Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AMD reckons it can win the high-end GPU battle

techradar.com

Post reply on HN
Print view View original

51–60 of 160 posts

Re: AMD reckons it can win the high-end GPU battle

#51

Post by jamesblonde »

atq2119 wrote: post #43
jamesblonde wrote: post #30

The big problem right now is that ROCm, AMD's equivalent to Cuda for deep learning, is not supported for the consumer Radeon Navi cards. I had hoped it would be supported in ROCm 3.0 when that was released - but there has been no official word, which is worrying. If AMD are to compete in the GPU/DeepLearning space, it will be because they have hardware that is competitive (not even as good as Nvidia), low cost, and d…

I agree that the lack of (official) Navi support is just bad strategy on AMD's part, but I disagree with your numbers. Their goal needs to be performance parity or dominance, and then perhaps offer a slight discount. That's how they're succeeding with Zen on CPUs. Offering the kind of steep discount you write about is just bad business.

At the high-end (V100), Nvidia are price gouging, so there is room for a profitable strategy here where AMD significantly undercut Nvidia. My reasoning comes from my belief that AMD will not reach performance parity in the next generation - for deep learning, at least. If the new Arcturus GPU is competitive with Nvidia's new Ampere chip, that would be great. But they won't sell if they are not significantly cheaper. But they won't be significantly cheaper (Arcturus will be the "Enterprise" GPU). However, Big Navi will be much cheaper and can be used in a data center for deep learning without the legal problems that using a 3080Ti would have. I see this as the same "war" between Enterprise hard-disks from 10-15 years ago. Google came out and built their infrastructure on cheap commodity disks. But Dell, HPE and other vendors would not let you use them - they said they were inferior, etc. Commodity GPUs could "democratize" deep learning - ordinary servers could just include a couple of GPUs as a cheap add-on.

Re: AMD reckons it can win the high-end GPU battle

#52

Post by jamesblonde »

missosoup wrote: post #45
jamesblonde wrote: post #30

The big problem right now is that ROCm, AMD's equivalent to Cuda for deep learning, is not supported for the consumer Radeon Navi cards. I had hoped it would be supported in ROCm 3.0 when that was released - but there has been no official word, which is worrying. If AMD are to compete in the GPU/DeepLearning space, it will be because they have hardware that is competitive (not even as good as Nvidia), low cost, and d…

AMD has no chance until there's broad support for Cuda. Cuda is so deeply entrenched at this point that trying to fragment the space and compete against it is a non-starter. The consumer/gaming market is nothing compared to the GPGPU market today.

Cuda is a proprietary Nvidia framework. You don't need cuda support to train deep learning models on TensorFlow. ROCm also works - and is built by AMD and is open-source.

Re: AMD reckons it can win the high-end GPU battle

#53

Post by fxtentacle »

_bxg1 wrote: post #48
fxtentacle wrote: post #46

AMD is blocked by being not versatile enough. Their consumer cards support neither AI nor do they have good Linux drivers. Of course, the average gamer doesn't care, but developers do. And as the result, almost all the video games - their core sales driver - are not as thoroughly optimized and tested on AMD as they are on the (likely overpriced) NVIDIA cards that developers use in their own home rigs.

I don't think there's much overlap between AI developers/Linux users and game developers. Of course the former market matters too, but I really doubt it has an effect on the gaming market.

Since some studios use Linux data processing pipelines or dabble with deep learning for procedural content generation, I don't see how they could buy anything except for NVIDIA for their staff at the moment.

Re: AMD reckons it can win the high-end GPU battle

#54

Post by craftyguy »

atq2119 wrote: post #44
omegabravo wrote: post #29

Earlier quoted context omitted.

I have 2x 24" monitors on Ubuntu 18.04.03 with a Radeon RX 580 I can't watch a video on Youtube without substantial tearing on Firefox or Chrome. I've followed close to 20 guides and tried a number of drivers which all fail provide any benefit. Running glxgears will run at 7400FPS, with minor tearing. My monitors don't have FreeSync or whatever it's called since they're pretty old (over 10 years). Since most guides a…

Tearing issues tend to be problems of the application, compositor, or Xorg. Nothing the driver can do much about, unfortunately. Xorg in particular is pretty bad at tearing with multi-monitor setups.

Some DDX drivers handle tearing (implementing vsync). Tearing is basically not a problem at all on Wayland compositors since the spec addresses it directly (from what I understand)

Re: AMD reckons it can win the high-end GPU battle

#55

Post by mehhh »

omegabravo wrote: post #29
chao- wrote: post #12

Earlier quoted context omitted.

Same with a 580. Linux, many monitors, work, light gaming. Drivers are the first time I've ever not had issues with Linux graphics. Three years in I'm still waiting for the other shoe to drop. Where are the driver issues? I'm certain it was like a law of physics in Linux, but somehow, not this time, not for me?

I have 2x 24" monitors on Ubuntu 18.04.03 with a Radeon RX 580 I can't watch a video on Youtube without substantial tearing on Firefox or Chrome. I've followed close to 20 guides and tried a number of drivers which all fail provide any benefit. Running glxgears will run at 7400FPS, with minor tearing. My monitors don't have FreeSync or whatever it's called since they're pretty old (over 10 years). Since most guides a…

Switch to Wayland, your issue is that Xorg is not tear free by default and Ubuntu has chosen to stick with this legacy stack in 18.04: https://www.omgubuntu.co.uk/2018/01/xorg-will-default-displa...

Re: AMD reckons it can win the high-end GPU battle

#56

Post by missosoup »

jamesblonde wrote: post #52
missosoup wrote: post #45

Earlier quoted context omitted.

AMD has no chance until there's broad support for Cuda. Cuda is so deeply entrenched at this point that trying to fragment the space and compete against it is a non-starter. The consumer/gaming market is nothing compared to the GPGPU market today.

Cuda is a proprietary Nvidia framework. You don't need cuda support to train deep learning models on TensorFlow. ROCm also works - and is built by AMD and is open-source.

Tensorflow is only a small subset of the massive engineering investment that has been done into cuda by various orgs and open source projects. We have GPU databases, GPU accelerated analytics packages, GPU accelerated DSP, etc. All with hand-written cuda kernels.

Unless AMD can provide a toolchain which takes cuda code and generates whatever it takes to run it with performance parity to Nvidia cards, it'll never take off.

ROCm is a decade too late to simply coexist with cuda and battle for market/mind share. Unless they can offer a massive perf/cost advantage, there is no incentive for anyone to invest in porting their code.

Re: AMD reckons it can win the high-end GPU battle

#57

Post by dnautics »

winter_blue wrote: post #22
perceptronas wrote: post #11

Earlier quoted context omitted.

Agree, I dropped Linux due bad graphics drivers (AMD)

Are you really saying that AMD is worse than Nvidia on Linux ??? I have a Nvidia GPU (on my laptop), and I've had a truly horrible experience on Linux with Nvidia proprietary drivers. I was actually considering ditching my laptop, and buying a new one with AMD chipsets, due to the issues I've had on Linux due to the presence of the Nvidia graphics card. I occasionally play games on my Linux laptop, using Proton (a St…

For machine learning, Nvidias driver situation is indeed terrible, but AMDs is worse in this sense: If you get the wrong kernel version, driver version or somehow mismatch the cuda version due to a pinned dependency, you're up the creek. However, everything downstream will work. With AMD, you will get the drivers on your machine no problem but heaven help you if you try to do anything with them (there are regressions and memory leaks, etc.)

Re: AMD reckons it can win the high-end GPU battle

#58

Post by dnautics »

jamesblonde wrote: post #52
missosoup wrote: post #45

Earlier quoted context omitted.

AMD has no chance until there's broad support for Cuda. Cuda is so deeply entrenched at this point that trying to fragment the space and compete against it is a non-starter. The consumer/gaming market is nothing compared to the GPGPU market today.

Cuda is a proprietary Nvidia framework. You don't need cuda support to train deep learning models on TensorFlow. ROCm also works - and is built by AMD and is open-source.

Have you tried installing rocm? Last I checked two months ago you had to run a Ubuntu 16.04 container on Ubuntu 18.04 because the amd drivers were only compiled for bionic and the rocm software was only compiled for xenial.

Re: AMD reckons it can win the high-end GPU battle

#59

Post by sebow »

The "problem" with AMD GPUs was always the lack of software 'compute' libraries, poorly optimized drivers, etc.

AMD could really gain a huge advantage(both in enterprise and in more consumer spheres, like gaming/media production) by opening up the development of drivers and software libraries to the interested folk.

So far they have done this (arguably) very well on Linux(by comparison to nvidia),yet the ecosystem of nvidia is still very big by comparison.

While it's true nvidia has invested hugely in their development ecosystem, AMD could gain some territory by integrating their software compute libraries(like RocM) in programming languages like C/C++/Python and libraries like Tensorflow,pytorch,etc. could gain them a huge advantage in multiple industries.(I personally would argue this should be done firstly for their very successful and potent CPUs)

OpenCL has been truly great and it still a go-to when you talk about massive heterogenous systems, but they should abuse the fact that gaming cards could accelerate similar compute workloads.

Re: AMD reckons it can win the high-end GPU battle

#60

Post by alkonaut »

_bxg1 wrote: post #49

AMD cards don't yet have any raytracing support, right? I know it's a young technology, but they shouldn't be dragging their feet on it if they want to make a serious push for the top.

I’d like a card that is a 2080 but instead used all the extra transistors compared to a 1080 for traditional rendering, not raytracing. Raytracing might be a thing in the future but now it’s just a gimmick that uses silicon that could have churned out more frames instead.
Post reply on HN