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

Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

blogs.nvidia.com

Post reply on HN
Print view View original

311–320 of 347 posts

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#311

Post by pjmlp »

14113 wrote: post #308
pjmlp wrote: post #175

Earlier quoted context omitted.

Anyone is free to target PTX and do their own compiler on top. In fact, given that it is there since version 3, there are compilers available for almost all major programing languages, including managed ones. While OpenCL is a C world, and almost no one cares about the C++ extensions and even less vendors care about SPIR-V. Also the community doesn't seem to be bothered that for a long time, the only SYCL implementat…

This is wrong - triSYCL is roughly the same age as ComputeCpp, and hipSYCL is only slightly younger. There has been a lot of academic interest in SYCL, but as with any new technology (especially niche technologies) it's always going to take time to get people on board. Also, from a quick look at your profile, you seem to have quite a lot of comments criticizing or commenting on CodePlay. Do you have some sort of rela…

I wish all the luck to CodePlay, the more success the better for them.

They are well appreciated among game developers, given their background.

My problem is how Khronos happens to sell their APIs, and let everyone alone to create their own patched SDKs and then act surprised that commercial APIs end up winning the hearts of the majority.

The situation has hardly changed since I did my thesis with OpenGL in late 90's, porting a particles visualization engine from NeXTSTEP to Windows.

Nothing that compares with CUDA, Metal, DirectX, LibGNMX, NVN tooling.

Hence my reference to CodePlay, as for very long time their SDK was the only productive way to use SYCL.

Khronos likes to oversell the eco-system, and usually the issues and disparities across OEMs tend to be "forgotten" on their marketing materials.

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#312

Post by fulafel »

pjmlp wrote: post #277
fulafel wrote: post #276

Earlier quoted context omitted.

OpenCL used to be supported on Android (But not required on by Google). Currently, Vulkan _is_ required by Google[1] and is probably the future path to GPU compute. [1] https://www.androidpolice.com/2019/05/07/vulkan-1-1-will-be-...

OpenCL is only supported via hacks to install shared libraries into one own's device. Vulkan even if optional until Android 10, it is supported by the SDK since version 7, which is something that OpenCL never had. No serious Android developer would make their life even harder than it already is with the official APIs, by making use of an API that requires device owners to manually install libraries via ADB.

Hmm, is the OpenCL situation really that fringe on android? Eg this OpenCL info app[1] description says "Even though OpenCL™ isn't part of the Android platform, it's available on many recent devices. On Android it's usually used as a back-end for other frameworks like Renderscript. Some manufacturers are providing SDKs for developers to use OpenCL™ on Android. "

I addition to the mentioned PowerVR and Intel platforms there seems to be Android OpenCL also for Mali and Adreno GPUs.

[1] https://apkpure.com/opencl%E2%84%A2-info/com.xh.openclinfo

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#313

Post by lumost »

fluffything wrote: post #290
diffrinse wrote: post #240

Earlier quoted context omitted.

Would you agree that that your observations somewhat imply that a competitive free market is not a fit for all governable domains (and don't mistake governable for government there, we're still talking about shepherding of innovation)?

Early tech investments are risky, but if your competition has tech 10 years more advanced than yours, there is probably no amount of money that would allow you to catch up, surpass, and make enough profits to recover the investment, mainly because you can't buy time, and your competitor won't stop to innovate, they are making a profit and you aren't, etc. So to me the main realization here is that in tech, if one com…

This is a good callout, unlike manufacturing the supply chain is almost universally vertically integrated for large software projects. While it's possible to make a kit car that at least some people would buy, most of the big tech companies have reached the point of requiring hundreds of engineers for years to compete.

Caveat that time has shown that the monopolies tend to decay over time for various reasons, the tech world is littered with companies that grew too confident in their monopoly.

- Cisco - Microsoft Windows - IBM

etc.

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#314

Post by wegs »

solidasparagus wrote: post #130
wegs wrote: post #113

Earlier quoted context omitted.

You're making a lot of lousy assumptions. As a few points of reference: * My list comprehensions run over gigabytes of data (but sometimes 3 orders of magnitude bigger or smaller). Stream processing of big data. It's not deep learning, but it's slow and potentially deeply parallel. It would move to MIMD trivially, and SIMD with just a little bit of work. * There are programming languages which support models almost e…

You're right about my assumptions, I apologize. Without knowing specifically what you do, I can't say if GPUs would make sense. But I don't agree that many workloads could be moved to GPUs cost effectively. It's very hard to feed work in fast enough to keep the GPU busy enough to be cost effective given the limited amount of GPU memory you have to work with.

Well, there's a question of whether the GPU has to be fed fast enough. My CPU is nominally rated at around 150-200 gigaflops. My GPU is rated at about 5 teraflops. That's about a 30x speed difference (and there are obviously faster GPUs out there). That's enough to move me from compute-bound to IO-bound and make things a lot faster. Once I'm IO-bound, I'll obviously see no more performance increase, but I figure I'll get a good 5-10x before I get there.

Right now, code runs anywhere from a few seconds to overnight, depending on what I'm doing. I'll also mention I'm working on many projects, so that's not overhead I incur every day, just once in a while.

Moving to GPU would move that to running anywhere from more-or-less instantly to an hour, I figure, based on similar very back-of-the-envelope benchmarks and guesstimates. That's totally worth dropping $1000 on a new GPU, if that's all it took and things worked out-of-the-box. It'd pay for itself in a few programmer hours.

On the other hand, that's totally not worth weeks of programmer / dev-ops time for switching to a proprietary tool chain. An alternative there is to wait for my computation, or to optimize my code. Both of those seem cheaper than maintaining a GPGPU workflow, where GPGPU is right now. If GPGPU came batteries-included in Ubuntu+numpy, it'd be an entirely different story.

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#315

Post by wegs »

pjmlp wrote: post #278
wegs wrote: post #216

Earlier quoted context omitted.

It's not about blame. It's about getting to an ecosystem where GPGPU is used for things beyond deep learning, bitcoin mining, video encoding, and similar niche applications to one where I can fluidly use MIMD to speed up my JavaScript and Python with first-class language constructs to support that. If that happens: 1) We'll get back on some kind of curve where computer performance starts increasing again. 2) The GPU…

It is easy to talk about the proprietary practices done by NVidia, yet none of the other GPGU device makers that are on Khronos weren't able to offer a better experience. So Khronos has 140 members, about 10 of them producing hardware, and they can't provide a proper developer experience, with software that looks like EE toolchains of the 90's. The market has already exploded, and CUDA has won.

You're missing the point, and taking everything as an attack on NVidia. It's not an attack on NVidia, or have anything to do with Nvidia versus Khronos. It's clear you've got enough baggage there that I won't go there (not like I was trying to go there in the first place).

But the market hasn't exploded. Most computers have built-in Intel graphics, and most apps can't make use of GPU. NVidia won the battle with AMD, but lost the battle with Intel. GPUs are still for gamers, deep learning applications, bitcoin miners, video editors, and a few other niche applications.

Given that CPUs aren't increasing in speed, and GPGPU is failing to make in-roads, for most workloads, computers are only marginally faster than they were a decade or two ago. If GPGPU made inroads into general computing, we'd still be on a Moore's Law curve, but that hasn't happened.

That's the problem.

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#316

Post by wegs »

paulmd wrote: post #170
wegs wrote: post #23

Earlier quoted context omitted.

Yeah. That's been my general problem with adopting NVidia for anything. They make good hardware, but there's a lot of lock-in, and not a lot of transparency. That introduces business risk. I'm not in a position where I need GPGPU, but if there wasn't that risk, and generally there were mature, open standards, I'd definitely use it. The major breakpoint would be when libraries like Numpy do it natively, and better yet…

NVIDIA has no ethical or moral responsibility to give their competitors the benefit of software they have paid to develop in-house. It is probably a safe bet that you yourself do not develop your projects under the Affero GPL, and so on some level you agree with this. What you see as "ecosystem lock-in" is properly viewed as software that you pay a premium for as part of your purchase price, above and beyond the pric…

It's not a safe bet. I've build project under AGPL, and made plenty of money doing it. There are places where open is good business, and there are places where proprietary is good business, and there's everything in between. AGPL was nice since I could be open, which had huge market advantage, but release code which my competitors would /never/ take advantage of. It had, quite literally, zero downsides, and a lot of upsides.

There are projects where I do 100% proprietary too, and a mix. It's a business decision. It's not as stupid as proprietary=profit and open=charity. It's a business calculation in every case.

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#317

Post by wegs »

fluffything wrote: post #241
wegs wrote: post #122

Earlier quoted context omitted.

Yes. I won't bet my business on a one-vendor solution with a medium-sized community which might disappear at some point. If CuPy supported NVidia and AMD, and was folded into Numpy, I'd buy the biggest, beefiest GPU I could find overnight.

What technology would you bet your business on then? Today, you can write numpy code, and that runs on pretty much all CPUs from all vendors, with different levels of quality. A one line change allows you to run all numpy code you write on nvidia GPUs, which at least today, are probably the only GPUs you want to buy anyways. In practice, you would probably be also running your whole software stack on CPUs, at least f…

shrug I'll bet my business on waiting an extra 15 minutes for analytics code to run.

Seriously. There's little most businesses really needs that I couldn't do on a nice 486 running at 33MHz. Now, if a $5000 workstations gives even 5% improvement to employee productivity, that's an obvious business decision. That doesn't mean it's necessary for a business to work. So dropping $1000 on an NVidia graphics card, if things ran faster and there were no additional cost, would be a no-brainer.

There are additional costs, though.

And no, you can't just go back from faster to slower. Try running Ubuntu 20.04 on the 486 -- it won't go. Over time, code fills up resources available. If I could take a 2x performance hit, it'd be fine. But GPUs are orders-of-magnitude faster.

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#318

Post by wegs »

bigdict wrote: post #269
wegs wrote: post #122

Earlier quoted context omitted.

Yes. I won't bet my business on a one-vendor solution with a medium-sized community which might disappear at some point. If CuPy supported NVidia and AMD, and was folded into Numpy, I'd buy the biggest, beefiest GPU I could find overnight.

CuPy has experimental support for Rocm.

Awesome! I did not know that.

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#319

Post by threeseed »

fluffything wrote: post #108
wegs wrote: post #31

Earlier quoted context omitted.

It's irrelevant to researchers. Research operates on rapid cycles: prototype, publish, move on. It does impact businesses. It doesn't prevent adoption for e.g. deep learning, but I haven't seen e.g. GPU-based databases reach broad adoption, or many other places where MIMD/SIMD would reduce costs or improve performance. Using classical hardware is clearly cheaper than the business risk and engineering time of relying…

> but I haven't seen e.g. GPU-based databases reach broad adoption NVIDIA announced today that Apache Spark 3.0 is built on RAPIDS and showed some benchmarks claiming same perf at 1/5 the cost. Not your classical database application though.

No they have built a GPU accelerated XGBoost library.

Spark 3.0 is in preview stage already and there is no mention of RAPIDS anywhere in the code.

Re: Nvidia CEO Introduces Nvidia Ampere Architecture, Nvidia A100 GPU

#320

Post by adev_ »

fluffything wrote: post #291
adev_ wrote: post #178

Earlier quoted context omitted.

> While Google did developed a PTX backend for LLVM, the student that worked on that as part of a GSOC got later hired by NVIDIA, and ended up contributing the current NVPTX backend that clang uses today. You more or less reformalized what I said. It might become used one day behind a proprietary blob, rebranded blob of NVidia, but fact is that today, close to nobody use it for production in the wild and it is not ev…

> CUDA fortran, yes you said it, CUDA fortran. The rest is OpenACC. You can also mix C, OpenACC, and CUDA C, and compile to x86-64. So I'm really not sure about what point you are trying to make here. You were claiming that OpenACC and CUDA only runs on nvidia's hardware, yet I suppose you now agree that this isn't true I guess. I do agree that PGI is still nvidia owned, but there are other compilers that do what PGI…

> You were claiming that OpenACC and CUDA only runs on nvidia's hardware, yet I suppose you now agree that this isn't true I guess.

I do not think I ever said that OpenACC runs only on NVidia hardware. However CUDA I still affirm that CUDA runs only on NVidia hardware yes. For anything else, it is based on code converter in best case.

Post reply on HN