Actualizado el 17/12/2024

icon Facebook icon Twiiter icon RSS icon EMAIL
Portada
  1. Portada
    2. >
  2. TecnonewsCat
    3. >
  3. Hyperscalers forge ahead with in-house semiconductor solutions

Hyperscalers forge ahead with in-house semiconductor solutions

Escrito por Agencias Externas el 18/06/2024 a las 15:24:26
595

Hyperscalers such as Meta, Microsoft, Google, and Amazon are strategically shifting towards developing proprietary semiconductor technologies in response to the soaring demand for Nvidia's GPUs in the burgeoning generative AI (GenAI) market. This strategic move is aimed at reducing dependency on Nvidia while fostering innovation and facilitating global expansion. Within this dynamic landscape, significant investments in AI technologies and strategic partnerships underscore the intense competition within the GenAI space, says GlobalData, a leading data and analysis company.


Nvidia, which sells graphic processing unit (GPU) semiconductors particularly suited to AI-specific workloads, has emerged as one of the clear winners in the nascent GenAI market as demand for its GPUs explodes. However, providers of GenAI services such as big cloud computing companies have been impacted by the massive financial burden of acquiring the expensive chips.


Beatriz Valle, Senior Enterprise Technology and Services Analyst at GlobalData, comments: “There is a significant imbalance between supply and demand when it comes to GPU processors, because GenAI models in general, and especially multimodal systems that produce images and videos, greatly benefit from the parallel processing capabilities of GPUs. As a result, the chips are expensive and in short supply.


“To counteract this trend, GenAI companies are coming to market with their own proprietary technologies to run these workloads. Google has its tensor processing unit (TPU) semiconductors and Amazon has its Inferentia and Trainium architectures. Meta, the social media company behind Facebook and Instagram, recently announced the next generation of custom-made chips to help power AI-driven rankings and recommendation ads on social media platforms.”


GlobalData analysis notes that money continues to flow in the lucrative GenAI space as hyperscalers jockey for position. Leading GenAI companies are ramping up investment in strategic partnerships across the world to remain competitive. Microsoft, now vying for a global presence to offer AI services across Microsoft Azure’s infrastructure, has recently invested $1.5 billion in new technology group G24, an Abu Dhabi (UAE) AI consortium. AWS is completing its final round of $4 billion in funding to partner Anthropic, striving to gain a foothold following Microsoft’s mega OpenAI investment.


Charlotte Dunlap, Research Director of Enterprise Technology and Services at GlobalData, comments: “Microsoft’s investment signals the start of a much larger AI technology role for the UAE. G24 is targeting various sectors for its AI solutions, including healthcare and energy. The deal is not without its complications. In return for the mega investment by Microsoft, G42 severed ties to China by eliminating hardware provisioning by Huawei systems, among other efforts in recent months to appease US concerns.”


GlobalData publishes GenAI newsletters on a quarterly basis as part of the Enterprise Technology and Services (ETS) syndicated research offering.

Facebook
Twitter
 Meneame
 Reddit



Ver más TecnonewsCat