Qualcomm, Intel, and Google have reportedly formed a new “strategic” coalition in an attempt to dethrone NVIDIA from the AI markets.
It Takes Not One But Three Big Tech Companies Such as Qualcomm, Intel & Google, To Have A Chance To Dethrone NVIDIA’s CUDA Supremacy In AI
Now, this does sound interesting, and it is probably a development to watch out for since we haven’t seen such a massive collaboration among companies to target a single entity. NVIDIA’s dominance in the AI market has shocked competitors to a vast extent since such financial growth and adoption weren’t seen previously. NVIDIA has gobbled up the bulk of the share of AI in tech industry, leaving no space for competitors to fill in, and this has troubled many of the firms who have now formulated a united front against NVIDIA.
Reuters reports that Qualcomm, Google, and Intel have reportedly teamed up to come for Team Green’s market share uniquely, as they plan on providing an alternative to potential clients for NVIDIA’s CUDA software platform, which has maintained its dominance for quite some time now. Instead of coming directly for NVIDIA’s hardware arsenal, the firms involved in the coalition believe that CUDA is the place to target, and they plan on doing so by leveraging the capabilities present in Intel’s oneAPI platform, which is the only CUDA alternative that could potentially out throw it.
We’re actually showing developers how you migrate out from an Nvidia platform.
– Vinesh Sukumar, Qualcomm’s head of AI
It’s about specifically – in the context of machine learning frameworks – how do we create an open ecosystem, and promote productivity and choice in hardware.
– Google’s director and chief technologist of HPC Bill Margo
For a quick history on Intel’s oneAPI, it is a unified programming model targeted explicitly at heterogeneous computing. It provides an interface to its users, ultimately allowing them to leverage it on various computing accelerator architectures, including GPUs, AI accelerators, and FPGAs. Intel’s oneAPI is ahead of NVIDIA’s CUDA in its ability to be more open towards architectures, and it allows more portability among developers when switching across various architectures. However, oneAPI hasn’t reached the optimization level compared to CUDA, but the firms in the coalition plan to get the same performance through rapid enhancements.
Exciting times are ahead for the AI markets, especially for NVIDIA and its competitors, as by the looks, NVIDIA is likely to see some heated competition in the AI segment moving forward but the company isn’t slowing down its progress in this segment as made evident during GTC 2024.
News Source: Reuters
