from Hacker News

GpuScan and SSD-To-GPU Direct DMA

by matsuu on 9/18/16, 8:16 AM with 26 comments

• by exDM69 on 9/18/16, 10:22 AM

  There is no explanation how it works. Does it work on top of existing APIs in user space? Or is there a custom kernel driver bypassing user space?

  I've done some high throughput streaming from HD/SSD to GPU before, and it's pretty easy to beat the naive solution but getting the most out of it would require kernel space code.

  I was doing random access streaming of textures using memory mapped files for input and copying to persistent/coherent mapped pixel buffers on the CPU with memcpy with background threads. This was intended to take advantage of the buffer caches (works great when a page is reused) and intended for random access. If I would have been working on a sequential/full file upload, my solution would be entirely different.

  Edit: here's the source: https://github.com/kaigai/ssd2gpu

  It has a custom kernel module.

• by zokier on 9/18/16, 10:11 AM

  This is very interesting in the light of recent AMD announcement of their "Solid State Graphics", ie GPU with SSD ducktaped on: http://www.anandtech.com/show/10518/amd-announces-radeon-pro...

• by foobar2020 on 9/18/16, 10:14 AM

  This would be incredibly useful for distributed machine learning - imagine a Tensorflow implementation that almost entirely bypasses CPU.

• by witty_username on 9/18/16, 9:44 AM

  So, if I understand correctly, data is being loaded directly from the SSD to the GPU and then filtered by the GPU before the CPU handles the more difficult queries.

  Neat.

• by justinclift on 9/18/16, 1:02 PM

  This is very awesome. If further developed + made into a feasible option for PostgreSQL, this has potential to do interesting things to TPC benchmarks. :)

• by nl on 9/18/16, 10:35 AM

  See also https://developer.nvidia.com/gpudirect and to some extent https://en.wikipedia.org/wiki/NVLink.

  NVLink is in the Power9 servers Google is using.

• by carbocation on 9/18/16, 5:04 PM

  I'm really hoping that Optane delivers on the hype, in which case our durable storage could be just 10x slower than RAM. At least, I imagine that it would be really helpful for speeding up even this approach.

• by Razengan on 9/18/16, 6:01 PM

  I hope this brings us closer to widespread external GPUs, where you could use a slower-than-PCIe bus like Thunderbolt 3 or USB 3.1 to upload all assets to the EGPU's SSD during a one-time loading screen.

• by foobarbecue on 9/18/16, 5:25 PM

  Direct Direct Memory Access? That's pretty direct.

• by musha68k on 9/18/16, 10:00 AM

  Amazing results! We need more of that kind of thinking - GPU/SSD accelerate all the things!

• by MrBuddyCasino on 9/18/16, 12:22 PM

  Who is providing the DMA engine in this case? Has the GPU access to PCIe device memory?