Nvidia announced its first Denver CPU - a dual-core mobile chip expected in 2H14 - under the Tegra K1 moniker in January. CEO Jen-Hsun Huang has previously said Nvidia is "seeing a lot of interest in putting something like Tegra in micro servers." This week, Nvidia showed off a solution that pairs its Tesla HPC GPUs with 3rd-party ARM (ARMH) server CPUs.
Meanwhile, a source tells the WSJ Samsung has decided to drop its ARM server CPU efforts. The company was previously reported to be working on ARM server parts with the help of AMD-poached talent.
AppliedMicro (AMCC) and Cavium (CAVM) must be pleased. Each has bet big on the budding ARM server CPU market, hoping to tap into growing interest in using them in microservers, application-specific hardware (e.g. networking or security gear), and hyperscale data centers. AMD, Marvell, and TI have also announced or launched parts.
The WSJ also reports Qualcomm (QCOM) is "working on [ARM] server chips," though no announcements have been made yet.
CEO Steve Mollenkopf hinted at CES Qualcomm, which has a huge R&D budget at its disposal, could go after the market. "I think there's going to be a tremendous amount of growth in computing and resources dedicated to supporting the cloud ... We look at that as an opportunity."