ARM CSS For Mobile 2 Platform Could Introduce DLSS 5-Like Neural Graphics

ARM CSS For Mobile 2 1

While NVIDIA is just starting to take its steps with DLSS 5, chipmaker ARM is preparing to offer something similar via its CSS for Mobile 2 mobile compute platform. And to do this, it’s introducing a combination of a new C2-Ultra CPU and Mali G2-Ultra NX GPU.

Mali G2-Ultra NX

The focus here is on the Mali G2-Ultra NX, which ARM is calling its first AI-native Mali GPU. The company is adding Neural Accelerators alongside the chip’s shader execution hardware, which should, technically speaking, allow it to execute the aforementioned DLSS 5-like neural rendering features, along with other preexisting technologies such as upscaling, frame generation, and ray reconstruction.

ARM CSS For Mobile 2 2
Image: ARM.

To that end, the CSS For Mobile 2 Platform and corresponding Mali G2-Ultra NX GPU will support three hardware accelerated neural graphics features: Neural Super Sampling (NSS), Neural Frame Rate Upscaling (NFRU), and Neural Super Sampling and Denoising (NSSD). And before you ask: yes, AI is at the heart of the platform’s design.

NSS will allow the chip to upscale content resolution from 540p to 1080p. NFRU essentially doubles the output from 30 fps to 60 fps, and NSSD will combine supersampling with ray reconstruction. “The GPU brings neural acceleration directly into the GPU, with neural accelerators tightly integrated into the shader cores, alongside a new execution engine and third-generation ray tracing unit. This complements the agentic AI capabilities of the C2 CPU cluster with dedicated GPU acceleration for increasingly demanding neural graphics workloads.”

C2-Ultra

ARM CSS For Mobile 2 3
Image: ARM.

Over on the CPU side, the new ARM C2-Ultra now includes SME2 support that, as is the trend, is made to support Agentic AI workloads, including speech, retrieval, and model execution. Performance gains include up to 1.7x improvement in AI models; up to 15% higher single-thread performance; 15% faster web browsing; 12% faster application launch; and 12% higher multi-threaded performance at the cluster level.

At the time of writing, ARM did not provide a specific launch date for its new mobile platform to partners.

(Source: ARM [1] [2] [3])

Leave a Reply

Your email address will not be published. Required fields are marked *

Need Help?