AMD launches 7nm datacenter CPUs and GPUs with performance competitive to NVIDIA.

Advanced Micro Devices launched its first 7-nm CPU and GPU targeting the data center market, delivering comparable performance to Intel's 14-nm Xeon and Nvidia's 12-nm Volta. A single 7-nm Epyc x86 processor narrowly beats a system with two Intel Skylake Xeons in a rendering job, while AMD's 7-nm Vega GPU performs on par with a Nvidia V100 in inference tasks. AMD is now competitive with Nvidia and positioned to gain significant market share in the server space.

The 14-nm Epyc, launched in May 2017, boosted AMD's negligible 0.5% share of x86 servers to 1.5%, and with customer relationships back on track, the 7-nm version could push AMD up to "high single digits" in x86 server market share by mid-2019. The Zen-based x86 chips have already boosted AMD's overall microprocessor market share to 9.23% in the second quarter of this year.

AMD employed creative packaging to deliver a lower-cost Epyc design. A single module includes up to eight 7-nm processor die linked with AMD's Infinity fabric to a single 14-nm I/O chip with memory controller, extending the approach used in the 14-nm Epyc that utilizes four die on a single package.

The 7-nm AMD Vega delivers 25% more performance than the previous 14-nm chip and contains 13.2 billion transistors. The high-end MI60 version for GPU computing features 64 compute units, 4,096 streaming processors, and up to 32 GBytes HBM2 memory with PCIe Gen 4 support. The AMD chip delivers within 7% of a Nvidia Volta's performance with less than half the die area, providing a head-to-head alternative for machine learning and commercial graphics applications.

Vega will deliver 29.5 Tera FP16 operations per second for AI training, while achieving 59 TOPS for 8-bit integer and 118 TOPS for 4-bit integer tasks in inference jobs. AMD claims that its 7-nm x86 chips will beat Intel's 10-nm versions, now expected late next year.