1060 Launch-No Wonder Nv is King

[Gregster]

The 1060 isn't smashing up the competition, it's winning in some-most DX11, but losing on DX12 in some to most - and absolutely getting ripped a new one in Vulcan.

So how is that smashing up the competition?

And I don't go on about the 1070/1080 as they have no direct competitor yet.

Also, throw in the knowledge that the 480 will be supported much longer, and not just forgotten like NVidia's previous generations have on MULTIPLE OCCASIONS(DX12 drivers for Fermi, Proper Async for Maxwell, and maybe even pascal etc..).

Pepper it up as much as you want, I don't care lol. I also don't have my crystal ball handy, so I can't see into the future.

ohhh, one thing I do remember though and quite important this really and a plus for AMD (maybe?) There has been a couple of drivers in the last few years and forgive me if my memory is wrong but I am sure someone will remind me.... 12.11 was the big GCN driver that gave AMD's 7xxx a big boost in frames, that came out 10 months after the 7970 launch and showed what the 7970 was capable of. It smashed up the 680 in most games and was like a new card for those that had 7970s. Next came the Crimson drivers more recently and they gave the 2xx/3xx/fury series a nice boost that was similar to the 12.11 drivers. Now NVidia either have great optimisation from the off (I think this is the case) or they rarely bother with any optimisations and what you see is what you get. People claim that NVidia gimp divers but what they are missing is there isn't any gimping, it is AMD improving and giving their customers the performance they paid for.

Funny old world but I am trying as hard today as yesterday

 

[Gripen90]

Spot on for you Gripen and I know you are a big mGPU user

Hehe yea, I just love 'em cards

 

[Dóminion]

I am future proofing myself with the AMD offering which is likely to perform much better under DX12/vulcan titles. Most of the review sites out there are benchmarking DX11 games, NVIDA optimised titles under DX12 and when benchmarking DOOM are opting openGL. This is heavily biased towards NVIDIA. The price to performance for the AMD cards will shine through. Add to that the lack of a-sync compute at a hardware level and sli on the 1060 and I just can't go NVIDIA. Also going by Nvidia's track record so far there's going to be a lot of cards that are going to be offered higher the the rrp. That's just my take.

 

[Scania]

NV are kings of milking their customers, I'm somewhat disappointed they don't include a small jar of Vaseline with each card!

 

[Gripen90]

I am future proofing myself with the AMD offering which is likely to perform much better under DX12/vulcan titles. Most of the review sites out there are benchmarking DX11 games, NVIDA optimised titles under DX12 and when benchmarking DOOM are opting openGL. This is heavily biased towards NVIDIA. The price to performance for the AMD cards will shine through. Add to that the lack of a-sync compute at a hardware level and sli on the 1060 and I just can't go NVIDIA. Also going by Nvidia's track record so far there's going to be a lot of cards that are going to be offered higher the the rrp. That's just my take.

I'd recommend reading these two articles about async compute. Nvidia does it one way, AMD the other way - but both do async compute which isn't just one thing, but a lot of different methods.

"Getting back to Pascal then, Pascal finally fixes the resource allocation issue. For Pascal, NVIDIA has implemented a dynamic load balancing system to replace Maxwell 2’s static partitions. Now if the queues end up unbalanced and one of the queues runs out of work early, the driver and work schedulers can step in and fill up the remaining time with work from the other queues.

In concept it sounds simple, and in practice it should make a large difference to how beneficial async compute can be on NVIDIA’s architectures. Adding more work to create concurrency to fill execution bubbles only works if the queue scheduling itself doesn’t create bubbles, and this was Maxwell 2’s Achilles’ heel that Pascal has addressed.

At the same time however I feel it’s important to note that the scheduling change alone won’t (and can’t) guarantee that Pascal will see significant gains from async compute across the board. Async compute itself is a catch-all term – there are lots of things you can do with asynchronous work submission/execution – so async doesn’t mean that a game is making significant use of concurrency. Furthermore the concurrency is still based on filling execution bubbles, and that means that there needs to be bubbles to fill in the first place. In other words, the greatest gains from async will come from scenarios where for whatever reason, the graphics queue and its synchronous shaders can’t completely saturate the GPU on its own."
http://www.anandtech.com/show/10325/the-nvidia-geforce-gtx-1080-and-1070-founders-edition-review/9

"Whether work placed in the COMPUTE queue is executed in parallel or in serial is ultimately the decision of the underlying driver. In DirectX 12, by placing items into a different queue the application is simply stating that it allows execution to take place in parallel - it is not a requirement, nor is there a method for making such a demand. This is similar to traditional multi-threaded programming for the CPU - by creating threads we allow and are prepared for execution to happen simultaneously. It is up to the OS to decide how it distributes the work."
http://www.futuremark.com/pressrele...time-spy?_ga=1.62804140.1212684934.1468965657

 

[Dóminion]

I'd recommend reading these two articles about async compute. Nvidia does it one way, AMD the other way - but both do async compute which isn't just one thing, but a lot of different methods.

"Getting back to Pascal then, Pascal finally fixes the resource allocation issue. For Pascal, NVIDIA has implemented a dynamic load balancing system to replace Maxwell 2’s static partitions. Now if the queues end up unbalanced and one of the queues runs out of work early, the driver and work schedulers can step in and fill up the remaining time with work from the other queues.

In concept it sounds simple, and in practice it should make a large difference to how beneficial async compute can be on NVIDIA’s architectures. Adding more work to create concurrency to fill execution bubbles only works if the queue scheduling itself doesn’t create bubbles, and this was Maxwell 2’s Achilles’ heel that Pascal has addressed.

At the same time however I feel it’s important to note that the scheduling change alone won’t (and can’t) guarantee that Pascal will see significant gains from async compute across the board. Async compute itself is a catch-all term – there are lots of things you can do with asynchronous work submission/execution – so async doesn’t mean that a game is making significant use of concurrency. Furthermore the concurrency is still based on filling execution bubbles, and that means that there needs to be bubbles to fill in the first place. In other words, the greatest gains from async will come from scenarios where for whatever reason, the graphics queue and its synchronous shaders can’t completely saturate the GPU on its own."
http://www.anandtech.com/show/10325/the-nvidia-geforce-gtx-1080-and-1070-founders-edition-review/9

"Whether work placed in the COMPUTE queue is executed in parallel or in serial is ultimately the decision of the underlying driver. In DirectX 12, by placing items into a different queue the application is simply stating that it allows execution to take place in parallel - it is not a requirement, nor is there a method for making such a demand. This is similar to traditional multi-threaded programming for the CPU - by creating threads we allow and are prepared for execution to happen simultaneously. It is up to the OS to decide how it distributes the work."
http://www.futuremark.com/pressrele...time-spy?_ga=1.62804140.1212684934.1468965657

I'm very much aware that the way NVIDIA (software) and AMD (hardware) handle this is different and of course a-sync is only part of the picture. There are certainly going to be other features on AMD cards and NVIDIA cards that will have an effect on performance.

What I have seen however is that AMD does seem to have a noticeable advantage under DX12 compared to NVIDA. This has been repeatedly reported. Frustratingly however DX12 seems to have been ignored by the reviewers. Some major reviewers did mention DX12 but concentrated on Tombraider which seems to be an NVIDIA optimised title, which unsurprisingly performs well for NVIDA, but with little performance gain under DX12 for them.

This is what I find difficult to stomach, that reviewers are not giving fair reviews, that they are failing to mention or cover NVIDIA's failings under DX12 but seem more than happy to point out all the AMD failings. And I don't say all of this because I am a fan of one card over the other. What I want is genuine, non biased reviews. NVIDIA does great at marketing itself and this one of the reasons they sell so well. Reviewers should offer sound reviews which are independent of this nonsense and focus on the consumer.

 

[humbug]

If Futuremark use an overall score it would be wrong for us to change that.

It would also be wrong not to list the graphics score in the tables, something a lot of review sites are guilty of.

If people really want to use a benchmark that relies 99% on graphics there is always Heaven or Valley but the red team have objections to them too.

You have a faster CPU than just about anyone, nothing to do with that, eh?

At the same time you've put yourself in here boasting about how high up the score board you got when its your CPU, not your GPU that got you there...

Riiiiight....

There is a 390X that beaten you and yet under you because of your CPU, there are also a bunch or Fury cards that are faster and yet also under you on the score board.

At least have the dignity to respect those users by not boasting about beating them where really you haven't.

 

[Kaapstad]

You have a faster CPU than just about anyone, nothing to do with that, eh?

At the same time you've put yourself in here boasting about how high up the score board you got when its your CPU, not your GPU that got you there...

Riiiiight....

There is a 390X that beaten you and yet under you because of your CPU, there are also a bunch or Fury cards that are faster and yet also under you on the score board.

At least have the dignity to respect those users by not boasting about beating them where really you haven't.

I did not boast about anything and unlike a lot of the stuff in this thread I posted the full information (the whole scoreboard) so people could view it how they wished.

It is a good job I don't take your view of things -

Do I complain when 8 Pack posts a score done on LN2 - of course not.

Do I complain when 8 Pack uses a binned CPU as he has an unlimited supply - of course not.

Do I complain when AMDMatt posts a score as he had one of the fastest CPUs, - of course not.

I am going to run my RX480s in the same PC would you prefer it if I kept the results to myself as you may accuse me of boasting ?

 

[humbug]

I did not boast about anything and unlike a lot of the stuff in this thread I posted the full information (the whole scoreboard) so people could view it how they wished.

It is a good job I don't take your view of things -

Do I complain when 8 Pack posts a score done on LN2 - of course not.

Do I complain when 8 Pack uses a binned CPU as he has an unlimited supply - of course not.

Do I complain when AMDMatt posts a score as he had one of the fastest CPUs, - of course not.

I am going to run my RX480s in the same PC would you prefer it if I kept the results to myself as you may accuse me of boasting ?

Very pleased with my 1060

Here it is (about half way up the scoreboard) on the new Time Spy bench and there are only 5 red scores above it.

You are clearly boasting about how few "Red Scores" are above you, there would be more above you on GPU scores as there are faster GPU's that are below you on account of your £800 CPU, including one with a 3 year old architecture.

Its complete disrespect for those with faster cards below you because of the way you set up the ranking system.
That just bad form...

 

[docsonic]

NV are kings of milking their customers, I'm somewhat disappointed they don't include a small jar of Vaseline with each card!

They are too tight for that!

I should really go to sleep

 

[tommybhoy]

You shouldn't drink and drive

Would never happen as I don't drink at all.

you should use drugs and post on forums.

I always have!

 

[Kaapstad]

You are clearly boasting about how few "Red Scores" are above you, there would be more above you on GPU scores as there are faster GPU's that are below you on account of your £800 CPU, including one with a 3 year old architecture.

Its complete disrespect for those with faster cards below you because of the way you set up the scoring system.

What I said is no different to what anyone else would have said, I was stating a fact.

I ask you again, do you think I should post any scores using my RX480s in that PC or would you class it as boasting.

I am also willing to delete all my scores from the Time Spy bench thread if you think they have no value or you consider them as boasting. what should I do ?

As to the layout of the thread, Time Spy is a system bench and the scoring reflects that. If it is good enough for Futuremark it is good enough for me.

 

[TheRealDeal]

What I said is no different to what anyone else would have said, I was stating a fact.

I ask you again, do you think I should post any scores using my RX480s in that PC or would you class it as boasting.

I am also willing to delete all my scores from the Time Spy bench thread if you think they have no value or you consider them as boasting. what should I do ?

As to the layout of the thread, Time Spy is a system bench and the scoring reflects that. If it is good enough for Futuremark it is good enough for me.

Would it be hard to have 2 lists. One as you have it and the other ranking gpu scores. I think one of the old bench threads had it this way. You do all the work so if it's to much don't bother. AMDMatt won't be happy as his 1core/duo would slide way down the list on the GS list.

 

[humbug]

What I said is no different to what anyone else would have said, I was stating a fact.

I ask you again, do you think I should post any scores using my RX480s in that PC or would you class it as boasting.

I am also willing to delete all my scores from the Time Spy bench thread if you think they have no value or you consider them as boasting. what should I do ?

As to the layout of the thread, Time Spy is a system bench and the scoring reflects that. If it is good enough for Futuremark it is good enough for me.

If you said the same thing and it was on the account of how you set up the scoring system, yes i would.

People have asked you to change it to make it more fair in a GPU forum benchmark, you refused, that's fine. really, its your choice.
But don't then go round boasting about how far you are up the scoreboard when the reason for that is the problem in the way you set up the benchmark rankings.
If i also had a GPU that was faster than yours but lost out in the ranking on account of your ranking system and then boasted about your positioning i would find that quite offensive.

 

[Kaapstad]

Would it be hard to have 2 charts. One as you have it and the other ranking gpu scores. I think one of the old bench threads had it this way. You do all the work so if it's to much don't bother.

I tried it once with the Skydiver thread, never again.

It is a real pain to do updating 2 tables and very easy to get it wrong.

I colour code the threads now so it is very easy for anyone to pick out whatever they need to see.

 

[Kaapstad]

If you said the same thing and it was on the account of how you set up the scoring system, yes i would.

People have asked you to change it to make it more fair in a GPU forum benchmark, you refused, that's fine. really, its your choice.
But don't then go round boasting about how far you are up the scoreboard when the reason for that is the problem in the way you set up the benchmark rankings.
If i also had a GPU that was faster than yours but lost out in the ranking on account of your ranking system and then boasted about your positioning i would find that quite offensive.

What I said is no different to what anyone else would have said, I was stating a fact.

I ask you again, do you think I should post any scores using my RX480s in that PC or would you class it as boasting.

I am also willing to delete all my scores from the Time Spy bench thread if you think they have no value or you consider them as boasting. what should I do ?

As to the layout of the thread, Time Spy is a system bench and the scoring reflects that. If it is good enough for Futuremark it is good enough for me.

You have not answered the questions ?

 

[TheRealDeal]

I tried it once with the Skydiver thread, never again.

It is a real pain to do updating 2 tables and very easy to get it wrong.

I colour code the threads now so it is very easy for anyone to pick out whatever they need to see.

Then maybe somebody could run there own thread based on Graphics scores. They could take the results from your thread and update there own only with graphics score as the main stat. I can easily read your results and often do look at the Gpu scores so no issue for me.

 

[layte]

If you said the same thing and it was on the account of how you set up the scoring system, yes i would.

People have asked you to change it to make it more fair in a GPU forum benchmark, you refused, that's fine. really, its your choice.
But don't then go round boasting about how far you are up the scoreboard when the reason for that is the problem in the way you set up the benchmark rankings.
If i also had a GPU that was faster than yours but lost out in the ranking on account of your ranking system and then boasted about your positioning i would find that quite offensive.

 

[humbug]

You have not answered the questions ?

I'm not asking you to delete your scores, i just think its poor form to boast about your rankings when that is down to the problem of how you setup those rankings, the same thing we asked you to change.

don't go around boasting about your GPU's rankings when you know its because of the way you set it up.

 

[tommybhoy]

You are clearly boasting about how few "Red Scores" are above you, there would be more above you on GPU scores as there are faster GPU's that are below you on account of your £800 CPU, including one with a 3 year old architecture.

Its complete disrespect for those with faster cards below you because of the way you set up the scoring system.

Sorry hummy, in regards to Kaaps BM's, although I agree completely with what you are saying, but-Big Picture it is the only way to do it, as it's always been.

@Kaap,

He has a point, imo, your 1060 is up there because of your ultra setup, so it obfuscates the whole picture.

I doubt it would be way up there running my [email protected] and similar setups?

I could be wrong though as the 1060 is a very strong gpu.

I know you do a simply massive amount for the OcUK ungrateful's, but any thoughts on integrating another table with GFx scores?

It'l probably put it to bed-then everyone's happy.