What PC hardware is needed to try out GPU programming? - gpu

I am interested to try out GPU programming. One thing not clear to me is, what hardware do I need? Is it right any PC with graphics card is good? I have very little knowledge of GPU programming, so the starting learning curve is best not steep. If I have to make a lot of hacks just in order to run some tutorial because my hardware is not good enough, I'd rather to buy a new hardware.
I have a retired PC (~10 year old) installed with Ubuntu Linux, I am not sure what graphics card it has, must be some old one.
I am also planning to buy a new sub-$500 desktop which to my casual research normally has AMD Radeon 7x or Nvidia GT 6x graphics card. I assume any new PC is good enough for the programming learning.
Anyway any suggestion is appreciated.

If you want to use CUDA, you'll need a GPU from NVidia, and their site explains the compute capabilities of their different products.
If you want to learn OpenCL, you can start right now with an OpenCL implementation that has a CPU back-end. The basics of writing OpenCL code targeting CPUs or GPUs is the same, and they differ mainly in performance tuning.
For GPU programming, any AMD or NVidia GPU made in the past several years will have some degree of OpenCL support, though there have been some new features introduced with newer generations that can't be easily emulated for older generations.
Intel's integrated GPUs in Ivy Bridge and later support OpenCL, but Intel only provides a GPU-capable OpenCL implementation for Windows, not Linux.
Also be aware that there is a huge difference between a mid-range and high-end GPU in terms of compute capabilities, especially where double-precision arithmetic is supported. Some low-end GPUs don't support double-precision at all, and mid-range GPUs often perform double-precision arithmetic 24 times slower than single-precision. When you want to do a lot of double-precision calculations, it's absolutely worth it to get a compute-oriented GPU (Radeon 7900 series or GeForce Titan and up).
If you want a low-end system with a non-trivial GPU power, you best bet at the moment is probably to get a system built around an AMD APU.

Related

Hardware for Deep Learning

I have a couple questions on hardware for a Deep Learning project I'm starting, I intend to use pyTorch for Neural Networks.
I am thinking about going for an 8th Gen CPU on a z390 (I'll wait month to see if prices drop after 9th gen CPU's are available) so I still get a cheaper CPU that can be upgraded later.
Question 1) Are CPU cores going to be beneficial would getting the latest Intel chips be worth the extra cores, and if cores on CPU will be helpful, should I just go AMD?
I am also thinking about getting a 1080ti and then later on, once I'm more proficient adding two more 2080ti's, I would go for more but it's difficult to find a board to fit 4.
Question 2) Does mixing GPU's effect parallel processing, Should I just get a 2080ti now and then buy another 2 later. And a part b to this question do the lane speeds matter, should I spend more on a board that doesn't slow down the PCIe slots if you utilise more than one.
Question 3) More RAM? 32GB seems plenty. So 2x16gb sticks with a board that can has 4 slots up to 64gb.
The matter when running multi GPU is also the number of available PCIe lanes. If you may go for up to 4 GPUs, I'd go for AMD Threadrippers for the 64 PCIe lanes.
For machine learning in a general manner, core & thread count is quite important, so TR is still a good option, depending on the budget of course.
Few poeple mention that running an instance on each GPU may be more interesting, if you do so, mising GPUs is not a problem.
32GB of ram seems good, no need to go for 4 sticks if your CPU does not support quad channel indeed.

How can I speed up deep learning on a non-NVIDIA setup?

Since I only have an AMD A10-7850 APU, and do not have the funds to spend on a $800-$1200 NVIDIA graphics card, I am trying to make due with the resources I have in order to speed up deep learning via tensorflow/keras.
Initially, I used a pre-compiled version of Tensorflow. InceptionV3 would take about 1000-1200 seconds to compute 1 epoch. It has been painfully slow.
To speed up calculations, I first self-compiled Tensorflow with optimizers (using AVX, and SSE4 instructions). This lead to a roughly 40% decrease in computation times. The same computations performed above now only take about 600 seconds to compute. It's almost bearable - kind of like you can watch paint dry.
I am looking for ways to further decrease computation times. I only have an integrated AMD graphics card that is part of the APU. (How) (C/c)an I make use of this resource to speed up computation even more?
More generally, let's say there are other people with similar monetary restrictions and Intel setups. How can anyone WITHOUT discrete NVIDIA cards make use of their integrated graphics chips or otherwise non-NVIDIA setup to achieve faster than CPU-only performance? Is that possible? Why/Why not? What needs to be done to achieve this goal? Or will this be possible in the near future (2-6 months)? How?
After researching this topic for a few months, I can see 3.5 possible paths forward:
1.) Tensorflow + OpenCl as mentioned in the comments above:
There seems to be some movement going on this field. Over at Codeplay, Lukasz Iwanski just posted a comprehensive answer on how to get tensorflow to run with opencl here (I will only provide a link as stated above because the information might change there): https://www.codeplay.com/portal/03-30-17-setting-up-tensorflow-with-opencl-using-sycl
The potential to use integrated graphics is alluring. It's also worth exploring the use of this combination with APUs. But I am not sure how well this will work since OpenCl support is still early in development, and hardware support is very limited. Furthermore, OpenCl is not the same as a handcrafted library of optimized code. (UPDATE 2017-04-24: I have gotten the code to compile after running into some issues here!) Unfortunately, the hoped for speed improvements ON MY SETUP (iGPU) did not materialize.
CIFAR 10 Dataset:
Tensorflow (via pip ak unoptimized): 1700sec/epoch at 390% CPU
utilization.
Tensorflow (SSE4, AVX): 1100sec/epoch at 390% CPU
utilization.
Tensorflow (opencl + iGPU): 5800sec/epoch at 150% CPU
and 100% GPU utilization.
Your mileage may vary significantly. So I am wondering what are other people getting relatively speaking (unoptimized vs optimized vs opencl) on your setups?
What should be noted: opencl implementation means that all the heavy computation should be done on the GPU. (Updated on 2017/4/29) But in reality this is not the case yet because some functions have not been implemented yet. This leads to unnecessary copying back and forth of data between CPU and GPU ram. Again, imminent changes should improve the situation. And furthermore, for those interested in helping out and those wanting to speed things up, we can do something that will have a measurable impact on the performance of tensorflow with opencl.
But as it stands for now: 1 iGPU << 4 CPUS with SSE+AVX. Perhaps beefier GPUs with larger RAM and/or opencl 2.0 implementation could have made a larger difference.
At this point, I should add that similar efforts have been going on with at least Caffe and/or Theano + OpenCl. The limiting step in all cases appears to be the manual porting of CUDA/cuDNN functionality to the openCl paradigm.
2.) RocM + MIOpen
RocM stands for Radeon Open Compute and seems to be a hodgepodge of initiatives that is/will make deep-learning possible on non-NVIDIA (mostly Radeon devices). It includes 3 major components:
HIP: A tool that converts CUDA code to code that can be consumed by AMD GPUs.
ROCk: a 64-bit linux kernel driver for AMD CPU+GPU devices.
HCC: A C/C++ compiler that compiles code into code for a heterogeneous system architecture environment (HSA).
Apparently, RocM is designed to play to AMDs strenghts of having both CPU and GPU technology. Their approach to speeding up deep-learning make use of both components. As an APU owner, I am particularly interested in this possibility. But as a cautionary note: Kaveri APUs have limited support (only integrated graphcs is supported). Future APUs have not been released yet. And it appears, there is still a lot of work that is being done here to bring this project to a mature state. A lot of work will hopefully make this approach viable within a year given that AMD has announced their Radeon Instinct cards will be released this year (2017).
The problem here for me is that that RocM is providing tools for building deep learning libraries. They do not themselves represent deep learning libraries. As a data scientist who is not focused on tools development, I just want something that works. and am not necessarily interested in building what I want to then do the learning. There are not enough hours in the day to do both well at the company I am at.
NVIDIA has of course CUDA and cuDNN which are libaries of hand-crafted assembler code optimized for NVIDIA GPUs. All major deep learning frameworks build on top of these proprietary libraries. AMD currently does not have anything like that at all.
I am uncertain how successfully AMD will get to where NVIDIA is in this regard. But there is some light being shone on what AMDs intentions are in an article posted by Carlos Perez on 4/3/2017 here. A recent lecture at Stanford also talks in general terms about Ryzen, Vega and deep learning fit together. In essence, the article states that MIOpen will represent this hand-crafted library of optimized deep learning functions for AMD devices. This library is set to be released in H1 of 2017. I am uncertain how soon these libraries would be incorporated into the major deep learning frameworks and what the scope of functional implementation will be then at this time.
But apparently, AMD has already worked with the developers of Caffe to "hippify" the code basis. Basically, CUDA code is converted automatically to C code via HIP. The automation takes care of the vast majority of the code basis, leaving only less than 0.5% of code had to be changed and required manual attention. Compare that to the manual translation into openCl code, and one starts getting the feeling that this approach might be more sustainable. What I am not clear about is where the lower-level assembler language optimization come in.
(Update 2017-05-19) But with the imminent release of AMD Vega cards (the professional Frontier Edition card not for consumers will be first), there are hints that major deep learning frameworks will be supported through the MIOpen framework. A Forbes article released today shows the progress MiOpen has taken over just the last couple of months in terms of performance: it appears significant.
(Update 2017-08-25) MiOpen has officially been released. We are no longer talking in hypotheticals here. Now we just need to try out how well this framework works.
3.) Neon
Neon is Nervana's (now acquired by Intel) open-source deep-learning framework. The reason I mention this framework is that it seems to be fairly straightforward to use. The syntax is about as easy and intuitive as Keras. More importantly though, this framework has achieved speeds up to 2x faster than Tensorflow on some benchmarks due to some hand-crafted assembler language optimization for those computations. Potentially, cutting computation times from 500 secs/epoch down to 300 secs/epoch is nothing to sneeze at. 300 secs = 5 minutes. So one could get 15 epochs in in an hour. and about 50 epochs in about 3.5 hours! But ideally, I want to do these kinds of calculations in under an hour. To get to those levels, I need to use a GPU, and at this point, only NVIDIA offers full support in this regard: Neon also uses CUDA and cuDNN when a GPU is available (and of course, it has to be an NVIDIA GPU). If you have access other Intel hardware this is of course a valid way to pursue. Afterall, Neon was developed out of a motivation to get things to work optimally also on non-NVIDIA setups (like Nervana's custom CPUs, and now Intel FPGAs or Xeon Phis).
3.5.) Intel Movidius
Update 2017-08-25: I came across this article. Intel has released a USB3.0-stick-based "deep learning" accelerator. Apparently, it works with Cafe and allows the user perform common Cafe-based fine-tuning of networks and inference. This is important stressing: If you want to train your own network from scratch, the wording is very ambiguous here. I will therefore assume, that apart from fine-tuning a network, training itself should still be done on something with more parallel compute. The real kicker though is this: When I checked for the pricing this stick costs a mere $79. That's nothing compared to the cost of your average NVIDIA 1070-80(ti) card. If you merely want to tackle some vision problems using common network topologies already available for some related tasks, you can use this stick to fine tune it to your own use, then compile the code and put it into this stick to do inference quickly. Many use cases can be covered with this stick, and for again $79 it could be worth it. This being Intel, they are proposing to go all out on Intel. Their model is to use the cloud (i.e. Nervana Cloud) for training. Then, use this chip for prototype inference or inference where energy consumption matters. Whether this is the right approach or not is left for the reader to answer.
At this time, it looks like deep learning without NVIDIA is still difficult to realize. Some limited speed gains are difficult but potentially possible through the use of opencl. Other initiatives sound promising but it will take time to sort out the real impact that these initiatives will have.
If your platform supports opencl you can look at using it with tensorflow. There is some experimental support for it on Linux at this github repository. Some preliminary instructions are at the documentation section of of this github repository.

How fast is PhysX on GPU compared to physics engines on CPU?

I have an application that is written to use the Bullet physics engine. I am running it on an Intel i7 2600K CPU with 8 cores. The application has to process millions of chunks of physics work, each of which can be done independently. It currently runs with 8 processes, each process working through its quota of the total independently. In summary, this work has a lot of easy parallelism.
Assuming that I can acquire the best NVIDIA consumer graphics card (say Titan), what is the ballpark improvement in the physics engine performance I can see by switching from Bullet on CPU to Physx on GPU? That is, approximately how much faster will this application run if rewritten for Physx?
I found a few papers that compare the result quality between Bullet and Physx, but could not find anything about the performance comparison.
Pierre Terdimann has done an extensive series of performance comparisons between Bullet 2.81 and PhysX 2.8.4, 3.2 and 3.3 here. These are comparisons between Bullet and PhysX, both running on CPU. It can be seen that the performance difference between the two is dependent on what features of the engine are being used. For a few features, the performance is about the same, while for most others there is a 3-5x speedup.
He also mentions in the addendum that not all physics features have been ported to PhysX on GPU. Cloth and particles can be accelerated on GPU, while rigid bodies is being currently ported to GPU, in a feature called GPU Rigid Bodies (GRB). If there is a feature that is GPU accelerated, then you can expect it to be faster than on CPU, but by how much is not clear.
I found this, it's not a comparison against any specific CPU physics engine but one hopes they are comparing like with like and running PhysX on the CPU.
So it's rather unspecific and from a FAQ by the makers of PhysX so take with a pinch of salt.
From here:
Running PhysX on a mid-to-high-end GeForce GPU will enable 10-20 times
more effects and visual fidelity than physics running on a high-end
CPU.
Lets say physx is doing particle interactions such as gravity of fluid movement. Then the cache control is very important since they are emberassingly parallel. You cannot directly control your CPU's cache but you can access to cache of titan which makes it maybe 100x faster than a 8-thread cpu.
If it is not so parallel and has many branching and doesnt have exhausting computations then it is around 10x-5x speedup(or whatever bandwidth ratio of graphics ram /main RAM).

Is it possible to do GPU programming if I have an integrated graphics card?

I have an HP Pavilion Laptop, it's so-called graphics card is some sort of integrated NVIDIA driver running on shared memory. To give you an idea of its capabilities, if a videogame was made in the last 5 years at a cost of more than a couple million dollars, it just won't be playable on my computer.
Anyways, I was wondering if I could do GPU programming, like CUDA, on this thing. I don't expect it to be fast, I'd just like to get the experience and not make my laptop catch fire in the meanwhile.
Find out what GPU your laptop is, and compare it against this list: http://en.wikipedia.org/wiki/CUDA#Supported_GPUs. Most likely, CUDA will not be supported.
This doesn't necessarily prevent you from doing "GPU programming", however. If the GPU supports fragment and vertex shaders, you can use the fixed pipeline to send data to the card (for example, through texture data) and do your processing in a fragment shader. You will then do a read from the pixel buffer to get the data back into system memory. Though hackish, this approach was quite popular until CUDA and other frameworks like OpenCL were introduced.

AMD Fusion how can CPU and GPU share the same memory?

AMD announced it's Fusion platform some time ago. Having read a bit about it I'm both excited and sceptic. For example it should make it possible that GPUs and CPUs share the same memory. (and the GPU and CPU are both in the same package) Now since GPUs have a much higher memory bandwidth (around 10x the bandwidth a CPU has) and that the way CPUs and GPUs use cache is fundamentally different, the question arises how the heck can they do this? I wonder if any details are known.
I have also searched for some detailed info on how this APU technically works, but haven't found anything better than AMD's whitepaper on the subject, which, in a slightly marketing-wise tone, does present a lot of good info.
By using high-bandwidth dual ported RAM. AMD is happy to explain.
Here is a good paper (from AMD) explaining the different memory access patterns in an APU
http://amddevcentral.com/afds/assets/presentations/1004_final.pdf
CPUs and GPUs have been sharing memory for years. Fusion is no different, in fact it's far from the first instance of a GPU being integrated into a general-purpose CPU core.
Like all other such solutions, it'll be fine for casual use, but it'll be far from cutting-edge 3D acceleration.