Scalable Pixel Streaming is here: TensorWorks has officially unlocked a new way for Unreal Engine users to seamlessly deploy Pixel Streaming applications in the cloud.
“The goal of this product is to empower Unreal Engine developers to deploy Pixel Streaming solutions at unprecedented scale and to make the technology more accessible to the growing Unreal Engine user community,” said Adam Rehn from TensorWorks.
Earlier this month, Epic Games released Unreal Engine 5, the next iteration of the world’s most advanced real-time 3D creation tool. UE5 gives creators more freedom and fidelity than ever before so they can bring the next-generation of photorealistic, interactive experiences to life.
“We’re thrilled that CoreWeave and TensorWorks are partnering to make it simpler and more cost-effective for Unreal Engine developers to deploy and scale Pixel Streaming applications in the cloud. This will lead to wider adoption of this exciting technology, and more dynamic virtual experiences that will be enjoyed by more people,” said Maurizio Sciglio, Director of Cloud Partnerships at Epic Games.
Pixel Streaming leverages the cloud to distribute real-time, interactive content to users on any device, removing the dependency for high-powered PCs or game consoles. Example applications include live events, e-sports, business-to-client presentations, product configurators, and collaborative product development, but the possibilities are endless.
In 2019, TensorWorks spearheaded the development of Linux support for Unreal Engine pixel streaming in the cloud. This allows engineers to take advantage of the numerous benefits associated with deploying applications in GPU-accelerated containers, such as increased portability, less complexity and lower costs.
Scalable Pixel Streaming makes it incredibly easy to deploy Pixel Streaming in the cloud. This allows businesses delivering real-time experiences to leverage the advantages of elastic computing resources at scale, without the hassle of manually managing infrastructure and scheduling workloads. Users can easily build images and deploy automatically scaled endpoints.
As use cases for Pixel Streaming continue to grow, forward-thinking companies like TensorWorks have been heavily involved in projects aimed at making this ultramodern technology faster, cheaper, and more accessible. And that’s where CoreWeave comes in.
CoreWeave is built for compute-intensive workloads
Due to the real-time nature of Pixel Streaming, high GPU availability and responsive scaling capabilities are critical to ensuring a positive experience for end users, and it can be difficult (and expensive) to find suitable resources.
CoreWeave’s infrastructure is specifically designed for high-performance, compute-intensive workloads served on-demand with high availability. Our industry-leading range of NVIDIA GPUs and Kubernetes native infrastructure is built to provide responsive auto-scaling and lightning fast spin-up times, so our clients can adapt quickly to end-user demand at a 50-80% lower performance-adjusted cost.
Access the most powerful data center GPU for visual computing
The NVIDIA A40 GPU delivers significant performance capabilities in graphics and compute workflows, making it the ideal GPU for Pixel Streaming.
CoreWeave currently owns the largest fleet of NVIDIA A40 GPUs in North America, and our proprietary auto-scaling technology allows customers to spin up instances in seconds based on end-user demand.
To get the best Scalable Pixel Streaming experience, TensorWorks recommends using Kubernetes orchestration that can handle capacity fluctuations in real-time. CoreWeave Cloud customers can deploy containerized workloads quickly and efficiently, scaling lightning fast to meet user demand. CoreWeave customers can take advantage of bare-metal performance without having to manage the infrastructure. Paired with SPS, consider your infrastructure problems solved!
Previously, companies had difficulties leaning-in to Pixel Streaming because of the challenges of scaling on-demand across legacy cloud providers and the cost of doing it at scale. CoreWeave is designed to deliver high-performance compute at a fraction of the cost. Compared to traditional cloud providers, CoreWeave reduces the per-user cost of Pixel Streaming by as much as 80%, on an infrastructure purpose built for large-scale on-demand workloads. We believe that this type of cutting-edge technology should be more accessible to today’s creators and engineers, and we’re proud to work with TensorWorks and Unreal Engine to make that possible.
If you’re an Unreal Engine user and you’re looking to give Scalable Pixel Streaming a try, here’s how you can get started with our closed beta prior to public release:
- Request a CoreWeave Cloud account.
- Compile your Unreal Engine project for Linux, and confirm compatibility.
- Follow these instructions to deploy SPS on CoreWeave.
.png)
.png)