---
title: "How Uber, OCI™, and Ampere® Co-Optimized OCI AmpereOne® M A4 Compute"
description: "Uber, Oracle Cloud Infrastructure, and Ampere co-optimized the OCI AmpereOne M A4 instance family after migrating large workloads from on-premises data centers to Arm-based cloud infrastructure. Early A1 and A2 deployments exposed dual-socket latency, Go performance degradation associated with smaller TLB and cache sizes, and insufficient single-thread performance for Gurobi services. Uber shared workload characteristics and targets that informed AmpereOne M silicon and an OCI FlexSKU, while the teams evaluated frequencies, core counts, deployment models, memory ratios, and storage. For latency-sensitive stateless services, the resulting guidance favors higher-frequency CPUs, about 8 GB per vCore, and virtual machines; throughput-oriented storage systems use parallelism, flexible memory-to-core ratios, and local NVMe or remote block storage according to locality. The collaboration is presented as improving price-performance and supporting energy-efficient, cost-optimized multi-architecture computing at scale."
---

# How Uber, OCI™, and Ampere® Co-Optimized OCI AmpereOne® M A4 Compute

[Uber](https://yomu.fyi/company/uber) · Aug 20, 2026

**Type:** Explainer

## Summary

Uber, Oracle Cloud Infrastructure, and Ampere co-optimized the OCI AmpereOne M A4 instance family after migrating large workloads from on-premises data centers to Arm-based cloud infrastructure. Early A1 and A2 deployments exposed dual-socket latency, Go performance degradation associated with smaller TLB and cache sizes, and insufficient single-thread performance for Gurobi services. Uber shared workload characteristics and targets that informed AmpereOne M silicon and an OCI FlexSKU, while the teams evaluated frequencies, core counts, deployment models, memory ratios, and storage. For latency-sensitive stateless services, the resulting guidance favors higher-frequency CPUs, about 8 GB per vCore, and virtual machines; throughput-oriented storage systems use parallelism, flexible memory-to-core ratios, and local NVMe or remote block storage according to locality. The collaboration is presented as improving price-performance and supporting energy-efficient, cost-optimized multi-architecture computing at scale.

## Context

Uber was moving massive workloads from on-premises data centers to cloud infrastructure while introducing Arm-based compute into an environment dominated by x86. Early OCI Ampere A1 and A2 deployments revealed issues involving cross-socket communication, Go memory management and performance, and single-threaded workloads that could not be addressed through horizontal scaling alone.

## Approach / What changed

Uber, Oracle Cloud Infrastructure, and Ampere used production workload characteristics and performance targets to co-optimize the A4 instance family from silicon through system configuration. They evaluated high-frequency Flex SKUs, bare-metal and virtual-machine deployments, core counts, memory-to-core ratios, DDR5 memory, and local NVMe versus remote block storage, with configurations selected for latency-sensitive or throughput-oriented workloads.

## Takeaways

- Dual-socket systems reduced performance for latency-sensitive operations because of cross-socket link bandwidth; Ampere’s high core counts per socket removed the need for multi-socket servers.
- Uber associated Go performance degradation and GOMEMLIMIT-related out-of-memory errors mainly with smaller TLB and cache sizes, which caused more page faults and slower garbage collection.
- A4 workload guidance separates latency-sensitive stateless services from throughput-oriented storage systems, matching higher frequency and predictable memory behavior to the former and flexible parallel storage configurations to the latter.

**Tags:** [Architecture](https://yomu.fyi/topic/architecture), [ARM](https://yomu.fyi/topic/arm), [Performance](https://yomu.fyi/topic/performance)

- Source: [Uber](https://www.uber.com/us/en/blog/how-uber-oci-and-ampere-co-optimized-oci-ampereone-m-a4-compute)
- Source URL: https://www.uber.com/us/en/blog/how-uber-oci-and-ampere-co-optimized-oci-ampereone-m-a4-compute
- Ingested by Yomu: 2026-09-09T14:22:06.824Z

[Read original post](https://www.uber.com/us/en/blog/how-uber-oci-and-ampere-co-optimized-oci-ampereone-m-a4-compute)