# Databricks and DualBird: the published facts in the same rows

Side by side, read 10 Nov 2026

**The short answer.** Databricks (Managed lakehouse built around Delta Lake) and DualBird (FPGA Spark plugin for AWS F2) are set out here in the same rows: published speed and cost results, code changes, table upkeep, table formats and where each runs. On the Compute Bill's reference job: Databricks on AWS (Jobs Compute): $32.40; DualBird on Amazon EMR (F2 instances): $49.50.

## The same rows for both

| Row | Databricks | DualBird |
| --- | --- | --- |
| Kind | Managed lakehouse built around Delta Lake | FPGA Spark plugin for AWS F2 |
| Does well | Per Databricks documentation, Photon keeps the Spark Catalyst planner, works with existing Spark API code and falls back to the Spark runtime for unsupported operations (read 2026-11-10). Per Databricks documentation, Delta tables can generate Iceberg metadata from Runtime 14.3 LTS, so Iceberg clients read Delta data without rewriting files (read 2026-11-10). | DualBird says onboarding is two changes: switch the EC2 instance type to F2 and add the DualBird Spark plugin, with Spark still handling planning and scheduling (read 2026-11-10). DualBird published a 100 GB sort-based Iceberg compaction run on June 11, 2026, run by DualBird, and states about $2 per TB for itself against about $6 per TB for CPU clusters. |
| Limited | Per Databricks documentation, serverless compute supports only Spark Connect APIs, so RDD APIs, R and Maven coordinates are unsupported, and ANSI mode on by default can change results of existing jobs. Per Databricks, compute is billed in DBUs in addition to the cloud bill for the same instances, so a Jobs Compute run pays Databricks and AWS separately; the Compute Bill line adds $12.24 of DBUs to $20.16 of EC2 (read 2026-11-10). | DualBird is not a lakehouse. It speeds up Spark on EMR and EKS; a team on a fully managed platform cannot add it. It documents no Spark version list and no table-format support list; Iceberg appears only in its own compaction benchmark. A price for the plugin and a benchmark run by a party other than DualBird: none found on 2026-11-10. |
| Published speed results | Docs claim Photon gives up to 5x better price/performance for data and analytics workloads versus other cloud data warehouses; no method is given on this page. Databricks announced an audited 100TB TPC-DS result of 32,941,245 QphDS for Databricks SQL, 2.2x the previous record held by Alibaba, with total system cost 10% lower on list pricing. This is a SQL warehouse result, not a Spark batch ETL result. The same Databricks blog says research from Barcelona Supercomputing Center found Databricks 2.7x faster and 12x better in price-performance than Snowflake on a TPC-DS derivative. This is a vendor-reported claim of a third-party study; the study itself was not fetched. | The home page body text states the headline claim as 10x-30x faster data processing with 50%-90% lower costs and no changes required. The same wording appears in the page's JSON-LD organization description and in the April 28, 2026 blog post. The raw HTML of the home page has a different range in its meta description and og:description (and twitter:description): 10x-100x faster. This is the meta tag, not body text. The February 5, 2026 funding blog post also says 10-100x faster and 50-90% cost savings. The body text of the home and product pages says 10x-30x, so the two ranges differ by site location. The product page repeats the 10x-30x end-to-end figure labeled 'Instance for instance' and 50%-90% cost reduction labeled 'Workload for workload'. The meta description of the product page is only 'Plug-and-play Spark acceleration.' Home and product pages show a '110GB Enrichment Pipeline (JOIN Denormalization)' benchmark: 50%-85% lower EC2 cost with 6x-18x higher performance ceiling. The pages give no hardware, baseline, dates or method, and no link to a write-up was found. Sort-based Iceberg compaction on 100 GB, compared with vanilla Spark and with an unnamed 'C++ accelerated Spark' (unnamed) with tuned EBS, memory and parallelism at each cluster size. Dataset is synthetic: about 370 million rows, 64 mixed-type columns, ZSTD level 3. Results: roughly 12-20x faster single-task execution, 55-85% lower EC2 cost, stable cost across scaling, about 6x higher practical performance ceiling. Cost per terabyte compacted is stated as about $2/TB for DualBird versus about $6/TB rising past $13/TB for CPU clusters. Part of the saving is attributed to F2 local NVMe replacing EBS volumes. The home page shows 'X15-X20' single-task speedup for the same benchmark, while the June 11, 2026 blog and the white paper landing page say 12-20x. The two do not match. The 'Single Task Performance' home-page tile says X15-X20 speedup for the same cluster size while reducing cost by over 50%. The blog describes the single-task comparison as against vanilla Spark, and the April 28 post as against 'a high-end C++ columnar engine running on a single CPU core'. The baseline wording is therefore not consistent across pages. |
| Published prices and cost results | Databricks bills compute in DBUs at per-second granularity with pay-as-you-go or committed-use discounts, and no up-front costs. The pricing pages render their rates by script; the AWS pricing data file those pages load lists Jobs Compute at $0.15 per DBU on the plan it names Premium (r6i.4xlarge: 4.08 DBU per hour). The file names no region. Jobs pricing offers serverless (Performance Optimized and Standard modes) and classic clusters; the jobs pricing page maps plan names across clouds (its Premium tier on Azure corresponds to Enterprise on AWS and GCP), while the AWS pricing data file lists the $0.15 Jobs Compute rate under a plan it names Premium. Photon instances consume DBUs at a different rate. | No pricing, price list or marketplace listing was found on the site. Cost claims are only percentages (50%-90% lower, 55%-85% lower EC2) and the about $2/TB versus $6-13/TB compaction figure. The pricing model for the plugin itself is not stated, so users cannot tell whether F2 instance price or license fee is included in the claimed savings. |
| Code changes and migration | Photon is a native C++ vectorized engine that keeps Spark's Catalyst planner and works with existing Spark API code; unsupported operations fall back to the Spark runtime. Serverless compute for notebooks and jobs supports only Spark Connect APIs; Spark RDD APIs, R, Maven coordinates, global temp views, DBFS mounts with instance profiles and Databricks Container Services are not supported. ANSI mode is on by default and the default time zone is UTC, which can change results of existing jobs. | Onboarding is two changes: switch the EC2 instance type to F2 and add the DualBird Spark plugin. The page shows three steps: change instance type, use our Spark plugin, see the results. The about page says 'No infrastructure changes. No code changes.' The compaction blog says Spark still handles planning, scheduling and fault tolerance and no pipeline rewrite or platform migration is needed. A DualBird engineer states the FPGA path accelerates a defined set of operators (sort, join, aggregation and others). Workloads dominated by Python UDFs or unsupported operators fall back to CPU and do not see the same benefit yet. Coverage is described as expanding. |
| Tuning and table upkeep | Photon is on by default for classic all-purpose compute, jobs compute and classic pipelines; API users must set runtime_engine to PHOTON. Photon instance types use DBUs at a different rate than the same instance without Photon. On serverless compute Databricks allocates and manages compute so users do not provision clusters; the Spark UI and Spark logs are not available, query profile and client-side logs are used instead. Serverless compute runs workloads without provisioning compute in your cloud account, and Databricks says this reduces the need to manage compute resources. It requires Unity Catalog to reach external data sources. | DualBird claims tens of Spark settings (executor sizing, CPU-memory ratios, parallelism) become irrelevant, disk spills effectively disappear and data skew stops causing spills, because only a few tasks run at once on the FPGA. The product page says the majority of common Spark performance configurations do not exist when using the solution. Cluster sizing reduces to the number of F2 nodes. These are vendor claims, not independently tested. Part 2 of the series (Sep 1, 2026) says skew still costs time, but less, and that it does not need repartitioning or salting. |
| Spark versions and table formats | Databricks Runtime 18 LTS ships Apache Spark 4.1.0 (released June 10, 2026); Runtime 17.3 LTS ships Spark 4.0.0; Runtime 16.4 LTS and 15.4 LTS ship Spark 3.5.x. Runtime 19 ships Spark 4.2.0 (July 23, 2026). Delta Lake tables can generate Iceberg metadata (UniForm, Iceberg reads) from Databricks Runtime 14.3 LTS so Iceberg clients read Delta data without rewriting files. | The home page says the plugin is compatible with Apache Spark, Amazon EMR and Amazon EKS. It lists 'Optimized out of the box', 'Migration-free' and 'No more performance tuning and troubleshooting'. No Spark version numbers, Databricks support or Iceberg/Delta/Hudi support list are stated on the home, product, about or resources pages. Iceberg appears only in the compaction benchmark and in the meta description of the Dec 1, 2025 demo post ('Accelerate Apache Spark and Apache Iceberg'). |
| Where it runs | Databricks on AWS lists many AWS regions including us-east-1 (US East, Northern Virginia), us-east-2, us-west-1, eu-west-1, eu-central-1, ap-southeast-2 and us-gov-west-1; the platform is also hosted on Azure and Google Cloud. Some features are only in a subset of regions. | DualBird needs AWS EC2 F2 instances. AWS describes F2 as second-generation FPGA instances with up to 8 AMD Virtex UltraScale+ HBM VU47P FPGAs, 16 GiB HBM per FPGA, 3rd-gen AMD EPYC (Milan) CPUs up to 192 vCPU, up to 2 TiB memory and up to 7.6 TiB NVMe. Sizes: f2.6xlarge (1 FPGA, 24 vCPU, 256 GiB), f2.12xlarge (2 FPGAs, 48 vCPU, 512 GiB), f2.48xlarge (8 FPGAs, 192 vCPU, 2,048 GiB). The AWS 'instance types by Region' doc lists F2 in eight regions: us-east-1 (N. Virginia), us-west-2 (Oregon), ap-northeast-1 (Tokyo), ap-northeast-2 (Seoul), ap-southeast-2 (Sydney), ca-central-1 (Canada Central), eu-central-1 (Frankfurt) and eu-west-2 (London). F2 is not listed in other regions on that page (for example us-east-2, eu-west-1). DualBird is AWS-only: no other cloud is mentioned. The product page says deployment runs on the customer's own servers in their AWS environment and 'Everything stays under your control'. |
| How open its benchmarks are | None found on 10 Nov 2026. | The white paper (dated July 1, 2026) has a landing page that restates the results and says methodology, system architecture and detailed analysis are in a download. The full document is gated behind a form and was not readable. The blog gives dataset shape and baselines but not instance counts, Spark version, configs or scripts. No public repo was found (github.com/dualbird returns 404). |
| Compute Bill | Databricks on AWS (Jobs Compute): $32.40 | DualBird on Amazon EMR (F2 instances): $49.50 |

Alphabetical order. Vendor statements are the vendor's own. Read 10 Nov 2026.

## How to read this page

The rows hold what each vendor or a named third party published, with no conclusion drawn. Benchmarks in the two columns ran on different hardware, data and baselines, so their numbers are not comparable across columns. The [Compute Bill method](/compute-bill/method/) explains the cost lines.

## Sources

- Databricks Data Intelligence Platform on AWS: docs.databricks.com/aws/en/compute/photon: https://docs.databricks.com/aws/en/compute/photon (read 10 Nov 2026)
- Databricks Data Intelligence Platform on AWS: docs.databricks.com/aws/en/compute/serverless/limitations: https://docs.databricks.com/aws/en/compute/serverless/limitations (read 10 Nov 2026)
- Databricks Data Intelligence Platform on AWS: docs.databricks.com/aws/en/compute/serverless: https://docs.databricks.com/aws/en/compute/serverless/ (read 10 Nov 2026)
- Databricks Data Intelligence Platform on AWS: docs.databricks.com/aws/en/release-notes/runtime: https://docs.databricks.com/aws/en/release-notes/runtime/ (read 10 Nov 2026)
- Databricks Data Intelligence Platform on AWS: docs.databricks.com/aws/en/delta/uniform: https://docs.databricks.com/aws/en/delta/uniform (read 10 Nov 2026)
- Databricks Data Intelligence Platform on AWS: docs.databricks.com/aws/en/resources/supported-regions: https://docs.databricks.com/aws/en/resources/supported-regions (read 10 Nov 2026)
- Databricks Data Intelligence Platform on AWS: databricks.com/blog/2021/11/02/databricks-sets-official-data-warehousing: https://www.databricks.com/blog/2021/11/02/databricks-sets-official-data-warehousing-performance-record.html (read 10 Nov 2026)
- Databricks Data Intelligence Platform on AWS: databricks.com/product/aws-pricing: https://www.databricks.com/product/aws-pricing (read 10 Nov 2026)
- Databricks Data Intelligence Platform on AWS: databricks.com/product/pricing/jobs: https://www.databricks.com/product/pricing/jobs (read 10 Nov 2026)
- DualBird: dualbird.io: https://dualbird.io/ (read 10 Nov 2026)
- DualBird: dualbird.io/product: https://dualbird.io/product (read 10 Nov 2026)
- DualBird: dualbird.io/resources/blog/a-simpler-spark: https://dualbird.io/resources/blog/a-simpler-spark (read 10 Nov 2026)
- DualBird: dualbird.io/resources/blog/iceberg-compaction-benchmark-faster-spark-low: https://dualbird.io/resources/blog/iceberg-compaction-benchmark-faster-spark-lower-cost (read 10 Nov 2026)
- DualBird: dualbird.io/resources/blog/spark-iceberg-compaction-benchmark-whitepaper: https://dualbird.io/resources/blog/spark-iceberg-compaction-benchmark-whitepaper (read 10 Nov 2026)
- DualBird: aws.amazon.com/ec2/instance-types/f2: https://aws.amazon.com/ec2/instance-types/f2/ (read 10 Nov 2026)
- DualBird: docs.aws.amazon.com/ec2/latest/instancetypes/ec2-instance-regions.html: https://docs.aws.amazon.com/ec2/latest/instancetypes/ec2-instance-regions.html (read 10 Nov 2026)

Last reviewed 10 Nov 2026. Canonical: https://datalakehousecompare.com/compare/databricks-vs-dualbird/
