diff --git a/get-started.mdx b/get-started.mdx index aa7e3789..1fe36119 100644 --- a/get-started.mdx +++ b/get-started.mdx @@ -34,6 +34,10 @@ Now that you've created your account, you're ready to deploy your first Pod: If you haven't set up payments yet, you'll be prompted to add a payment method and purchase credits for your account. + +If your selected GPU is out of capacity, click **Deploy When Available** instead of **Deploy On-Demand**. Runpod queues your Pod and deploys it automatically once the GPU frees up, then notifies you by email and in your console inbox. See [Choose a Pod](/pods/choose-a-pod#deploy-when-available) for details. + + diff --git a/pods/choose-a-pod.mdx b/pods/choose-a-pod.mdx index 2c1d208d..5c8eaf94 100644 --- a/pods/choose-a-pod.mdx +++ b/pods/choose-a-pod.mdx @@ -61,7 +61,31 @@ For data-intensive workloads, ensure sufficient volume disk or network volume ca 1. **Right-size your resources**: Start with the minimum viable configuration, then scale up based on actual usage. Development and testing often need less power than production. -2. **Consider savings plans**: For extended usage, Runpod's [savings plans](/pods/pricing#savings-plans) reduce costs for committed usage. +2. **Use Deploy When Available for scarce GPUs**: When a high-demand GPU is out of capacity, [Deploy When Available](#deploy-when-available) lets you queue your Pod and deploy automatically once it frees up, instead of paying for a larger or pricier GPU to start sooner. + +## Deployment options + +When you deploy a Pod, the deploy screen surfaces a few features that help you find available capacity and choose the best GPU for your workload. + +### Recommended tab + +The **Recommended** tab highlights the GPUs best suited to your selected template. It lists the template's recommended GPUs first, then adds up to three more GPUs that are currently available, ranked by VRAM. This lets you compare strong recommendations against options that are in stock and ready to deploy right away. + +### NVLink availability + +When you deploy a multi-GPU Pod (a GPU count greater than 1) on an H100, H200, or B200, a banner tells you whether the GPU supports NVLink. Only SXM and NVL cards provide NVLink for high-bandwidth GPU-to-GPU communication, which benefits distributed training and other workloads that exchange large amounts of data between GPUs. + +### Deploy when available + +When your chosen GPU is out of capacity, you don't have to deploy immediately or keep retrying. Select **Deploy When Available** to queue your Pod, and Runpod deploys it automatically as soon as the GPU frees up. + +You set a window for when Runpod should attempt the deployment: + +- **Now**: Deploy as soon as capacity becomes available. +- **In 24 hours** (default): Keep trying for the next 24 hours. +- **Custom**: Use the calendar picker to set a specific From and To time, with hour and minute precision. + +Runpod sends an email and a console inbox notification when your Pod deploys or finishes migrating, so you know when it's ready without watching the console. ## Secure Cloud vs Community Cloud diff --git a/pods/manage-pods.mdx b/pods/manage-pods.mdx index 07961ed1..fa33c389 100644 --- a/pods/manage-pods.mdx +++ b/pods/manage-pods.mdx @@ -249,6 +249,12 @@ Pods provide two log types: Access logs from the [Pods page](https://www.console.runpod.io/pods) by expanding your Pod and clicking **Logs**. +## View template details + +Expand a Pod and open the **Template details** section to review the template your Pod was deployed from, including its readme and configuration. + +If the template isn't working as expected, click **Report problem with template** to open a feedback form and describe the issue. Your feedback goes directly to the Runpod team to help improve the template. + ## Troubleshooting | Issue | Solution | diff --git a/pods/overview.mdx b/pods/overview.mdx index e8b63956..32bc6f45 100644 --- a/pods/overview.mdx +++ b/pods/overview.mdx @@ -55,7 +55,7 @@ Runpod offers two cloud options: ## Pricing -Pods are billed by the minute with no fees for ingress/egress. Runpod also offers long-term [savings plans](/pods/pricing#savings-plans) for extended usage patterns. See [Pod pricing](/pods/pricing) for details. +Pods are billed by the minute with no fees for ingress/egress. See [Pod pricing](/pods/pricing) for details. ## Limitations diff --git a/pods/pricing.mdx b/pods/pricing.mdx index e66c7ab9..5130886c 100644 --- a/pods/pricing.mdx +++ b/pods/pricing.mdx @@ -33,12 +33,12 @@ You must have at least one hour's worth of credits for your selected configurati ### Savings plans -Commit to a 3-month or 6-month term upfront for significant discounts on compute costs. When you stop a Pod, the savings plan automatically applies to your next deployment of the same GPU type. - -Savings plans only cover GPU compute costs—[storage costs](/pods/storage/types) are billed at standard rates. Storage charges continue to accrue on stopped Pods. If your balance reaches \$0, your Pods stop: those with a network volume are preserved, while those without one are terminated and their data cannot be recovered. Plans are non-refundable and have fixed expiration dates. +Savings plans have been discontinued. You can no longer create new savings plans, but existing plans are unaffected and remain active until their commitment period expires. +Savings plans let you commit to a 3-month or 6-month term upfront in exchange for discounts on compute costs. If you have an active plan, it continues to apply automatically: when you stop a Pod, the plan applies to your next deployment of the same GPU type until it expires. Savings plans cover GPU compute costs only—[storage costs](/pods/storage/types) are billed at standard rates. Storage charges continue to accrue on stopped Pods, and plans are non-refundable with fixed expiration dates. + ## Storage pricing | Storage type | Running Pod | Stopped Pod | Notes | diff --git a/release-notes.mdx b/release-notes.mdx index 26de4622..a163721b 100644 --- a/release-notes.mdx +++ b/release-notes.mdx @@ -4,6 +4,15 @@ sidebarTitle: "Product updates" description: "New features, fixes, and improvements for the Runpod platform." --- + +## Pod deployment experience improvements + +- **[Deploy When Available](/pods/choose-a-pod#deploy-when-available):** When your chosen GPU is out of capacity, you can queue your Pod and let Runpod deploy it automatically as soon as the GPU frees up. Set a deployment window using Now, In 24 hours, or a custom range, and you'll get an email and a console inbox notification when your Pod deploys or finishes migrating. +- **[Recommended tab](/pods/choose-a-pod#recommended-tab):** The deploy screen now pairs each template's recommended GPUs with additional high-availability options that are currently in stock. +- **[NVLink availability](/pods/choose-a-pod#nvlink-availability):** When you deploy a multi-GPU Pod on an H100, H200, or B200, a banner shows whether the GPU supports NVLink. + + + ## Flash beta: Run Python functions on cloud GPUs