Intel Flex graphics cards are positioned for video streaming and cloud gaming data center workloads.

Intel's datacenter-focused Flex-series GPUs, codenamed Arctic Sound, are now ready for deployment, with computer makers expected to begin shipping systems over the next few months. The GPUs target media and game streaming applications where stream density outweighs computational power, distinguishing them from Intel's AI and HPC-focused Ponte Vecchio GPUs. The Flex-series comes in two form factors: the 75W Flex 140 with 12GB of memory and the 150W Flex 170 with 16GB, both featuring PCIe 4.0 interfaces and passive cooling in single-slot configurations. The Flex 140's half-height design enables deployment of up to 10 cards in high-density systems.

Intel claims significant performance advantages in media transcoding workloads. The Flex 140 offers five times the media transcode throughput and twice the decoding performance of Nvidia's A10 GPUs, capable of supporting up to 36 concurrent 1080p 60fps streams or eight 4K 60fps streams. A 10-card host can deliver 360 simultaneous 1080p streams or 80 4K streams using the HEVC H.265 format. For cloud gaming, a single Flex 170 can deliver 68 game streams at 720p 30fps and is validated on nearly 90 popular Android titles. However, the Flex-series falls short for AI and HPC applications: the Flex 140 delivers peak performance of 8 teraflops in FP32 calculations, on par with Nvidia's $450 MSRP A2000, while the Flex 170 offers 16 teraflops of FP32—roughly half the 31.2 teraflops claimed by Nvidia's A10.

A key differentiator is native support for AV1 encoding, which Intel says is the first to deliver in a datacenter form factor. The royalty-free codec, developed under the Alliance for Open Media by companies including Amazon, Netflix, and Google, promises 30 percent bandwidth savings compared to the HEVC format. Intel contends this stream density capability significantly reduces operating costs for media streaming and Android game streaming workloads, with applicability for streaming AI and metaverse use cases.

Nvidia reported that in the three months to July 31 (the second quarter of its fiscal 2023), it recorded $6.7 billion in revenue, up three percent year-on-year, and a profit of $656 million, down 72 percent. Gaming sales fell 33 percent to $2 billion, while Datacenter revenue increased 61 percent to $3.8 billion. The Flex-series GPUs will be available from more than 15 OEM partners, including Cisco, Dell Tech, Hewlett Packard Enterprise, Supermicro, Inspur, and H3C in the coming months.