Jump to content

Why does the GTX 1060 perform better than the rx 480 if the 480 has more tflops

Go to solution Solved by don_svetlio,

Cause DX11? Look at DX12/Vulkan and the 480 is 10% faster.

2 minutes ago, App4that said:

Damn man, don't be afraid to show a little bias once in awhile. It's OK. /s

 

Async is still locked on Nvidia cards with Doom, so once that is fixed the 1060 will show a gain.

Lol, nah I try to be neutral on as much as I can. But between the AMD AIBs and slightly lower price of AMD cards, AMD tends to be preferable.

 

Though I had been considering going with an EVGA GTX 1070 but the cost was kinda off putting because I'm not sure how long it will take me to save up enough money for everything, so trying to reduce the total price makes it a little more obtainable.

 

Anyways is Async locked in all games for Nvidia cards because I just checked the AMD subreddit and it looks like there are some retested benches in DX12 showing a different story about the performance differences. Because DX12 and Vulkan heavily benefit from async. But if async is locked on the Nvidia cards in all games then I could see why DX12 benches might show some preference to the 480 because that wouldn't be as fair of a comparison.

a Moo Floof connoisseur and curator.

:x@handymanshandle x @pinksnowbirdie || Jake x Brendan :x
Youtube Audio Normalization
 

 

 

Link to comment
Share on other sites

Link to post
Share on other sites

1 minute ago, wcreek said:

Lol, nah I try to be neutral on as much as I can. But between the AMD AIBs and slightly lower price of AMD cards, AMD tends to be preferable.

 

Though I had been considering going with an EVGA GTX 1070 but the cost was kinda off putting because I'm not sure how long it will take me to save up enough money for everything, so trying to reduce the total price makes it a little more obtainable.

 

Anyways is Async locked in all games for Nvidia cards because I just checked the AMD subreddit and it looks like there are some retested benches in DX12 showing a different story about the performance differences. Because DX12 and Vulkan heavily benefit from async. But if async is locked on the Nvidia cards in all games then I could see why DX12 benches might show some preference to the 480 because that wouldn't be as fair of a comparison.

It's lock in Doom since id concentrated on the 480 first since it launched first. The new update should hit any day for Nvidia.

If anyone asks you never saw me.

Link to comment
Share on other sites

Link to post
Share on other sites

57 minutes ago, SamStrecker said:

Well considering see it's performance on Linux in other applications doing better then DX12 on windows. I can safely say it Vulkan becomes mainstream in every game there is no need to use windows anymore

That wasn't my point. My point was that you cannot discern a pattern from only one occurrence. You cannot judge Vulkan as a platform, and the 1060 or 480's performance using it, on a game that for all we know could have been very AMD biased. If a whole bunch of games use Vulkan and they all follow the same pattern then you can make that kind of statement.

 

I hope Vulkan does take off instead of Dx12. That wasn't my issue with what you said.

 

Edit: I'm actually quite excited about Vulkan. True cross platform gaming, not dependant on drivers... Why would anyone pay for Windows if this takes off?

Link to comment
Share on other sites

Link to post
Share on other sites

This is something thats always puzzled me.

 

Why are AMD stream processors inferior to Nvidia cuda cores?

 

I have a GTX460 with 336 cuda cores and I also have an HD4850 with 800 stream processors. Then shouldn't the HD4850 kick my 460's ass? But no the reverse is true. My 460 is nearly twice as fast as my old 4850. 

Link to comment
Share on other sites

Link to post
Share on other sites

3 minutes ago, 42kq said:

This is something thats always puzzled me.

 

Why are AMD stream processors inferior to Nvidia cuda cores?

 

I have a GTX460 with 336 cuda cores and I also have an HD4850 with 800 stream processors. Then shouldn't the HD4850 kick my 460's ass? But no the reverse is true. My 460 is nearly twice as fast as my old 4850. 

Because a lot of those cores are, a lot of the time, sitting around doing nothing waiting for other processes to finish so they can be utilised. This is why there are diminishing returns to just adding in more cores, and why the Fury X doesn't perform anything like it should in Dx11.

 

This is the whole point of Asynchronous Shaders. Those underutilised cores can be getting on with other tasks that can be done in parallel while other parts of the chip can be processing the graphics. This is why Asynchronous Compute benefits big GPUs like the Fury more than small GPUs like the 480, and why Nvidia didn't bother with it with Maxwell. Nvidia claimed that their GPUs were already utilised enough that this wouldn't give any benefit. This is why their solution for Pascal is preemption -- reducing the time taken to switch activities, rather than have different parts of the chip working on entirely different workloads simultaneously.

Link to comment
Share on other sites

Link to post
Share on other sites

@othertomperson

So just how inefficient are Stream processors in DX11? I remember my 8800gts rivaling an HD3850.

 

P.S. I dont plan on downgrading to windows 10 so dx12 is irrelevant to me.

Link to comment
Share on other sites

Link to post
Share on other sites

FLOPS is a measure of the theoretical power of a card. It measures how much floating point math a GPU can do per second if you keep it's shader cores fed 100% of the time. In most real world activities this is just far from achievable due to the enthropic and dynamic nature of programs such as games. Synthetic benchmarks can measure FLOPS because they are designed to be a predictable and easy to process workload.

CPU: Intel Core i3 4370 (3.8GHz, 2C/4T) GPU: AMD R9 380X 4GB

Link to comment
Share on other sites

Link to post
Share on other sites

To clear things up on which version of openGL Doom used, on the demo it used 4.3 for me, full game uses 4.5

 

The performance was fine, sure the boost from vulcan is huge, made no difference to me as i lock my frame rate to 75 vsync. both in opengl and vulcan it was fine.

 

its a myth that AMD's performance in DX11 is bad, it's fucking good (unless PhysX is baked in), it's just Nvidia's is better...

 

Both the 1060 and 480's are in the same performance ballpark, sure the 1080 is a bit ahead in dx11 not enough to put it into another tier like you would think, the way people are going on about it (both are 1080p60-75fps cards yes they get above 75 but not far enough above to be 1080@120,144 class across the board), but the 480 AIB's haven't release their cards yet, and given you can get a 1080p 50-75 (not the best) or 35-144 freesync monitor now for £120 and £190 respectively, AMD still offers better value for your money if you want a great high frame rate with no tearing gaming experience....

 

Tflops, wise the 480 is coming under what i would expect out of a 5.9tflop card, maybe there is more performance to be tapped into yet, AMD does improve performance historically with drivers...Doom showed the potential of the card for sure, and the first driver revision yielded a 3% gain in performance apparently...so with time they maybe able to squeeze a fair bit more out of it...

 

who knows

----Ryzen R9 5900X----X570 Aorus elite----Vetroo V5----240GB Kingston HyperX 3k----Samsung 250GB EVO840----512GB Kingston Nvme----3TB Seagate----4TB Western Digital Green----8TB Seagate----32GB Patriot Viper 4 3200Mhz CL 16 ----Power Color Red dragon 5700XT----Fractal Design R4 Black Pearl ----Corsair RM850w----

Link to comment
Share on other sites

Link to post
Share on other sites

6 minutes ago, super_skank said:

To clear things up on which version of openGL Doom used, on the demo it used 4.3 for me, full game uses 4.5

 

Nope, try again. The internet never forgets ;) 

 

 

If anyone asks you never saw me.

Link to comment
Share on other sites

Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

×