According to fudzilla, a number of well-informed sources told them that there was a large-scale uprising in Silicon Valley in response to the proposed NVIDIA arm acquisition. Fudzilla further said that they had spoken with people from a number of technology companies and that according to the discussion, everyone except arm and NVIDIA thought the deal would be bad for the industry. Regulators from China and the European Union may object strongly, but it is a matter for the future, as it traditionally takes 18 to 24 months to complete a major merger.
The report further pointed out that an exciting alliance of Intel, Qualcomm, Tesla and other major players in the chip market is under discussion to take concerted action and express their concerns to the relevant authorities in the United States and the world. According to reports, the relevant progress of the alliance will be announced soon.
Apple has also been involved in discussions, the report said, but Apple may withdraw from the exciting group because they may be with ARM
AMD and NVIDIA at NVIDIA's GPU Technology Conference and arm's developer summit stressed how great the merger would be.
But the current Sino US relations, Huawei issues and the relationship between the United States and the European Union have seriously troubled the deal. More importantly, arm licensees don't want dominant chip companies like NVIDIA to see their technology and determine the future of arm's roadmap.
The report also said that every large arm license holder began to explore alternatives, and risc-v stood out.
Will the British government step in to stop the deal?
At the same time, according to thisismoney reports, British ministers at all levels said they were
But yesterday, officials paved the way for possible intervention. The spokesman said:
Former British Deputy Prime Minister Lord Madsen joined the call for an investigation, saying in an interview City AM the London newspaper that the acquisition was
However, NVIDIA said arm would still license chip designs to competitors and vowed to invest in arm's UK operations.
To reach a deal, NVIDIA formed an alliance?
In order to close the deal, NVIDIA seems to have other actions besides various commitments.
According to foreign media HPC, in order to complete the landmark transaction of acquisition of chip IP supplier arm Ltd., NVIDIA has focused on its partner ecosystem to seek to expand its broad business scope to AI, HPC and network edge. Not surprisingly, these platform alliances are using arm based architectures.
At this week's arm Developers Conference, NVIDIA's chief executive, Jensen Huang, expanded his chip partner roster while highlighting their strategy to retain and integrate the arm business model. In his efforts to create synergies between AI hardware and software, Huang mentioned partnerships with Fujitsu, Marvell technologies and amber computing, an arm based chip start-up founded in 2018 by Renee James, a former Intel executive.
The three, Huang say, combine Arm platforms with a range of CPU,GPU and new datacenter processors (DPU).(Agencies
Nvidia collaboration with Fujitsu focuses on the Arm -based A64FX processor used by Japanese suppliers in supercomputers, which currently ranks top in the global supercomputing rankings. Huang said:
NVIDIA said its engineers were working to support scalable vector extension (SVE) in Fujitsu a64fx.
In November, Marvell announced that its arm based thunderx series server processor will support NVIDIA GPU. Marvell also ported NVIDIA cuda-x AI tools, HPC library, AI framework and software development to its thunderx platform.
NVIDIA said it would target GPU accelerated 64 bit server processors based on arm v8-a architecture for HPC, cloud and edge computing. But Huang did not provide more details this week.
NVIDIA released the ampere A100 GPU in May. In order to promote more synergy, Huang also announced further cooperation with an arm chip start-up of the same name.
Since its launch two years ago, ampere computing has been adding more arm based kernels to its cloud native Altra processors, and announced its 128 core design in June. It is expected to start sending samples of the chip by the end of 2020.
Duncan Poole, director of NVIDIA Platform Alliance, pointed out that ampere computing and NVIDIA will use cuda-x tools to combine arm based CPU and its GPU architecture to jointly launch a reference design. Recently, NVIDIA has expanded its collaboration with ampere computing Mt.Jade Server platform to cloud game field. The Altra based system consists of two arm based 80 core SOC, four NVIDIA T4 GPUs and one NVIDIA mellanox bluefield-2 DPU.
Most of the chip partnerships came months before NVIDIA announced its $40 billion acquisition of arm. These alliances and their deal with mellanox provide clues to Huang Renxun's overall strategy of dominating AI hardware, software and development tools, challenging Intel's server dominance in the data center, and extending NVIDIA's influence to HPC.
British chip IP suppliers are missing:
Arm CEO Simon