Overview Getting Started S3 API Regions Consistency Examples API Ref

Documentation

ACS Documentation

Complete guide to integrating with Accelerated Cloud Storage. Learn how to use our S3-compatible API, understand our global infrastructure, and get started in minutes.

Overview

Accelerated Cloud Storage (ACS) provides globally distributed, performant, S3-compatible object storage designed for AI workloads. Our service offers enterprise-grade reliability with 99.99% availability and 99.999999999% data durability.

Key Features

S3-compatible API
Global data distribution
Strong consistency within regions
SOC-2 Type 2 compliant

Quick Start

1Contact ACS for API key
2Update endpoint URL
3Start using ACS

Getting Started

Step 1: Get Your API Key

Contact our team to get onboarded and receive your credentials. We'll provide you with an access key ID and secret access key.

Email Sales Schedule Meeting

Step 2: Update Your Code

Make a single change to your existing S3 code by updating the endpoint URL and client configuration.

main.py

import boto3

import os

from botocore.config import Config

s3 = boto3.client(

"s3",

endpoint_url="https://acceleratedprod.com",

region_name="global",

aws_access_key_id=os.getenv("AWS_ACCESS_KEY_ID"),

aws_secret_access_key=os.getenv("AWS_SECRET_ACCESS_KEY"),

config=Config(signature_version="s3v4", s3={"addressing_style": "virtual"}),

)

Step 3: Start using ACS

Use ACS with your existing S3-compatible tools and libraries.

example.py

import boto3

import uuid

# Generate a unique bucket name for every run

bucket_name = f"acs-examples-bucket-{uuid.uuid4().hex[:8]}"

object_key = "goals.txt"

# Create a bucket

s3.create_bucket(Bucket=bucket_name)

# Upload an object

s3.put_object(Bucket=bucket_name, Key=object_key, Body=b"We're building the storage layer for AI!")

# Fetch and print the object content

print(s3.get_object(Bucket=bucket_name, Key=object_key)["Body"].read().decode("utf-8"))

# Delete the object

s3.delete_object(Bucket=bucket_name, Key=object_key)

# Delete the empty bucket

s3.delete_bucket(Bucket=bucket_name)

S3 API Compatibility

ACS, as an S3‑compatible service, supports the modern S3 authentication and authorization mechanisms recommended by AWS for AuthN and AuthZ workflows. ACS provides a single authentication mechanism: AWS Signature Version 4.

This method uses an HMAC-based signature derived from a canonicalized request string and a signing key computed from the secret access key; standard S3 SDKs generate the signature and Authorization header for clients. The server validates requests by recomputing the signature using the provided access_key_id and comparing it to the incoming signature to establish the authenticated principal's context.

Authentication Details

• Uses AWS Signature Version 4 (SigV4)
• Access keys are generated by ACS
• Keys are encrypted at rest with AES‑256
• Compatible with all standard S3 SDKs

Global Infrastructure

ACS is deployed to 10 regions around the world, ensuring your data is always close to your users for optimal performance and reliability.

North America

us-v - United States, Virginia
us-oh - United States, Ohio
us-ca - United States, California
us-or - United States, Oregon
ca-q - Canada, Quebec

International

uk-l - UK, Greater London
de-he - Germany, Hesse
jp-13 - Japan, Tokyo Prefecture
kr-11 - South Korea, Seoul
au-nsw - Australia, New South Wales

Consistency Model

ACS provides strict read-after-write consistency within the same region by default, and eventual consistency across regions globally.

How It Works

In practice, a write performed in the US Virginia region will be strongly consistent when read from US Virginia, while reads from another region like US Oregon may briefly return stale data and serve an older version under the default model.

This behavior reflects the default consistency policy, which favors local strong guarantees with globally eventual replication for performance and latency benefits.

Within Regions

Strong consistency - immediate read-after-write guarantees

Across Regions

Eventual consistency - optimized for global performance

Examples & Code Snippets

Explore our comprehensive collection of examples and code snippets to get started with ACS quickly.

Complete Examples Repository

Visit our GitHub repository for detailed examples, tutorials, and integration guides.

View Examples on GitHub

API Reference

ACS will continue expanding its S3 compatibility by iterating toward broader coverage of standard IAM, bucket and object-level APIs while preserving drop‑in use of AWS S3 SDKs, tooling, and libraries.

Bucket Level APIs

• ListBuckets

• CreateBucket

• HeadBucket

• DeleteBucket

Object Level APIs

• ListObjectsV2

• ListObjects

• PutObject

• GetObject

• HeadObject

• DeleteObject

• CopyObject

• CreateMultipartUpload

• UploadPart

• UploadPartCopy

• CompleteMultipartUpload

• AbortMultipartUpload

• ListParts

• ListMultipartUploads

Advanced Features

• Pre-signed URLs for secure access
• Conditional parameters for efficient operations