The available evidence does not establish whether Salad currently performs proof-of-work mining. The distinction depends on what its assigned computing task does, not simply on whether it uses a GPU. To verify a mining claim, current, dated technical documentation would need to show that the task contributes work to a named proof-of-work network. No Salad-specific documentation or workload configuration is provided here, so a definitive conclusion about its current tasks is not supported.

What counts as crypto mining?

Proof of work is a consensus model that uses a hard-to-solve but relatively easy-to-verify computational problem to select a participant for publishing a block. It consumes time, energy, and computational cycles. The details—including the algorithm, difficulty rules, and hardware—depend on the particular network. This makes proof-of-work mining a specific network function, not a general label for demanding computation. source

Bitcoin illustrates how that function can work. Bitcoin mining software constructs a candidate block and block header from transactions, then sends header work to hashing hardware. The hardware searches for qualifying work associated with that candidate. This describes Bitcoin’s workflow; it should not be treated as a universal specification for all networks or workloads. source

For classification, ask whether the computation participates in a named proof-of-work network’s block-production process. If a device performs some other computation, that task does not become mining merely because it runs on a GPU or is computationally intensive. The network role and purpose of the task are the key distinctions.

How to assess Salad crypto mining claims

Begin with current, first-party technical documentation that identifies the task assigned to a device. A useful description would name the proof-of-work network and explain how the assigned computation contributes to that network’s block-production process. A general statement about using computing resources, without identifying the network role, would not establish proof-of-work mining.

Where available, technical details can help clarify the task. For example, a mining software configuration or pool connection details may provide evidence of the network a workload is configured to contact. Such evidence must be current and tied to the particular task being discussed. A software name or endpoint, on its own, may not establish what a given device is currently doing; the documentation needs to connect the detail to the workload and network purpose.

Keep different kinds of evidence separate. First-party technical documentation can support a claim about the task it describes, within its date and scope. User reports may describe individual situations, but do not by themselves verify a current or universal product workload. Broad marketing language is not a substitute for technical evidence identifying a proof-of-work network and the task’s role in it. Observations such as GPU activity likewise show neither the computation’s purpose nor its contribution to block production.

The evidence available for this article does not include current, Salad-specific task documentation, a verifiable device configuration, or pool connection details. It therefore cannot establish that Salad currently performs proof-of-work mining. Nor does the absence of such evidence establish which other workloads, if any, it assigns. A careful account should leave both points unresolved unless current, attributable documentation supports a more specific conclusion.

GPU computing that is not proof-of-work mining

A graphics processing unit (GPU) is hardware that can be used for computation, but the device alone does not identify the workload. A task can use a GPU without participating in a proof-of-work network. The practical question is not simply whether a GPU is active; it is what the task computes and whether its output is part of a named network’s block-production process.

Device behavior is not a reliable substitute for task documentation. High utilization, heat, or fan noise may indicate that the device is doing work, but these observations do not reveal whether it is hashing for proof of work or performing a different computation. Classifying a workload from those observations would confuse the hardware activity with the purpose of the task.

Proof-of-work networks can also have network-specific difficulty rules. For example, difficulty can be adjusted as mining power changes, but adjustment intervals and formulas differ across networks. That is a feature of the particular proof-of-work design being examined, not a rule that identifies every GPU workload as mining. source

Tie any conclusion to dated, verifiable documentation. If a description refers to a particular task at a particular time, do not silently extend it to a different task or a later period. Check whether the document identifies the workload, the relevant network, and the relationship between them. If one of these links is missing, describe the classification as unverified rather than assuming that GPU use proves mining.

What the distinction does—and does not—tell you

Classifying a workload as proof-of-work mining would answer a technical question about its role. It would not, by itself, establish profitability, security, or suitability. Likewise, concluding that a task is not proof-of-work mining would not establish its fees, yields, availability, or recovery options. Those are separate questions requiring current, relevant evidence; they cannot be inferred from hardware use or from the workload label.

The same boundary applies to terms, availability, and compliance. A technical classification alone does not show what product terms apply or what obligations exist in any jurisdiction. For operational details, consult current product documentation. For jurisdiction-specific questions, consult authoritative sources that address the relevant jurisdiction. Without that separate evidence, those matters remain outside what a workload classification can establish.

Sources and references