Tom's Hardware Superposition Benchmark Thread

[EquineHero]

This is why his score was not added. Poor overall attitude and complete disregard to proof against his claim.

1. his card was thermal throttling. This is extremely obvious.
2. If his claim was true, my 1070 wouldn't be at 2100Mhz throughout the benchmark.
3. The rules of this forum prevent me from expressing how I truly feel towards this individual.

Just to play devils advocate here EH

While attitude may not be on point, it doesn't violate any rules of the forum I see, nor does it violate any rules you've put in place for this thread.

Whether they agree or disagree with the benchmark itself is irrelevant - they wish to participate here.

You certainly could cite the fact they didn't adhere to:

2. Post the screenshot, with both GPUZ and the Superposition window in the same image.

Might want to keep personal feelings out of it.

Yeah if you read my reply instead of reacting emotionally you'd know that I went to pichost and it was super blurry.

Agreed, I did the same and it was - maybe you need to log in, I don't know. I know I wasn't logging in to find out.

Yep, that's HWInfo and not GPU-Z

 

[QwerkyPengwen]

There are many things I could easily say about you and the way you handle this thread.

One thing I will say is that you were clearly offended by my original post where I was lightly venting about some things. And instead of maybe communicating such things and providing some feedback you just stayed silent and ignored it.

My attitude wasn't poor, not in the slightest. You must have just misinterpreted my intentions.

But again, that's neither here nor there. That post was forever ago and I've moved on from that GPU and caring about the score.

At the end of the day things are what they are. My apologies if you've misinterpreted anything I've said so far.
Conversation and communication generally becomes degraded quite a bit when in text form since there's no way for anyone to feel the context in which things are said.

You have my latest submission. I pushed my current card to it's limits without shunt modding and throwing it into a custom loop and that's what I got for my score.

If my submission isn't too your liking for reasons that make it not fit within your criteria in your original post then please feel free to let me know what needs to change. I did use HWinfo instead of GPU-Z and that's simply because I don't like GPU-Z as a program and find HWinfo to be a more accurate tool for real time monitoring and information.

And I'm not going to go through the process of running superposition multiple times with those overclocks again just to get another score that's identical and then take a screenshot with GPU-Z. Don't see how GPU-Z vs HWinfo makes any difference. Unless there's a very specific reason you want to see GPU-Z in the picture that you failed to state in your original post. Otherwise, it's just to show that I am running the card I say I am and I can do that with HWinfo just fine.

 

[EquineHero]

did use HWinfo instead of GPU-Z and that's simply because I don't like GPU-Z as a program and find HWinfo to be a more accurate tool for real time monitoring and information.

Rules are rules. Show GPU-Z or your submission will not be added.

As for how I handle the thread, it is my thread at the end of the day, and as long as it conforms to the site rules I can add or prevent submissions at my discretion. Tom's is a fairly open and welcoming forum, if you don't like the way I operate my thread (which is still moderated by The Big Guys™) you're probably more than welcome to start a new thread with a different benchmark suite of your choosing.

I am the [strike]keeper[/strike] librarian of this thread and the Passmark thread over in the CPUs forum, which has its own similar rules. Passmark itself is a flawed benchmark, as is Unigine Heaven, Firestrike, Timespy, etc. All synthetic benchmarks are flawed. I can't prevent you from not liking a certain benchmark title, but if you want your score added to the leaderboard, I need the score and a GPU-Z window.

Other submissions do not have this as the rule was added after someone photo-shopped their results and their submissions were grandfathered in.

The only results I remove are from banned or no longer active users. I've had my own moderation history due to my personality type and my knack for strongly disagreeing with others. The Tom's Hardware moderators are very direct. I'm trying to stay here.

Once more so that it's clear:

Your submission must include a GPU-Z screenshot and the Superposition results window within the same frame to be added to the leaderboard. This is non negotiable.

 

[QwerkyPengwen]

did use HWinfo instead of GPU-Z and that's simply because I don't like GPU-Z as a program and find HWinfo to be a more accurate tool for real time monitoring and information.

Rules are rules. Show GPU-Z or your submission will not be added.

As for how I handle the thread, it is my thread at the end of the day, and as long as it conforms to the site rules I can add or prevent submissions at my discretion. Tom's is a fairly open and welcoming forum, if you don't like the way I operate my thread (which is still moderated by The Big Guys™) you're probably more than welcome to start a new thread with a different benchmark suite of your choosing.

I am the [strike]keeper[/strike] librarian of this thread and the Passmark thread over in the CPUs forum, which has its own similar rules. Passmark itself is a flawed benchmark, as is Unigine Heaven, Firestrike, Timespy, etc. All synthetic benchmarks are flawed. I can't prevent you from not liking a certain benchmark title, but if you want your score added to the leaderboard, I need the score and a GPU-Z window.

Other submissions do not have this as the rule was added after someone photo-shopped their results and their submissions were grandfathered in.

The only results I remove are from banned or no longer active users. I've had my own moderation history due to my personality type and my knack for strongly disagreeing with others. The Tom's Hardware moderators are very direct. I'm trying to stay here.

Once more so that it's clear:

Your submission must include a GPU-Z screenshot and the Superposition results window within the same frame to be added to the leaderboard. This is non negotiable.

That is fine. I understand. I assume then I can't take the saved screen of superposition and open it up in a photo viewer and open up GPU-Z alongside it? I would need to rerun the benchmark until I get my score again?

 

[EquineHero]

That is fine. I understand. I assume then I can't take the saved screen of superposition and open it up in a photo viewer and open up GPU-Z alongside it? I would need to rerun the benchmark until I get my score again?

You will need to run the benchmark again.

 

[QwerkyPengwen]

Here you go. A new screenshot with GPU-Z

And just as an added bonus, here's a couple pictures taken with my phone.

 

[EquineHero]

Here you go. A new screenshot with GPU-Z

And just as an added bonus, here's a couple pictures taken with my phone.

>takes picture with phone

 

[fagetti]

Got a 5 years old msi 780ti but at stock its same as my 780 Lightning overclocked, i will post once i hit 3000 score stable. .im trying to get 1300-1350 core without raising voltage too much. Im comfortable below 1.3v since its air cooled.

EDIT: Equinehero have you tried raising your mem or gpu voltage on that 1070? What voltage controller your model has? Maybe it can be unlocked like my 780 Lightning and 780ti i can go upto 1,45v without hard mod (anything above 1.25-1.3v needs water cooling)

 

[EquineHero]

EDIT: Equinehero have you tried raising your mem or gpu voltage on that 1070? What voltage controller your model has? Maybe it can be unlocked like my 780 Lightning and 780ti i can go upto 1,45v without hard mod (anything above 1.25-1.3v needs water cooling)

I'd prefer a hard mod over a soft mod. There's less to screw up.

My 1070 Ti is at +180 core and +500 mem, any higher on the memory and it starts artifacting, and it crashes above 2100MHz, which is already insane for Pascal.

My old 1050 Ti used to get 2150-2200Mhz, but I sold it off a while back

 

[EquineHero]

Consider the suggested thread as started!

Here are some scores from when I first installed my Vega:

Yo, have you tried the Vega PowerPlay mod?

https://www.youtube.com/watch?v=w6gpxe0QoUs

Vega 56 outperforms the RTX 2070 in most cases but draws about 300 to 400 watts.

 

[fagetti]

Consider the suggested thread as started!

Here are some scores from when I first installed my Vega:

Yo, have you tried the Vega PowerPlay mod?

https://www.youtube.com/watch?v=w6gpxe0QoUs

Vega 56 outperforms the RTX 2070 in most cases but draws about 300 to 400 watts.

Thats same wattage what my 780 Lightning took in furmark with modded bios and voltage 1.3

 

[EquineHero]

Thats same wattage what my 780 Lightning took in furmark with modded bios and voltage 1.3

Too bad Pascal is so locked down, I'd love to unlock my 1070 Ti's power limits and max out my PSU to see how good Rosewill really is

 

[Rogue Leader]

Consider the suggested thread as started!

Here are some scores from when I first installed my Vega:

Yo, have you tried the Vega PowerPlay mod?

https://www.youtube.com/watch?v=w6gpxe0QoUs

Vega 56 outperforms the RTX 2070 in most cases but draws about 300 to 400 watts.

My Vega 64 LC already draws 330 average and can spike to 360 as is. i can only imagine if i did this mod it would light my PSU on fire. No thanks.

 

[EquineHero]

Consider the suggested thread as started!

Here are some scores from when I first installed my Vega:

Yo, have you tried the Vega PowerPlay mod?

https://www.youtube.com/watch?v=w6gpxe0QoUs

Vega 56 outperforms the RTX 2070 in most cases but draws about 300 to 400 watts.

My Vega 64 LC already draws 330 average and can spike to 360 as is. i can only imagine if i did this mod it would light my PSU on fire. No thanks.

Get an AX200i then

 

[Rogue Leader]

Consider the suggested thread as started!

Here are some scores from when I first installed my Vega:

Yo, have you tried the Vega PowerPlay mod?

https://www.youtube.com/watch?v=w6gpxe0QoUs

Vega 56 outperforms the RTX 2070 in most cases but draws about 300 to 400 watts.

My Vega 64 LC already draws 330 average and can spike to 360 as is. i can only imagine if i did this mod it would light my PSU on fire. No thanks.

Get an AX200i then

No thanks, already upgraded my PSU somewhat because 650 wasn't cutting it. Driver updates have increased power consumption. There comes a point....

 

[EquineHero]

No thanks, already upgraded my PSU somewhat because 650 wasn't cutting it. Driver updates have increased power consumption. There comes a point....

If I had the money I'd buy an AX1200i in a heartbeat because of the low load efficiency.

 

[JaSoN_cRuZe]

Here are the results of my Sapphire RX 570 Nitro +
Kindly update the master results table
Thanks

Guess i won the silicon lottery...

 

[EquineHero]

Here are the results of my Sapphire RX 570 Nitro +
Kindly update the master results table
Thanks
Guess i won the silicon lottery...

Your results would be better if you had a better motherboard and faster memory. I'll update the table later. I'm busy.

 

[JaSoN_cRuZe]

Here are the results of my Sapphire RX 570 Nitro +
Kindly update the master results table
Thanks
Guess i won the silicon lottery...

Your results would be better if you had a better motherboard and faster memory. I'll update the table later. I'm busy.

Yeah, Thanks. I'm planning on upgrade next year to Zen2!!

 

[TechyInAZ]

Doubt it's his CPU and RAM. I ran my 1060 6GB on my A8-7600 with superposition, basically identical results vs my 4.5ghz 4690K.

 

[binderasaf1]

Hello dear superposition people.
There's a mystery i have yet to solve , and I'm super curious as to what's the solution.
Upon browsing through superpositions hof in unigine's website, i stumbled upon this score (attached in the pic).
http://i63.tinypic.com/vpx0xw.jpg

There's a 700ish points difference between us, and 5-6 fps, which is a lot.
Now I'm baffled by this, as both of us have ryzen systems, the only difference being me having a 6 core but running 3.95 ghz (he is running 3.9ghz). The impact on results, regardless, is marginal at best. If anything, my 50mhz advantage should help more than 2 extra cores. We both run dual channel 16 gigs, and my mem is actually clocked higher. Gpu Clockwise, this guy runs his core 50mhz more than me, and 90mhz more on the memory. This difference can never account for such a big difference in score/fps. In the real world its 1 fps at best. I tried all the regular nvidia control panel tweaks, but i can't go past 6346ish....
Am i missing something? What's the secret sauce for this big discrepency in results? I'm pretty sure its not the clocks...
Help anyone ?

 

[jankerson]

Hello dear superposition people.
There's a mystery i have yet to solve , and I'm super curious as to what's the solution.
Upon browsing through superpositions hof in unigine's website, i stumbled upon this score (attached in the pic).
http://i63.tinypic.com/vpx0xw.jpg

There's a 700ish points difference between us, and 5-6 fps, which is a lot.
Now I'm baffled by this, as both of us have ryzen systems, the only difference being me having a 6 core but running 3.95 ghz (he is running 3.9ghz). The impact on results, regardless, is marginal at best. If anything, my 50mhz advantage should help more than 2 extra cores. We both run dual channel 16 gigs, and my mem is actually clocked higher. Gpu Clockwise, this guy runs his core 50mhz more than me, and 90mhz more on the memory. This difference can never account for such a big difference in score/fps. In the real world its 1 fps at best. I tried all the regular nvidia control panel tweaks, but i can't go past 6346ish....
Am i missing something? What's the secret sauce for this big discrepency in results? I'm pretty sure its not the clocks...
Help anyone ?

Yeah, big difference... He OCed his GTX 1080Ti to the moon, look at the memory clocks for the GPU.... 6300 MHz...

That's the difference.

 

[binderasaf1]

Hello dear superposition people.
There's a mystery i have yet to solve , and I'm super curious as to what's the solution.
Upon browsing through superpositions hof in unigine's website, i stumbled upon this score (attached in the pic).
http://i63.tinypic.com/vpx0xw.jpg

There's a 700ish points difference between us, and 5-6 fps, which is a lot.
Now I'm baffled by this, as both of us have ryzen systems, the only difference being me having a 6 core but running 3.95 ghz (he is running 3.9ghz). The impact on results, regardless, is marginal at best. If anything, my 50mhz advantage should help more than 2 extra cores. We both run dual channel 16 gigs, and my mem is actually clocked higher. Gpu Clockwise, this guy runs his core 50mhz more than me, and 90mhz more on the memory. This difference can never account for such a big difference in score/fps. In the real world its 1 fps at best. I tried all the regular nvidia control panel tweaks, but i can't go past 6346ish....
Am i missing something? What's the secret sauce for this big discrepency in results? I'm pretty sure its not the clocks...
Help anyone ?

Yeah, big difference... He OCed his GTX 1080Ti to the moon, look at the memory clocks for the GPU.... 6300 MHz...

That's the difference.

Nope...mine is clocked at 6210mhz...i actually benched at 6300 as well and it certainely wasnt a 5 fps difference. Its also not the "looser memory timings from increased clocks" which degrades the performance, as my scores and fps scales directly with the memory clocks consistently...im simply at 6210mhz as its my actual gaming mem clock. Its Not the memory.

 

[jankerson]

Hello dear superposition people.
There's a mystery i have yet to solve , and I'm super curious as to what's the solution.
Upon browsing through superpositions hof in unigine's website, i stumbled upon this score (attached in the pic).
http://i63.tinypic.com/vpx0xw.jpg

There's a 700ish points difference between us, and 5-6 fps, which is a lot.
Now I'm baffled by this, as both of us have ryzen systems, the only difference being me having a 6 core but running 3.95 ghz (he is running 3.9ghz). The impact on results, regardless, is marginal at best. If anything, my 50mhz advantage should help more than 2 extra cores. We both run dual channel 16 gigs, and my mem is actually clocked higher. Gpu Clockwise, this guy runs his core 50mhz more than me, and 90mhz more on the memory. This difference can never account for such a big difference in score/fps. In the real world its 1 fps at best. I tried all the regular nvidia control panel tweaks, but i can't go past 6346ish....
Am i missing something? What's the secret sauce for this big discrepency in results? I'm pretty sure its not the clocks...
Help anyone ?

Yeah, big difference... He OCed his GTX 1080Ti to the moon, look at the memory clocks for the GPU.... 6300 MHz...

That's the difference.

Nope...mine is clocked at 6210mhz...i actually benched at 6300 as well and it certainely wasnt a 5 fps difference. Not the memory.

SP is all GPU memory bandwidth..

 

[binderasaf1]

Hello dear superposition people.
There's a mystery i have yet to solve , and I'm super curious as to what's the solution.
Upon browsing through superpositions hof in unigine's website, i stumbled upon this score (attached in the pic).
http://i63.tinypic.com/vpx0xw.jpg

There's a 700ish points difference between us, and 5-6 fps, which is a lot.
Now I'm baffled by this, as both of us have ryzen systems, the only difference being me having a 6 core but running 3.95 ghz (he is running 3.9ghz). The impact on results, regardless, is marginal at best. If anything, my 50mhz advantage should help more than 2 extra cores. We both run dual channel 16 gigs, and my mem is actually clocked higher. Gpu Clockwise, this guy runs his core 50mhz more than me, and 90mhz more on the memory. This difference can never account for such a big difference in score/fps. In the real world its 1 fps at best. I tried all the regular nvidia control panel tweaks, but i can't go past 6346ish....
Am i missing something? What's the secret sauce for this big discrepency in results? I'm pretty sure its not the clocks...
Help anyone ?

Yeah, big difference... He OCed his GTX 1080Ti to the moon, look at the memory clocks for the GPU.... 6300 MHz...

That's the difference.

Nope...mine is clocked at 6210mhz...i actually benched at 6300 as well and it certainely wasnt a 5 fps difference. Not the memory.

SP is all GPU memory bandwidth..

I can promise you, and im saying this from personal experience, the difference between 6210mhz and 6300mhz on the memory is never 5-6 fps. 1fps at most if not 0.5 in sp. Its most certainly not the memory.