Tom's Hardware Superposition Benchmark Thread

[SoNic67]

"11. Video cards that cannot run 1080P Extreme due to memory limits can be run at 1080p Medium, where 35% of that score will be used."
Is that a precise number?

 

[EquineHero]

"11. Video cards that cannot run 1080P Extreme due to memory limits can be run at 1080p Medium, where 35% of that score will be used."
Is that a precise number?

Well, basically, yes. A GTX 580 should pperform equally with a reference GTX 1050Ti in terms of raw power. The 1050Ti does about 1400 in Superposition 1080p Extreme, and running Superposition at 1080p medium on both the GTX 580 and the 1050Ti yeilds about the same score.

Using this method of testing I've determined that this is fairly accurate so I appended this rule so people can add their scores even when their cards cannot effectively run the 1080p Extreme tests due to memory limitations.

The 900-ish score I have for the 580 is an actual 1080p Extreme score so I left it alone.

 

[SoNic67]

For my RX580 "extreme" is only 30% of "medium", see below. Applying the rule 11 to my medium result, I would get higher benchmark on extreme (2951).
I think there should be either two distinct categories or... just one (like Medium), no conversions of scores.
PS: It is really a RX580 (like I have shown in the GPU-Z above) but the test software didn't get updated to recognize this (basically a refreshed RX480).

 

[fagetti]

Finally i can run my gtx 670 windforce 2gb at Pcie 2.0 x16 slot. Here are the results.
Its not really that big of a boost i expected. Result with pcie 1.1 x4 slot was 5896 for medium superposition.



35% of this score would be 2136 so im already beating stock GTX 970 by few points. Im going to volt mod this card to 1.3v, lets see if i can beat Gtx 1060 stock.

 

[EquineHero]

Finally i can run my gtx 670 windforce 2gb at Pcie 2.0 x16 slot. Here are the results.
Its not really that big of a boost i expected. Result with pcie 1.1 x4 slot was 5896 for medium superposition.



35% of this score would be 2136 so im already beating stock GTX 970 by few points. Im going to volt mod this card to 1.3v, lets see if i can beat Gtx 1060 stock.

My R9 270x 2GB ran 1080p Extreme just fine, give 1080p Extreme a go. I'll add the score when you do.

 

[fagetti]

I have no idea why for my 670 2gb windforce 1080p extreme is a lagfest. I was getting results from 1000 to 1150. Maybe my card is not utilizing all the vram it can or your card has better vram chips overall.
However superposition tells that 1080 high and 1080 extreme modes require same amount of vram, but i still tested 1080 high setting and looks like its the sweet spot im getting decent 20-30fps.

 

[EquineHero]

I have no idea why for my 670 2gb windforce 1080p extreme is a lagfest. I was getting results from 1000 to 1150. Maybe my card is not utilizing all the vram it can or your card has better vram chips overall.
However superposition tells that 1080 high and 1080 extreme modes require same amount of vram, but i still tested 1080 high setting and looks like its the sweet spot im getting decent 20-30fps.

Raise your DRAM speed, raise the memory clocks, and disable your pagefile. Then try 1080p Extreme. It will tell you you need more vRAM, ignore that. My GTX 580 ran Superposition and got a lower score (900s) so a score of 1100 makes perfect sense.

 

[andrewtsang]

Here's mine, running 8700k, 32GB ram, Zotac 1080TI Amp Extreme

]http://[/url]

 

[pete_101]

I was running a couple of benchmarks for my own records before upgrading my old R9 290X, these benchmarks included Superposition. Looking at the result chart, I'm surprised not to see any other R9 290X and that my score beats the RX 580 and is about equal with the GTX 1060 6gb.

Anyway, here's what my old system with it's 290X toaster scores.

 

[EquineHero]

I was running a couple of benchmarks for my own records before upgrading my old R9 290X, these benchmarks included Superposition. Looking at the result chart, I'm surprised not to see any other R9 290X and that my score beats the RX 580 and is about equal with the GTX 1060 6gb.

Anyway, here's what my old system with it's 290X toaster scores.

Yeah, AMD's lack of performance improvement in consumer cards really shows. A single 1070 beats out all of AMD's flagships except the Vega64. Which is why stores get raided for AMD cards, because the only thing AMD cards are good for is mining. Lol.

 

[Rogue Leader]

I was running a couple of benchmarks for my own records before upgrading my old R9 290X, these benchmarks included Superposition. Looking at the result chart, I'm surprised not to see any other R9 290X and that my score beats the RX 580 and is about equal with the GTX 1060 6gb.

Anyway, here's what my old system with it's 290X toaster scores.

Yeah, AMD's lack of performance improvement in consumer cards really shows. A single 1070 beats out all of AMD's flagships except the Vega64. Which is why stores get raided for AMD cards, because the only thing AMD cards are good for is mining. Lol.

The Vega 56 is competitive with the 1070, in some cases its faster. However IF you can find one its twice the price. The 1070ti however beats it, and is often cheaper as well.

 

[pete_101]

I was running a couple of benchmarks for my own records before upgrading my old R9 290X, these benchmarks included Superposition. Looking at the result chart, I'm surprised not to see any other R9 290X and that my score beats the RX 580 and is about equal with the GTX 1060 6gb.

Anyway, here's what my old system with it's 290X toaster scores.

Yeah, AMD's lack of performance improvement in consumer cards really shows. A single 1070 beats out all of AMD's flagships except the Vega64. Which is why stores get raided for AMD cards, because the only thing AMD cards are good for is mining. Lol.

The Vega 56 is competitive with the 1070, in some cases its faster. However IF you can find one its twice the price. The 1070ti however beats it, and is often cheaper as well.

What's interesting is I just re-tested with my new card, I got a used EVGA 1060 6GB SC very cheap, so that's my upgrade. My benchmark scores are up about 20% for Heaven and Valley, but are actually worse for Superposition, so this might be one benchmark that runs better on AMD.

As for my old R9 290X, my brother wants to buy it off me to replace his Radeon HD 6870 crossfire setup, so it's a win/win situation. He's getting an upgrade, I'm getting a slight upgrade but one that's much newer and uses far less power and is much quieter and it's only costing me $100 to change over. Anyway, he's got a 1500w PSU where as my 600w unit can now take it a bit easier and I can overclock my CPU a bit.

 

[EquineHero]

What's interesting is I just re-tested with my new card, I got a used EVGA 1060 6GB SC very cheap, so that's my upgrade. My benchmark scores are up about 20% for Heaven and Valley, but are actually worse for Superposition, so this might be one benchmark that runs better on AMD.

As for my old R9 290X, my brother wants to buy it off me to replace his Radeon HD 6870 crossfire setup, so it's a win/win situation. He's getting an upgrade, I'm getting a slight upgrade but one that's much newer and uses far less power and is much quieter and it's only costing me $100 to change over. Anyway, he's got a 1500w PSU where as my 600w unit can now take it a bit easier and I can overclock my CPU a bit.

Eh, my older CX600 can push 800W if it wants. 600W is quite a lot.

 

[Rogue Leader]

What's interesting is I just re-tested with my new card, I got a used EVGA 1060 6GB SC very cheap, so that's my upgrade. My benchmark scores are up about 20% for Heaven and Valley, but are actually worse for Superposition, so this might be one benchmark that runs better on AMD.

As for my old R9 290X, my brother wants to buy it off me to replace his Radeon HD 6870 crossfire setup, so it's a win/win situation. He's getting an upgrade, I'm getting a slight upgrade but one that's much newer and uses far less power and is much quieter and it's only costing me $100 to change over. Anyway, he's got a 1500w PSU where as my 600w unit can now take it a bit easier and I can overclock my CPU a bit.

Eh, my older CX600 can push 800W if it wants. 600W is quite a lot.

Pushing 800w from a CX600 (no way in hell, OCP would cut you off well below 800, but for arguments sake) is an easy way to fry all of your equipment. At a minimum the CX600 would immolate itself.

In any event you are correct, 600w is way more than he could ever need for a 1060 6gb and his CPU even overclocked to the moon.

 

[loki1944]

With my 6850K/1080Ti SLI system:

[/url][/img]

 

[TechyInAZ]

wow!

 

[EquineHero]

With my 6850K/1080Ti SLI system:

*incoherent swearing*

 

[aquielisunari]

With my 6850K/1080Ti SLI system:

*incoherent swearing*

Your pic says it all. You're looking at Loki's score.

 

[EquineHero]

With my 6850K/1080Ti SLI system:

*incoherent swearing*

Your pic says it all. You're looking at Loki's score.

Yeah...I'll just have to buy another GTX 1070 for unofficially supported triple 1070 SLI!!!1!1!

 

[TechyInAZ]

You'd need 4 way 1070 SLI to beat 2 highly overclocked GTX 1080 TIs in SLI.

 

[loki1944]

With my 6850K/1080Ti SLI system:

*incoherent swearing*

Your pic says it all. You're looking at Loki's score.

Yeah...I'll just have to buy another GTX 1070 for unofficially supported triple 1070 SLI!!!1!1!

Pretty sure if you had 2x1080Tis you would smoke that with a 5.1Ghz 8700K.

 

[SverreMunthe]

Have a score of 12312, with 2 GPU's. No idea how that compares, think it's rated as 13, but can't choose other than 1 GPU on the leader board. Have shared it in this post, from FB, but I don't see it. Tried Google Drive earlier, with same result.

 

[EquineHero]

With my 6850K/1080Ti SLI system:

*incoherent swearing*

Your pic says it all. You're looking at Loki's score.

Yeah...I'll just have to buy another GTX 1070 for unofficially supported triple 1070 SLI!!!1!1!

Pretty sure if you had 2x1080Tis you would smoke that with a 5.1Ghz 8700K.

My 4790K outperforms two 7700Ks here, I could definitely do triple 1070 SLI.

Have a score of 12312, with 2 GPU's. No idea how that compares, think it's rated as 13, but can't choose other than 1 GPU on the leader board. Have shared it in this post, from FB, but I don't see it. Tried Google Drive earlier, with same result.

Upload the photo to Imgur.

 

[SverreMunthe]

How much does the CPU influence the score?

 

[EquineHero]

How much does the CPU influence the score?

It shouldn't, it's a GPU benchmark.

You rich people and your money. Pah!