Sold by: Firemind
Quick Launch
AWS Free Tier
A free, read-only assessment of AI training and inference workloads running on specialist GPU clouds, for migration to AWS. Firemind maps your clusters, accelerators, storage and job patterns to AWS targets and returns a like-for-like business case and a migration plan.
Overview
Overview: The GPU Cloud to AWS Migration Assessment, built on Firemind's IT Operations Engine, is for teams running AI training or inference on specialist GPU clouds, sometimes called neoclouds, who want those workloads next to the rest of their estate on AWS. It reads how your clusters are built and used, and tells you what each workload should move to, what it would take, and how it compares with today. You receive a migration plan for every workload, free and with no commitment.
What the assessment covers
- GPU instance types, counts and utilisation
- Kubernetes and Slurm cluster topology, schedulers and node pools
- High-speed interconnect and the jobs that depend on it
- Shared file systems and object storage, by size and throughput
- Training, fine-tuning and inference job patterns
- Data transfer volumes between the GPU cloud and the rest of your estate
Typical targets include Amazon EC2 accelerated computing instances with Amazon EC2 Capacity Blocks for ML, Amazon SageMaker HyperPod for training clusters, Amazon EKS for containerised inference, AWS Trainium and AWS Inferentia where the models suit them, Amazon FSx for Lustre and Amazon S3 for data, and Amazon Bedrock where a managed model can replace a self-hosted one.
What you receive
- An inventory of every workload in scope, grouped into applications by dependencies and tags
- A recommended migration path for each workload, from rehost, relocate and replatform to repurchase, refactor, retain or retire, with the AWS target and the reason
- Dependencies and blockers, with the evidence behind each finding
- A like-for-like comparison that includes the cases where staying put is cheaper, alongside capacity options, accelerator choice, data transfer and the cost of running two platforms
- A draft wave plan, ordered so lower-risk workloads move first
- Right-sized AWS estimates, ready to load into the AWS Pricing Calculator, including the post-migration annual run rate that MAP eligibility is based on
- A run-readiness view: what it would take to operate each workload once it lands on AWS
- A readout with a Firemind cloud architect, and a summary you can take into a Migration Acceleration Program (MAP) Assess conversation with your AWS account team
Access: read-only, and your data stays out of scope
Access is read-only. Connect a read-only API credential where your provider offers one and, for Kubernetes-based clusters, a credential bound to the built-in Kubernetes view role, which excludes Secrets. Alternatively, upload an inventory export. Training data, model weights, checkpoints and secrets stay out of scope.
Optionally, deploy the IT Operations Engine read-only AWS role with Quick Launch so the assessment also checks the AWS landing zone your workloads would move into. It reads configuration and metadata only and makes no changes. If you already run the IT Operations Engine Free Assessment, this step is already done.
How it works
- Subscribe here. It is free, no card, no commitment.
- Connect a read-only credential for your GPU cloud, or upload an inventory export. Firemind provides the steps for your provider and the read-only collection script, which you review before running.
- Optionally, deploy the read-only AWS role using Quick Launch. It registers with Firemind automatically.
- Firemind returns your assessment and a readout with a Firemind cloud architect.
- Revoke the credential with your provider whenever you choose. Access ends with it.
Scope: Every cluster in your GPU cloud account, with no cluster cap. Bring as much of your estate as you can: the more the assessment sees, the more complete your plan and business case.
Next step: GPU Cloud to AWS Migration by Firemind on AWS Marketplace takes the plan through a full MAP Assess, landing zone, migration and operations, delivered as a private offer.
About Firemind: AWS Premier Tier Services Partner with the AWS Migration and Modernization Competency. 200+ clients since 2018, ISO 27001 certified, in partnership with AWS, NVIDIA and Anthropic. Core markets: UK and Ireland, Finland, Germany and the Netherlands.
Highlights
- Built for AI infrastructure: GPU types and utilisation, cluster topology, interconnect, storage throughput and job patterns, mapped to AWS accelerated computing, SageMaker HyperPod, Amazon EKS and Trainium.
- A like-for-like business case: it shows where staying put is cheaper, alongside capacity, accelerator choice, data transfer and the cost of running two platforms.
- Read-only by design: training data, model weights, checkpoints and secrets stay out of scope. The assessment reads configuration, scheduling and utilisation only.
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Quick Launch
Leverage AWS CloudFormation templates to reduce the time and resources required to configure, deploy, and launch your software.
Pricing
This product is available free of charge. Free subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimensions summary
This listing has one pricing dimension, offered at no charge. You pay nothing to use the GPU Cloud to AWS Migration Assessment. There are no tiers, instance sizes, or usage add-ons to choose from. The assessment reads your source environment and maps components so you can evaluate moving GPU workloads to AWS. Because this is a free dimension, your cost does not scale with the number of accounts, resources, or workloads assessed. Any follow-on engagement would be arranged separately with the vendor.
Top-of-mind questions for buyers
The assessment reads your source environment and produces an architecture mapping of your GPU workloads. It cross-references resources against best practice and maps components for a move to AWS. It uses read-only access and changes nothing. The free dimension covers this assessment work at no charge.
No. This is a free dimension, so your cost stays at zero regardless of how many accounts, regions, or workloads the assessment covers. The assessment can inventory every region and enabled service without adding charges. Nothing accrues as scope grows.
No. The free dimension covers the assessment and mapping only. Any migration build, rebuild, or ongoing operations would be arranged separately with the vendor. The assessment produces the mapping and evidence; decisions on next steps and any paid engagement stay with you.
www.firemind.com+1
Helpful?
Vendor refund policy
This product is free. AWS Marketplace collects no fees from you and Firemind does not invoice for it, so there is nothing to refund. You can cancel your subscription at any time in the AWS Marketplace console, and you can revoke Firemind's access immediately by removing the connection you created. If you believe you have been charged for this product in error, contact Firemind at hello@firemind.com and we will respond within one business day.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Software as a Service (SaaS)
SaaS delivers cloud-based software applications directly to customers over the internet. You can access these applications through a subscription model. You will pay recurring monthly usage fees through your AWS bill, while AWS handles deployment and infrastructure management, ensuring scalability, reliability, and seamless integration with other AWS services.
Additional details
Usage instructions
Use Quick Launch above if you can. These steps are for buyers whose policy is to review and deploy CloudFormation themselves.
What gets created, and nothing else
- One IAM role, FaePlatformRole, which Firemind assumes to run the assessment
- Three IAM managed policies, all of them Deny, layered over the role's baseline
- One execution role and one Lambda function, used once to report the created role ARN back to Firemind, then never invoked again
- One custom resource that invokes that function
What the role can and cannot do
The baseline is the AWS managed SecurityAudit and ViewOnlyAccess policies. Three Deny policies sit over the top, covering object and storage data, database and stream contents, and secrets, key material and privilege escalation. An explicit Deny always wins in IAM evaluation, so this holds regardless of what the baseline would otherwise permit.
The role can list a bucket and count its objects, but cannot read one. It can describe a database, its engine and its encryption configuration, but cannot read a row. It cannot decrypt anything, read a secret or parameter value, download log contents, or open a session to a host. Remediation is disabled and the role holds no write permissions.
Steps
- Download the template from the URL shown in the Quick Launch panel and review it. Everything above is visible in the file.
- Create the stack in any region you like. The role is global, so the region only determines where the stack object lives. Assessment covers every region regardless.
- Acknowledge CAPABILITY_NAMED_IAM. The role name is fixed so that Firemind can derive the ARN without asking you for it.
- Set FaeTrustedAccountId to the Firemind account shown on the Quick Launch page, and FaeCallbackUrl to the registration endpoint shown there. Leave the remaining parameters at their defaults.
- On completion the stack reports the role back to Firemind automatically. If that call fails the stack rolls back, by design, so a silent half-configured state is not possible.
Trust
The trust policy is conditioned on an external ID equal to your own AWS account ID. Both sides derive it, so there is no secret to exchange and nothing to rotate.
Ending access
Delete the stack. That removes the role and every policy with it, and Firemind's access ends at that moment. No notice is required.
Support
Vendor support
For support, or questions about your assessment, contact Firemind at hello@firemind.com . We respond within one business day (UK business hours, Monday to Friday).
This assessment includes:
- Help setting up the read-only connection, including reviewing it with your security team before you run it
- A migration assessment and a readout with a Firemind cloud architect
- Help removing Firemind's access, and deleting any uploaded exports, when you are finished
There is no charge for this product and no paid support tier attached to it.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
Move your GPU cloud workloads to AWS, from a MAP-aligned assessment through landing zone, wave-by-wave migration and operations. Capacity, accelerator choice and performance are planned and benchmarked before anything moves. Every change runs as governed code in your AWS account, and the same engine runs each wave once it lands.
A free, read-only assessment of your Google Cloud estate for migration to AWS. Connect keyless, read-only access and Firemind maps every workload to its AWS target, recommends a migration path for each, and returns a draft wave plan and the inputs for your AWS business case.
A free assessment of the AI models, agents and applications you run outside AWS, for migration to Amazon Bedrock and Amazon SageMaker AI. Firemind maps each workload to an AWS model and service, plans how to prove it performs at least as well, and returns a migration plan.
A free, read-only assessment of your Microsoft Azure estate for migration to AWS. Grant keyless, Reader-only access and Firemind maps every workload to its AWS target, recommends a migration path for each, and returns a draft wave plan, licensing findings and the inputs for your AWS business case.
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.