The Chief Executive Officer of AMD, Lisa Su, committed not to offer any projections on expanding the market share after the release of the first-generation EPYC server CPUs. Jean Hu, the organisation’s chief financial officer, did not impose such limits while she was attending the Barclays technology conference on December 7. Instead, she openly gave facts and forecasts about this particular field.
According to her, AMD processors already occupy 29.7% of the market share in the data centre technology area in terms of value. They already account for almost fifty per cent of the server market in the cloud segment; however, the situation is somewhat more complicated in the corporate segment of the server industry. EPYC processors still only account for 15–16% of the market in this area, and to improve this position, AMD needs to devote more resources to promoting its products in the market.
Because the total cost of ownership for server processors is a significant factor for corporate purchasers, the company is trying to target customers with a low total cost of ownership. AMD will introduce the Turin generation of processors next year. These processors are anticipated to blend performance and operational operating expenditures successfully.
In addition, Jean Hu offered his thoughts on the recent policy shifts that NVIDIA has implemented about the regularity with which new generations of computational accelerators are released for the server market. To bring to your attention, AMD’s main rival is now getting ready to launch a new generation of accelerators annually instead of the old pattern of releasing them every two years.
The representative of AMD explained that the market is the primary factor behind the rapid pace at which new chips are being introduced, and this pattern is observed for both NVIDIA and AMD. Considering that AMD unveiled its initial solutions for this market segment in 2020, it has introduced numerous generations of accelerators over the past four years. Currently, AMD is redistributing its resources to devote a substantially greater amount of money and expertise to the development of accelerators for data centres than was previously the case.
