Jump to content

Since I just ordered a 7970, I was wondering about the "GPU compute" power of AMD cards compare to Nvidia cards. In loads of reviews people mention that AMD is drastically better when it comes to computing power, but what does this actually mean in real life? When will the superior compute power of AMD cards come into play, and which applications will use them? Does this mean that having a 7970 will speed up Sony Vegas rendering more than a GTX 680 or 780? 

Have you tried turning it off and on again?

Link to comment
https://linustechtips.com/topic/45207-amd-graphics-cards-compute-power/
Share on other sites

Link to post
Share on other sites

Amd cards preform much better in tasks like bit coin mining. A lot of them out preform nvidia gpus.

Desk: monitors 3x Asus VE248h(eyefinity), Keyboard Cm Strom Trigger(mx red), Mouse Corsair m65, Headset Audio Technica ATH-M50

Black Friday 2013 Build: i7 4770k, Gigabyte Z87X UD5H, 16gb Corsair, Msi R9 290, Corsair Axi 760, Corsair 750D, 2x intel 530 240gb ssd, 2x Seagate 400gb

Older Machine amd x640, msi 760g mobo, 8gb gskillz, Sapphire 6870, Corsair hx650, Cooler master haf 922, ocz agility 3 120gb ssd || HTPC: i7 3770k, shuttle xpc z77, 16gb gskillz, Asus GTX 650 ti, intel 120gb msata ssd

Link to post
Share on other sites

depends on what you compute with, in opencl amd wipes the floor with nvidia but  with anything else nvidia has pure compute power in thier higherend cards, as the time of writing this post,

Cpu: Intel i7 4770k @4.4 Ghz | Case: Corsair 350D | Motherbord: Z87 Gryphon | Ram: dominator platinum 4X4 1866 | Video Card: SLI GTX 980 Ti | Power Supply: Seasonic 1000 platinum | Monitor: ACER XB270HU | Keyboard: RK-9100 | Mouse: R.A.T. 7 | Headset : HD 8 DJ | Watercooled

Link to post
Share on other sites

Since I just ordered a 7970, I was wondering about the "GPU compute" power of AMD cards compare to Nvidia cards. In loads of reviews people mention that AMD is drastically better when it comes to computing power, but what does this actually mean in real life? When will the superior compute power of AMD cards come into play, and which applications will use them? Does this mean that having a 7970 will speed up Sony Vegas rendering more than a GTX 680 or 780? 

 

Just mentioning the words OpenCl causes Nvidia GPUs to kill over, a Titan gets its ass whooped by a 7970 to the tune of around $600 cheaper. All other lower level Nvidia cards might as well be considered as case warmers in Open Cl.

Link to post
Share on other sites

In Hitman Absolution, they use DirectCompute to accelerate DOF and AO processing, which is why AMD GPU excels here.

 

 

Direct Compute accelerated Bokeh Depth of Field

 

On top of Hitman Absolution implementing depth of field, the PC version features a much more realistic and visually interesting depth of field by using direct compute to simulate the effect caused by inter-lens reflection of very bright out-of focus objects.

 

Direct Compute Accelerated Ambient Occlusion

 

When objects are located close to each other, the amount of light that can reach the surface of an object is reduced. While many games have been simulating this effect by applying algorithms like AMDs HDAO, Nixxes decided to take it a step further and create a proprietary direct compute based algorithm which combines the ideas behind the most widely used ambient occlusion algorithms to one proprietary high quality, high performance technique.

Sauce: Hitman Absolution Performance and IQ Review - HardOCP.com

| Intel i7-3770@4.2Ghz | Asus Z77-V | Zotac 980 Ti Amp! Omega | DDR3 1800mhz 4GB x4 | 300GB Intel DC S3500 SSD | 512GB Plextor M5 Pro | 2x 1TB WD Blue HDD |
 | Enermax NAXN82+ 650W 80Plus Bronze | Fiio E07K | Grado SR80i | Cooler Master XB HAF EVO | Logitech G27 | Logitech G600 | CM Storm Quickfire TK | DualShock 4 |

Link to post
Share on other sites

Let's just say this... 

http://www.anandtech.com/bench/GPU13/587

680 SLI < 7790 

Console optimisations and how they will effect you | The difference between AMD cores and Intel cores | Memory Bus size and how it effects your VRAM usage |
How much vram do you actually need? | APUs and the future of processing | Projects: SO - here

Intel i7 5820l @ with Corsair H110 | 32GB DDR4 RAM @ 1600Mhz | XFX Radeon R9 290 @ 1.2Ghz | Corsair 600Q | Corsair TX650 | Probably too much corsair but meh should have had a Corsair SSD and RAM | 1.3TB HDD Space | Sennheiser HD598 | Beyerdynamic Custom One Pro | Blue Snowball

Link to post
Share on other sites

Let's just say this... 

http://www.anandtech.com/bench/GPU13/587

680 SLI < 7790 

 

 

LuxMark is an OpenCL based GPU benchmark renderer. Of course AMD cards are going to out-perform Nvidia...

 

 

For OP: Which one is better depends on what type of computing is required. OpenCL? go AMD. CUDA? go Nvidia. Bitcoin farming is dominated by AMD, Folding is better on Nvidia cards

 

Sony Vegas Pro 12 supports both. As long as the cards are somewhat similar, there will be little difference between them unless it's something like a Quadro card which will destroy AMD in render speeds.

4930k @ 4.5GHz - EVGA 780ti Superclocked ACX - ASUS x79 Deluxe - 32GB Tactical Tracer @ 2133

Link to post
Share on other sites

But what does direct compute actually do??? I have never come across a situation where I have needed this feature (aside from bitcoin mining which I do not do). What practical advantages does AMD have over Nvidia?

Have you tried turning it off and on again?

Link to post
Share on other sites

Sony Vegas Pro 12 supports both. As long as the cards are somewhat similar, there will be little difference between them unless it's something like a Quadro card which will destroy AMD in render speeds.

Quite the opposite, Firepro cards destroy Quadro cards several hundred dollars more expensive in any OpenCL workload and AMD cards are better at folding as well.

http://www.tomshardware.com/reviews/best-workstation-graphics-card,3493-30.html

http://www.tomshardware.com/reviews/best-workstation-graphics-card,3493-25.html

Link to post
Share on other sites

But what does direct compute actually do??? I have never come across a situation where I have needed this feature (aside from bitcoin mining which I do not do). What practical advantages does AMD have over Nvidia?

Direct Compute is DX11 based, so it can be implemented in games, an example of this is TressFX in Tomb Raider.

Generally speaking AMD cards (especially the 7000 series) will perform any OpenCL accelerated workload orders of magnitude faster than Nvidia cards.

This is due to the new GCN architecture, which focuses on compute, it was designed specifically to be implemented onto APUs to handle intensive parallel workloads.

Link to post
Share on other sites

Quite the opposite, Firepro cards destroy Quadro cards several hundred dollars more expensive in any OpenCL workload and AMD cards are better at folding as well.

http://www.tomshardware.com/reviews/best-workstation-graphics-card,3493-30.html

http://www.tomshardware.com/reviews/best-workstation-graphics-card,3493-25.html

 

Both of those benchmarks utilize OpenCL. Of course they show AMD to destroy everything...

 

On that TomsHardware article, look at any of the other benchmark results that use real world examples and then say the Firepro cards destroy the Quadro.

4930k @ 4.5GHz - EVGA 780ti Superclocked ACX - ASUS x79 Deluxe - 32GB Tactical Tracer @ 2133

Link to post
Share on other sites

Both of those benchmarks utilize OpenCL. Of course they show AMD to destroy everything...

 

On that TomsHardware article, look at any of the other benchmark results that use real world examples and then say the Firepro cards destroy the Quadro.

So according to you anything that uses OpenCL is not a real world example ? even if it was a program analysts use for financial markets ? or professional video production suites ? what about photoshop ? that's not real-world to you?

 

Link to post
Share on other sites

So according to you anything that uses OpenCL is not a real world example ? even if it was a program analysts use for financial markets ? or professional video production suites ? what about photoshop ? that's not real-world to you?

 

 

Op asked if AMD cards have more compute power over Nvidia. I pointed out it depends on the software and workload you give them. You give benchmarks that can only favor AMD. I'm not saying OpenCL isn't a real world thing, of course it is. My point is, you can't use a biased benchmark to give an overall impression between two brands of GPU.

 

Like I said, OpenCL is AMD's territory and CUDA is Nvidia's. Similar consumer-grade desktop (not workstation) cards will perform equally well when given a workload that suits that GPU.

4930k @ 4.5GHz - EVGA 780ti Superclocked ACX - ASUS x79 Deluxe - 32GB Tactical Tracer @ 2133

Link to post
Share on other sites

Okay, OP originally asked if choosing one card over the other will drastically accelerate video rendering:

 

If you're doing a lot of editing and putting in a lot of affects that are GPU accelerated, then you'll want to pay attention to what video card you get, but if all you're doing in Vegas is cutting together a few clips of gaming for youtube or whatever (in other words not doing any editing to the actual video, just cutting and pasting), then no GPU is going to accelerate your rendering because that scenario is basically just encoding which is going to be CPU and possibly disk bound.

Workstation: 3930k @ 4.3GHz under an H100 - 4x8GB ram - infiniband HCA  - xonar essence stx - gtx 680 - sabretooth x79 - corsair C70 Server: i7 3770k (don't ask) - lsi-9260-4i used as an HBA - 6x3TB WD red (raidz2) - crucia m4's (60gb (ZIL, L2ARC), 120gb (OS)) - 4X8GB ram - infiniband HCA - define mini  Goodies: Røde podcaster w/ boom & shock mount - 3x1080p ips panels (NEC monitors for life) - k90 - g9x - sp2500's - HD598's - kvm switch

ZFS tutorial

Link to post
Share on other sites

Op asked if AMD cards have more compute power over Nvidia. I pointed out it depends on the software and workload you give them. You give benchmarks that can only favor AMD. I'm not saying OpenCL isn't a real world thing, of course it is. My point is, you can't use a biased benchmark to give an overall impression between two brands of GPU.

 

Like I said, OpenCL is AMD's territory and CUDA is Nvidia's. Similar consumer-grade desktop (not workstation) cards will perform equally well when given a workload that suits that GPU.

OpenCL is an open standard it's not really something only AMD has control over unlike CUDA which is really Nvidia's proprietary technology.

Link to post
Share on other sites

OpenCL is an open standard it's not really something only AMD has control over unlike CUDA which is really Nvidia's proprietary technology.

 

Never said it wasn't :)

4930k @ 4.5GHz - EVGA 780ti Superclocked ACX - ASUS x79 Deluxe - 32GB Tactical Tracer @ 2133

Link to post
Share on other sites

Okay, OP originally asked if choosing one card over the other will drastically accelerate video rendering:

 

If you're doing a lot of editing and putting in a lot of affects that are GPU accelerated, then you'll want to pay attention to what video card you get, but if all you're doing in Vegas is cutting together a few clips of gaming for youtube or whatever (in other words not doing any editing to the actual video, just cutting and pasting), then no GPU is going to accelerate your rendering because that scenario is basically just encoding which is going to be CPU and possibly disk bound.

So in what scenario would an AMD GPU accelerate video rendering over an Nvidia GPU?

 

Direct Compute is DX11 based, so it can be implemented in games, an example of this is TressFX in Tomb Raider.

Generally speaking AMD cards (especially the 7000 series) will perform any OpenCL accelerated workload orders of magnitude faster than Nvidia cards.

This is due to the new GCN architecture, which focuses on compute, it was designed specifically to be implemented onto APUs to handle intensive parallel workloads.

Does this mean that in future games AMD will have a huge advantage? What role does DirectCompute play in upcoming PC games?

Have you tried turning it off and on again?

Link to post
Share on other sites

So in what scenario would an AMD GPU accelerate video rendering over an Nvidia GPU?

Read what people have said above, that's pretty much been answered, but again, it depends completely on the program you're using and what specific affects you're using in that program,

(aka I don't have enough experience to know enough about it)

Like I said though, if you're just doing minor edits and then rendering, you're not really going to be using much in the way of GPU power; you're going to be CPU bound.

Workstation: 3930k @ 4.3GHz under an H100 - 4x8GB ram - infiniband HCA  - xonar essence stx - gtx 680 - sabretooth x79 - corsair C70 Server: i7 3770k (don't ask) - lsi-9260-4i used as an HBA - 6x3TB WD red (raidz2) - crucia m4's (60gb (ZIL, L2ARC), 120gb (OS)) - 4X8GB ram - infiniband HCA - define mini  Goodies: Røde podcaster w/ boom & shock mount - 3x1080p ips panels (NEC monitors for life) - k90 - g9x - sp2500's - HD598's - kvm switch

ZFS tutorial

Link to post
Share on other sites

So in what scenario would an AMD GPU accelerate video rendering over an Nvidia GPU?

 

Does this mean that in future games AMD will have a huge advantage? What role does DirectCompute play in upcoming PC games?

I would say that with AMD hardware in both next-gen consoles developers will work with AMD to maximize performance so rest assured that direct compute and PpenCL acceleration will be a major part of game development in the near future.

Link to post
Share on other sites

Op asked if AMD cards have more compute power over Nvidia. I pointed out it depends on the software and workload you give them. You give benchmarks that can only favor AMD. I'm not saying OpenCL isn't a real world thing, of course it is. My point is, you can't use a biased benchmark to give an overall impression between two brands of GPU.

Like I said, OpenCL is AMD's territory and CUDA is Nvidia's. Similar consumer-grade desktop (not workstation) cards will perform equally well when given a workload that suits that GPU.

CUDA is biased towards Nvidia. OpenCL is open standard and not biased.

Just because AMD is better than Nvidia at OpenCL doesn't mean it's biased. It just means it's better.

If you ever need help with a build, read the following before posting: http://linustechtips.com/main/topic/3061-build-plan-thread-recommendations-please-read-before-posting/
Also, make sure to quote a post or tag a member when replying or else they won't get a notification that you replied to them.

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

×