Salad

Salad introduces a revolutionary approach to cloud computing through its distributed, people-powered cloud platform. At its core, Salad operates on a compute-sharing economy model, tapping into the vast, latent power of millions of consumer GPUs that sit idle for most of the day. By creating a global network of these devices, Salad provides businesses, developers, and AI researchers with access to immense computational power at a fraction of the cost of traditional hyperscale cloud providers. This democratizes access to the resources needed for cutting-edge AI innovation, high-performance computing, and large-scale data processing.

The platform is built on the idea of sustainability and efficiency, utilizing already-existing hardware to reduce the environmental impact and financial barriers associated with large data centers. For individuals who contribute their PC's power (known as 'Chefs'), Salad offers a simple desktop application that runs in the background, turning idle time into tangible rewards like gift cards, subscriptions, and games. For businesses, SaladCloud delivers a secure, reliable, and highly scalable infrastructure for deploying containerized applications.

How to use Salad

Using Salad is straightforward, with distinct processes for businesses and individual contributors:

For Businesses (Deploying on SaladCloud):

1. Containerize Your Application: Package your workload, such as an AI model or a data processing job, into a container image.
2. Choose Your Resources: Use the SaladCloud portal or API to specify your exact requirements, including the type of GPU (from GTX 10-series to RTX 50-series and A100s), the number of vCPUs, and the amount of RAM.
3. Deploy: Launch your containerized workload on the Salad Container Engine (SCE). Salad handles the orchestration, finding available nodes on the network that match your specifications and deploying your containers.
4. Scale and Monitor: Easily scale your instances up or down based on demand. You only pay for the compute time you use, with no charges for data ingress/egress or instance initialization.

For PC Owners (Earning with Salad):

1. Download the App: Install the free Salad desktop application on your Windows PC.
2. Run the App: Let the application run when your computer is idle. Salad automatically detects when your GPU and CPU are not in use and assigns compute jobs from the SaladCloud network.
3. Earn Rewards: You accumulate Salad Balance for the compute time you share. This balance can be redeemed for a wide variety of rewards in the Salad Storefront, including video games, gift cards, and more.

Core Features of Salad

• Distributed GPU Network: Access to a massive, globally distributed network of over 60,000 daily active consumer and datacenter GPUs.
• Salad Container Engine (SCE): A fully managed container service that simplifies deployment and orchestration, eliminating the need to manage VMs or individual instances.
• Extreme Cost-Effectiveness: Reduce compute costs by up to 90% with GPU instances starting from as low as $0.02/hour.
• On-Demand Scalability: Effortlessly scale from one to thousands of GPU instances in real-time to meet fluctuating demands without pre-provisioning.
• Global Edge Network: Deploy workloads on low-latency edge nodes located in over 190 countries, bringing computation closer to the user.
• SOC2 Certified Security: A patented approach isolates customer environments and data, ensuring secure and reliable deployments across the network.
• Transparent, Usage-Based Pricing: A simple pricing model with no hidden fees, no data transfer costs, and no long-term commitments.
• Support for Diverse AI Workloads: Optimized for GPU-heavy tasks including Text-to-Image, Text-to-Speech, LLMs, computer vision, and scientific computing.

Use Cases for Salad

Salad is ideal for a wide range of GPU-intensive applications:

• AI/ML Inference: Companies like Civitai use Salad to serve millions of AI image generations daily at a low cost. It's perfect for deploying models like Stable Diffusion, Llama, and Whisper.
• Voice AI: Achieve massive cost reductions on Speech-to-Text (transcription, captioning) and Text-to-Speech (TTS) workloads, with performance up to 99% cheaper than managed API services.
• High-Performance Computing (HPC): Distribute complex scientific and engineering simulations, such as molecular dynamics (GROMACS), across thousands of nodes.
• 3D Rendering & Digital Content Creation: Render complex scenes and animations by distributing the workload across the network, significantly speeding up production pipelines.
• Data Processing: Run large-scale, parallel data batch jobs efficiently and affordably.

Advantages of Salad

Salad offers a compelling alternative to traditional cloud services:

• Unbeatable Cost Savings: The primary advantage is the dramatic reduction in cloud spending, allowing startups and enterprises to scale their AI operations without breaking the bank.
• Massive Scalability: The decentralized nature of the network provides access to a vast pool of GPU resources that can be difficult to secure from major providers.
• Sustainable Computing: By leveraging existing idle hardware, Salad promotes a greener, more sustainable approach to cloud computing, reducing e-waste and energy consumption from new data centers.
• Democratized Access: Salad breaks the cloud monopoly held by big tech, enabling smaller companies to compete and allowing individuals to participate in and profit from the cloud economy.
• Simplified Operations: The managed container engine reduces DevOps overhead, allowing teams to focus on their applications rather than infrastructure management.

Pricing and Plans

Salad's pricing is designed to be transparent and highly competitive. It operates on a pay-as-you-go model with no upfront costs or long-term contracts.

• GPU Instances: Pricing is granular and depends on the specific GPU model, vCPU, and RAM configuration. Prices start as low as $0.02/hour for a GTX 1050 Ti and scale up for more powerful cards like the RTX 4090 ($0.18/hour) or A100 ($0.40/hour).
• Customizable Instances: Users can fully customize their instances to match their exact workload requirements, paying only for the resources they need.
• No Hidden Fees: Salad famously has no data ingress or egress fees, which can be a significant cost on other platforms.
• Volume Discounts: For large-scale deployments (50+ GPU instances) and enterprise contracts, Salad offers discounted pricing upon consultation with their sales team.

A detailed pricing calculator is available on their website to estimate costs accurately before deployment.