r6id.8xlarge on Databricks
r6id.8xlarge is a memory optimized AWS worker type with 32 vCPUs and 256 GB of memory. On Databricks it uses 8.16 DBUs an hour, so 1 node on Jobs Compute in US East (N. Virginia) costs $3.64 an hour: $1.22 for the DBUs and $2.42 to AWS for the machine.
AWS · Memory optimized · r6id
r6id.8xlarge at a glance
- vCPUs
- 32
- Memory
- 256 GB
- Own SSD
- 1,900 GB
- DBUs an hour
- 8.16
- Classic compute
- With Photon
- 23.66
- DBUs an hour, Jobs Compute
- 1 node an hour
- $3.64
- Jobs Compute, US East (N. Virginia)
1 node an hour on each kind of compute
In US East (N. Virginia) on the Premium plan. Databricks charges the DBUs, AWS the machine. A cluster pays this for every worker and for the driver.
| Compute | DBUs an hour | A DBU | Databricks | Machine | An hour |
|---|---|---|---|---|---|
| Jobs Compute | 8.16 | $0.15 | $1.22 | $2.42 | $3.64 |
| Jobs Compute + Photon | 23.66 | $0.15 | $3.55 | $2.42 | $5.97 |
| All-Purpose Compute | 8.16 | $0.55 | $4.49 | $2.42 | $6.91 |
| All-Purpose Compute + Photon | 16.32 | $0.55 | $8.98 | $2.42 | $11.40 |
| Lakeflow pipelines Core | 8.16 | $0.20 | $1.63 | $2.42 | $4.05 |
| Lakeflow pipelines Core + Photon | 23.66 | $0.20 | $4.73 | $2.42 | $7.15 |
| Lakeflow pipelines Pro | 8.16 | $0.25 | $2.04 | $2.42 | $4.46 |
| Lakeflow pipelines Pro + Photon | 23.66 | $0.25 | $5.92 | $2.42 | $8.34 |
| Lakeflow pipelines Advanced | 8.16 | $0.36 | $2.94 | $2.42 | $5.36 |
| Lakeflow pipelines Advanced + Photon | 23.66 | $0.36 | $8.52 | $2.42 | $10.94 |
In each region
1 node an hour on Jobs Compute, Premium, cheapest first.
| Region | Machine | An hour |
|---|---|---|
| US East (N. Virginia) | $2.42 | $3.64 |
| US East (Ohio) | $2.42 | $3.64 |
| US West (Oregon) | $2.42 | $3.64 |
| Asia Pacific (Mumbai) | $2.54 | $3.76 |
| Europe (Ireland) | $2.69 | $3.91 |
| Asia Pacific (Sydney) | $2.78 | $4.01 |
| Europe (London) | $2.84 | $4.06 |
| Europe (Frankfurt) | $2.91 | $4.13 |
| Asia Pacific (Seoul) | $2.91 | $4.13 |
| Asia Pacific (Singapore) | $2.92 | $4.15 |
| Asia Pacific (Tokyo) | $2.92 | $4.15 |
Other memory optimized sizes
- r6id.xlarge4 vCPUs, 32 GB, 1.02 DBUs an hour
- r6id.2xlarge8 vCPUs, 64 GB, 2.04 DBUs an hour
- r6id.4xlarge16 vCPUs, 128 GB, 4.08 DBUs an hour
- r6id.8xlarge32 vCPUs, 256 GB, 8.16 DBUs an hour
- r6id.12xlarge48 vCPUs, 384 GB, 12.24 DBUs an hour
- r6id.16xlarge64 vCPUs, 512 GB, 16.32 DBUs an hour
- r6id.24xlarge96 vCPUs, 768 GB, 24.48 DBUs an hour
- r6id.32xlarge128 vCPUs, 1,024 GB, 32.64 DBUs an hour
Questions
- How many DBUs does r6id.8xlarge use on Databricks?
- r6id.8xlarge uses 8.16 DBUs an hour on Databricks classic compute without Photon. With Photon it uses 23.66 DBUs an hour on Jobs Compute and 16.32 on All-Purpose Compute, as Databricks' price list gives them.
- How much does r6id.8xlarge cost an hour on Databricks?
- In US East (N. Virginia), 1 r6id.8xlarge node on Jobs Compute costs $3.64 an hour: $1.22 to Databricks for 8.16 DBUs at $0.15 each, and $2.42 to AWS for the machine. On All-Purpose Compute the same node costs $6.91 an hour, and on Lakeflow pipelines Core $4.05. A cluster pays this for every worker and for the driver.
- Does Photon cost more on r6id.8xlarge?
- Yes, by the hour: with Photon, 1 r6id.8xlarge node on Jobs Compute in US East (N. Virginia) costs $5.97 an hour instead of $3.64, because it uses 23.66 DBUs an hour instead of 8.16. It saves money only when the job finishes fast enough to make up the difference.
- Which region is cheapest for r6id.8xlarge on Databricks?
- US East (N. Virginia), at $3.64 an hour for 1 node on Jobs Compute, against $4.15 in Asia Pacific (Tokyo), the dearest of the 11 regions listed here that offer it. Databricks charges the same for a DBU of Jobs Compute in every AWS region listed, so the gap is the machine. US West (N. California), Canada (Central), South America (São Paulo), Europe (Paris) and Asia Pacific (Jakarta) do not offer it.
- What is the same size as r6id.8xlarge on the other clouds?
- The nearest worker types of the same kind are Standard_E32ds_v5 on Azure and n2-highmem-32 on Google Cloud. Each cloud gives its own DBUs an hour for its types, so compare what a cluster costs in the calculator, not the DBUs alone.
- What is the next size up or down from r6id.8xlarge?
- One size down is r6id.4xlarge, at 4.08 DBUs an hour; one size up is r6id.12xlarge, at 12.24. Within a family, DBUs and the machine price grow with the vCPUs, so 2 nodes of a size cost about what 1 node of twice the size costs.
Sources
Where each fact comes from
- 1Databricks: pricing calculator and instance types (the DBUs an hour of each worker type, with and without Photon)
- 2Databricks: pricing on AWS (the price of a DBU for each kind of compute and plan)
- 3AWS: Amazon EC2 on-demand pricing
- 4Databricks: compute configuration reference (the driver is the workers' type by default)
- 5Databricks: Photon
- 6AWS: r6id sizes