Hacker News Reader

Comment by phamilton | Hacker News Reader

I used to do research on gene sequencing on a GPU. For small sets it was quite fast (it's arguably a O(n^4) algorithm, though really a O(n^2m^2)) but once you couldn't fit the data set on the GPU it was dead slow.

This would solve that problem nicely.

View on HN

Well, maybe. It would certainly improve. But you probaly still won't get anywhere near the GDDR5(etc) speeds from those memory accesses. Depending on the memory access pattern it might end up dropping down significantly. But you might still end up considering it "dead slow."

I think the main advantage is the GPU doesn't have to go through the CPU to get data.

i.e. the transaction goes from this: GPU -> CPU -> HDD -> GPU

to this: GPU -> HDD -> GPU

Sure, that part was clear. But the difference between "quite fast" for this application and "dead slow" might be that hundreds of GB/s with GDDRx/HBM is so much faster than anything else. Even the M.2 SSD onboard is probably only ~5GB/s. So if the walk pattern is something that is predictable enough by the GPU (usually pretty linear, but some memory features have 2d/3d locality), then you could crunch those numbers as fast as the SSD could deliver them. Now it becomes a question of how much computation/kernel time can you spend on a chunk of data?

Not with an NGFF SSD. You can have direct memory access to one. That said, internal GPU bus might or might not be faster than PCIe.

Oh and the fact that nobody write a specialised enough driver for a GPU to use this.

On a CPU you'd have the same bottlenecks unless you fit the entire dataset in memory (which, admittedly is a possibility). Assuming you can't get the full dataset in memory, apples-to-apples you should be able to see a significant speedup on the GPU over the CPU. Previously this was not the case for large datasets.

Except that Pascal class GPUs support true unified virtual memory so some minor code changes might provide mostly the same performance using memory-mapped files. Caveat: I don't own a Pascal class GPU yet.

Though unified virtual memory isn't limited to Pascal -- AMD's GCN architecture has supported it for years already, as do these products.

Unless I'm misunderstanding you, unified virtual memory (while cool) doesn't really help get around the bottleneck of sending data to the GPU.

It would send data from CPU space at PCIE speeds (~10.5 GB/s). What's the bandwidth from that SSD? <1 GB/s? Seems like plenty of margin...

That said, onboard SSD is cool if it can read without pre-empting the PCIE bus.

Reply on news.ycombinator.com