GPUMini offers two contact channels: email support@gpumini.com or sign in to the console and submit a ticket. Email is best for pre-sales and business inquiries; for issues linked to an order, instance, or bill, use a ticket so the full handling record stays together.
Workload, memory, storage, term, and node recommendations
Connection issues, build failures, disk, and network diagnostics
Order amounts, terms, add-ons, and payment records
Bulk nodes, team permissions, delivery schedule, and review materials
Risk details, redacted evidence, and impact scope
This form does not submit content to the site server. After you click Generate, it opens the email app on your device and organizes your entries into an email addressed to support@gpumini.com. Review sensitive information again before sending.
Best for pre-sales selection before an order is created, business purchasing, security consultations, and general questions that do not depend on a specific instance record.
Best for connection issues, build failures, configuration checks, and payment record questions linked to an existing order, instance, or bill. A ticket keeps follow-up information in one handling record.
“We need a faster Mac” alone is rarely enough for a reliable recommendation. The information below helps us compare GPUMini M4 Core and GPUMini M4 Plus and determine whether additional storage or a Thunderbolt 5 expansion option is needed.
List the tools you use most—Xcode, xcodebuild, Fastlane, CI Runner, a remote Mac desktop, or AI experiments—and say whether the work is interactive development or long-running background processing.
Provide normal and peak concurrency, and note whether simulators, tests, archives, or multi-repository pipelines run at the same time. Concurrency reflects resource pressure better than the duration of a single build.
Record typical memory usage during a local build and note whether editors, browsers, simulators, and container tools run simultaneously. This helps compare 16GB and 24GB configurations.
Estimate the combined space used by source code, dependencies, DerivedData, archives, models, and temporary files. Include weekly growth to assess whether 256GB, 512GB, or additional storage is appropriate.
Describe where developers are based day to day and how team members are distributed. Nodes in Singapore, Japan (Tokyo), South Korea (Seoul), and Hong Kong are all available to order; choose based on real-world network testing.
If you cannot provide exact figures, share your current computer configuration, a typical build time, project directory size, and team size. Support will clarify what can be determined directly and what should be tested first.
Submit technical issues through a console ticket whenever possible. The information below helps identify the exact failure point in the connection path, node status, credential use, disk space, or build process.
Specify the Singapore, Japan (Tokyo), South Korea (Seoul), or Hong Kong node, and use a timezone-aware timestamp for the first occurrence and most recent reproduction.
State whether you use a remote Mac desktop, SSH, or file transfer, and identify the office network, home broadband, or mobile network. Record whether the result changes when switching networks.
Copy the complete error line, exit code, and failed step; do not write only “connection failed” or “build error.” Screenshots should retain context while masking addresses, credentials, and project secrets.
List actions and results in order, such as retrying the connection, switching networks, checking disk space, and rerunning commands. This prevents support from asking you to repeat steps already verified.
Keep commands, timestamps, exit codes, and necessary stack traces; remove private keys, access tokens, repository credentials, certificate contents, personal data, and unreleased source code.
Bulk dedicated physical machines involve more than quantity. Teams usually also need to define who manages orders, who uses the nodes, how delivery is accepted in batches, and which existing materials are needed for security or procurement review.
Send your business inquiry to support@gpumini.com and include “Business Purchasing” in the subject. We will organize the items requiring follow-up based on your target configuration, node distribution, and delivery schedule.
Provide the target machine count, the ratio of M4 Core to M4 Plus, node distribution, expected start date, and usage term. Actual availability is determined by the real-time response from the console.
Define the responsibilities of the purchasing owner, technical administrator, and node users, and specify which roles need access to orders, instances, billing, and support records.
Specify whether delivery is one-time or staged by project phase, and set acceptance criteria for each batch, including configuration checks, connection verification, and test builds.
List the questions required by procurement, security, privacy, or compliance processes. Requests for security reports and materials should likewise be sent to support@gpumini.com.
We use your name, reply email, issue category, node, term, and description to identify and respond to inquiries, troubleshoot issues, and retain necessary support records. Retention is limited according to business processing, account security, and legal obligations.
Do not submit account passwords, private keys, access tokens, complete certificate contents, unredacted build logs, or personal information unrelated to the issue. When logs are needed, include only the timestamps, commands, exit codes, and context required to locate the failure.
View the Privacy Policy and User Rights NoticeCheck order confirmations, instance status, billing records, and support tickets in the same management portal. If your issue concerns an existing service, sign in first and submit a ticket.