better find some friend who can help you.

 

[STORMComponents]

Do you have any friend who has a good rig or someone who is an enthusiast?

I can use my home rig, I don't need to find a better one. I have a 6 core at 4.0GHz and 16GB of 1600MHz RAM. I just didn't want to be using a home machine as a donkey test box. Looks like that's what's going to happen though. I might run them all in linux or something so I don't mess about with my own drivers or settings!

Thanks anyway.

 

[Sangeet Khatri]

Do you have any friend who has a good rig or someone who is an enthusiast?

I can use my home rig, I don't need to find a better one. I have a 6 core at 4.0GHz and 16GB of 1600MHz RAM. I just didn't want to be using a home machine as a donkey test box. Looks like that's what's going to happen though. I might run them all in linux or something so I don't mess about with my own drivers or settings!

Thanks anyway.

Just get a new 50 dollar HDD and install a new OS on it and run benchmarks on it. It would be better and would not at all mess your settings at all. So just get a new 7200RPM HDD.

 

[muzzpro]

Do you have any friend who has a good rig or someone who is an enthusiast?

I can use my home rig, I don't need to find a better one. I have a 6 core at 4.0GHz and 16GB of 1600MHz RAM. I just didn't want to be using a home machine as a donkey test box. Looks like that's what's going to happen though. I might run them all in linux or something so I don't mess about with my own drivers or settings!

Thanks anyway.

Just get a new 50 dollar HDD and install a new OS on it and run benchmarks on it. It would be better and would not at all mess your settings at all. So just get a new 7200RPM HDD.

from the very first post you posted something which was not required
@OP yes if it even is a fx6100, it will do good.

 

[STORMComponents]

@OP yes if it even is a fx6100, it will do good.

So you think stress testing on a POS platform would still show issues? We're talking Athlon X2 with 2GB RAM or something similar.
I know for a thorough test I should benchmark for 2-3 hours on a high-end system, but doing that on 30 different cards on my home rig will be rather disruptive.

 

[muzzpro]

@OP yes if it even is a fx6100, it will do good.

So you think stress testing on a POS platform would still show issues? We're talking Athlon X2 with 2GB RAM or something similar.
I know for a thorough test I should benchmark for 2-3 hours on a high-end system, but doing that on 30 different cards on my home rig will be rather disruptive.

your 3d mark scores will show only 1/6 of the performance those high end cards have.
an i5 or fx6350 is recommended for something like a hd7970

I am POSITIVE that your rig will immensely bottleneck these cards.

 

[smeezekitty]

Even a low end CPU won't bottleneck a GPU heavy benchmark.

 

[STORMComponents]

@OP yes if it even is a fx6100, it will do good.

So you think stress testing on a POS platform would still show issues? We're talking Athlon X2 with 2GB RAM or something similar.
I know for a thorough test I should benchmark for 2-3 hours on a high-end system, but doing that on 30 different cards on my home rig will be rather disruptive.

your 3d mark scores will show only 1/6 of the performance those high end cards have.
an i5 or fx6350 is recommended for something like a hd7970

I am POSITIVE that your rig will immensely bottleneck these cards.

I never said I cared about the scores.
I need to test the cards to see if they're faulty or not, and simply want to know if a bottlenecked system would still fairly test if the cards are faulty or functional under standard testing.