3 ms·
Agree. In my market (AI & physics simulation) AMD has been the default choice for years. And with lots of investment into SSE & AVX performance gains, I don't s
by fxtentacle 6y ago
Agree. In my market (AI & physics simulation) AMD has been the default choice for years. And with lots of investment into SSE & AVX performance gains, I don't see these tools switching to ARM anytime soon.
Edit: I didn't know before, but AMD also already has 20% of the notebook market.
https://seekingalpha.com/news/3615331-amd-to-take-20-share-of-notebook-processor-market-in-2020 https://seekingalpha.com/news/3615331-amd-to-take-20-share-o...
So that makes it even more weird to say "expand Apple’s share of the PC market — at Intel’s expense" and not mention AMD.
- read_if_gay_ 6y agoWhy is it weird? If AMD is competitive with Apple and Intel isn’t, then it’s absolutely at Intel’s expense.
- fxtentacle 6y agoHmm I kind of have to agree with that. Apple is pushing into the market, AMD will stand their ground, so it's Intel that will lose the most.
- Dracophoenix 6y agoIs that so? I would have thought MKL and CUDA would have kept AMD from dominating in AI and simulations for at least the next 3-5 years. That is unless you're working supercomputers.
- fxtentacle 6y agoCUDA yes. I'm using an AMD Ryzen for data pre-processing and then an NVIDIA GPU for the actual AI training. The TensorFlow pre-processing pipeline is heavily optimized for parallelism, so it works perfectly on AMD CPUs with lots of cores. As for MKL, yes it is impressive for some applications like real-time audio processing, but AFAIK with TensorFlow MKL is only used if you don't use a GPU, so pretty much never.