"The so-called CUDA moat could be getting narrower. ROCm, if you're not familiar, is a suite of software libraries and development tools, including HIP frameworks, that provides developers a low-level programming interface for running high-performance computing (HPC) and AI workloads on GPUs. The software stack is reminiscent in many ways of the CUDA runtime, but for AMD GPUs rather than Nvidia."
"ROCm 7 is arguably AMD's biggest update yet. Compared to ROCm 6, AMD says that customers can expect a roughly 3.5x uplift in inference performance on the MI300X. Meanwhile, the company says it has managed to boost the effective floating point performance achieved in model training by 3x. AMD claims that these software enhancements combined give its latest and greatest GPU, the MI355X, a 1.3x edge in inference workloads over Nvidia's B200 when running DeepSeek R1 in SGLang."
ROCm 7.0 delivers major software improvements that increase inference and training performance across AMD GPUs, including MI355X and MI300-series devices. The update adds datatype support, better compatibility with popular runtimes and frameworks, and hardware-specific optimizations in the ROCm runtime. Compared with ROCm 6, ROCm 7 yields roughly 3.5x inference uplift on the MI300X and about 3x higher effective floating-point training performance. These enhancements help the MI355X gain an inference edge over certain Nvidia B200 workloads while AMD cards also offer greater HBM3e capacity versus comparable Nvidia parts.
Read at Theregister
Unable to calculate read time
Collection
[
|
...
]