| ▲ | Matmul on Blackwell: Part 2 – Using Hardware Features to Optimize Matmul(modular.com) |
| 21 points by robertvc 2 days ago | 12 comments |
| |
|
| ▲ | saagarjha 20 hours ago | parent | next [-] |
| Is anyone using Modular? Curious how you find it compares against the competitors in this space. |
| |
| ▲ | totalperspectiv 19 hours ago | parent | next [-] | | I have used Mojo quite a bit. It’s fantastic and lives up to every claim it makes. When the compiler becomes open source I fully expect it to really start taking off for data science. Modular also has its paid platform for serving models called Max. I’ve not used that but heard good things. | |
| ▲ | subharmonicon 20 hours ago | parent | prev [-] | | I’ve also been curious to see actual users compare/contrast their experiences with other options, but so far haven’t seen that. There seem to be enthusiasts who have experimented a bit and like what they see but I haven’t seen much else. |
|
|
| ▲ | subharmonicon 20 hours ago | parent | prev [-] |
| TLDR: In order to get good performance you need to use vendor-specific extensions that result in the same lock-in Modular has been claiming they will enable you to avoid. |
| |
| ▲ | totalperspectiv 19 hours ago | parent | next [-] | | I don’t follow your logic. Mojo can target multiple gpu vendors. What is the Modular specific lock in? | | |
| ▲ | subharmonicon 13 hours ago | parent | next [-] | | The blog post is about using an NVIDIA-specific tensor core API that they have built to get good performance. Modular has been pushing the notion that they are building technology that allows writing HW-vendor neutral solutions so that users can break free of NVIDIA's hold on high performance kernels. From their own writing: > We want a unified, programmable system (one small binary!) that can scale across architectures from multiple vendors—while providing industry-leading performance on the most widely used GPUs (and CPUs). | | |
| ▲ | totalperspectiv 5 hours ago | parent [-] | | They allow you to write a kernel for Nvidia, or AMD, that can take full advantage of the Hardware of either one, then throw a compile time if-statement in there to switch which kernel to use based on the hardware available. So, you can support either vendor with as-good-vendor-library performance. That’s not lock-in to me at least. It’s not as good as the compiler being able to just magically produce optimized kernels for arbitrary hardware though, fully agree there. But it’s a big step forward from Cuda/HIP. |
| |
| ▲ | smilekzs 19 hours ago | parent | prev [-] | | Not OP but I think this could be an instance of leaky abstraction at work. Most of the time you hand-write an accelerator kernel hoping to optimize for runtime performance. If the abstraction/compiler does not fully insulate you from micro-architectural details affecting performance in non-trivial ways (e.g. memory bank conflict as mentioned in the article) then you end up still having per-vendor implementations, or compile-time if-else blocks all over the place. This is less than ideal, but still arguably better than working with separate vendor APIs, or worse, completely separate toolchains. | | |
| |
| ▲ | imtringued 13 hours ago | parent | prev [-] | | Correct. There is too much architectural divergence between GPU vendors. If they really wanted to avoid vendor specific extensions in user level code, they would have gone with something that could be said to be loosely inspired by tiny grad (which isn't ready yet). Basically, you need a good description of the hardware and the compiler automatically generates the state of the art GEMM kernel. Maybe it's 20% worse than Nvidia's hand written kernels, but you can switch hardware vendors or build arbitrary fused kernels at will. |
|
