M60可以用于深度学习

你好
首先，让我们解决许可问题......最简单的方法是使用90天的评估。
你可以从这里得到它：
http://www.nvidia.com/object/grid-evaluation.html#utm_source=shorturl&utm_medium=referrer&utm_campaign=grid-eval
关于两个GPU的使用......与K80一样，它们都可以使用。
多GPU利用率的任何限制都取决于您的软件，而不是硬件:-)
您还可能希望确保两个GPU都处于“计算”模式，如果您正在使用AI，则不是“图形”。
您可以在注册评估后使用正确的M60驱动程序包获得的Linux Boot Utility来实现。
问候
本



以下为原文

Hi

Firstly, let's get that licensing resolved... The easiest way to deal with that is to use a 90 day evaluation. You can get that from here:

http://www.nvidia.com/object/grid-evaluation.html#utm_source=shorturl&utm_medium=referrer&utm_campaign=grid-eval

Regarding usage of both GPUs... As with a K80, they're both there to be used. Any limitation in multi-GPU utilization is down to your software, not the hardware :-)

You'll also more than likely want to make sure that both GPUs are in "Compute" mode, NOT "Graphics" if you're playing with AI. You can do that using the Linux Boot Utility you'll get with the correct M60 driver package after you've registered for the evaluation.

Regards

Ben

感谢您指出我正确的方向，非常感谢.. 90天试用将是完美的。
经过一番研究后，至少根据我的理解，我最近购买的GTX1080ti Strix比用于人工智能的M60更有利。
虽然我可以访问M60，但我会探索vGPU，这实际上更符合我的长期计划，但是我知道我可以双启动并将其用于“计算”而不受限制“下班后”是有用的。
再次感谢..，



以下为原文

Thanks for pointing me in the right direction, much appreciated.. 90 day trial will be perfect.

After a bit more research it appears, at least from my understanding, the GTX1080ti Strix I bought recently would be more beneficial then the M60 for AI purposes.

While I have access to the M60 I'll explore vGPU, this actually more closely aligns with my longer term plans, however knowing I can dual boot and leverage it towards 'compute' without restrictions 'after hours' is useful to know.

Thanks again..,

我没有任何争论。
你无法解决GRID仍然停留在麦克斯韦的简单事实，而NVIDIA制造的所有其他东西现在（并且自去年以来一直是？！）一代，而且最近，现在是2代架构。
那些较新的Pascal和Volta架构更高效，更强大，更快速，更适合您的工作，无论是GeForce，Quadro还是Tesla。
让我们知道你如何继续使用AI的东西:-)
问候
本



以下为原文



No arguments from me about that.

You can't get around the simple fact that GRID is still stuck on Maxwell, and everything else NVIDIA make is now (and has been since last year?!) a generation and just recently, now 2 generational architectures ahead. Those newer Pascal and Volta architectures are just more efficient, more powerful, faster and more relevant to what you're working with, doesn't matter whether it's GeForce, Quadro or Tesla.

Let us know how you get on with the AI stuff :-)


Regards

Ben

嗨，
我想问同样的问题，但关于特斯拉M6;
我认为这是大致相同的问题所以我不妨在这里问。
是M6，是网格架构，也意味着或适合深度学习研究？
我认为它们是专为虚拟桌面设计的，但客户告诉我，他们更容易购买。
您如何使用GPU进行网格深度学习？
谢谢！
马坦



以下为原文

Hi,

I wanted to ask the same question, but about the Tesla M6; I figured this is roughly the same question so I might as well ask here. Is the M6, and is the Grid architecture, also meant or suited for deep learning research?

I kind of assume they were designed for virtual desktops, but a customer tells me they are easier for them to buy.

How would you even use a GPU for deep learning over the grid?

Thanks!
Matan

只是为了使术语正确...“GRID”是放置在给定的特斯拉（当前为M10，M6，M60）（以及之前的Quadro（K1 / K2））GPU上的软件组件。
在最基本的形式中（如果你可以称之为），GRID软件目前用于在“图形”模式下使用GPU时创建FrameBuffer配置文件，这允许用户在访问相同的物理GPU时共享GPU的一部分FrameBuffer
。
GRID始终得到增强和开发，以提供更好的性能，功能增强和功能，这就是NVIDIA选择软件定义模型的原因，而不是硬件模型，因为存在更多限制。
不，M10，M6和M60不是特别适合AI。
但是，它们可以工作，而不是像其他GPU那样有效。
NVIDIA为特定工作负载和行业（技术）领域创建特定的GPU，因为每个领域都有不同的要求。
稍微偏离主题，但如果您想要AI的最佳（官方）资源，那么您正在寻找DGX-1（https://www.nvidia.com/en-us/data-center/dgx-1/
）或DGX站（https://www.nvidia.co.uk/data-center/dgx-station/）。
DGX-1将P100与NVLINK结合使用。
然而，P100最近已被V100取代。
由于这是一个全新的产品，DGX Station从V100开始，再次使用NVLINK。
请注意，他们使用“计算”​​聚焦GPU，而不是图形聚焦。
所以，回答你的问题，是的，你可以使用M6，如果你处于“图形”模式，你可以在VM之间共享FrameBuffer并更好地利用GPU。
然而，为了获得最佳性能，请引导他们使用更集中的GPU线。
问候



以下为原文

Just to get the terminology correct ... "GRID" is the software component that lays over a given set of Tesla (Currently M10, M6, M60) (and previously Quadro (K1 / K2)) GPUs. In its most basic form (if you can call it that), the GRID software is currently for creating FrameBuffer profiles when using the GPUs in "Graphics" mode, which allows users to share a portion of the GPUs FrameBuffer whilst accessing the same physical GPU. GRID is always being enhanced and developed to offer better performance, feature enhancements and functionality, and this is why NVIDIA have opted for a software defined model, as opposed to a hardware model, where there are far more limitations.

No, the M10, M6 and M60 are not specifically suited for AI. However, they will work, just not as efficiently as other GPUs. NVIDIA creates specific GPUs for specific workloads and industry (technological) areas of use, as each area has different requirements.

Slightly off topic, but if you want the best (official) resources for AI, then you're looking at a DGX-1 (https://www.nvidia.com/en-us/data-center/dgx-1/) or a DGX Station (https://www.nvidia.co.uk/data-center/dgx-station/). The DGX-1 uses P100s in combination with NVLINK. However, the P100s have very recently been replaced with V100s. As it is a brand new offering, the DGX Station starts with V100s and again, uses NVLINK. Note that they use "Compute" focused GPUs, not Graphics focused.

So, in answer to your question, yes, you can use the M6, and if you're in "Graphics" mode you can share FrameBuffer between VMs and make better use of the GPU. However for the best performance, guide them to a more focused GPU line.

Regards