Introducing Telnyx Storage: Low-Cost Object Storage

We're excited to announce Telnyx Cloud Storage: cloud object storage powered by distributed infrastructure for lower costs and lower latency. Our S3-compatible APIs will help you migrate, and we'll never charge egress fees, so you can scale your apps and experiences with confidence.

Update November 8, 2022: Telnyx Cloud Storage is now available for you to try for free. Simply log into your Mission Control Portal account and start building on more affordable, low-latency object storage.

Telnyx Cloud Storage fills a need that cloud giants can't meet

Cloud object storage has quickly become a cornerstone of the web with apps and services relying on S3 or a comparable service from other cloud giants.

We believe the next generation of connected experiences will outgrow the storage services provided by cloud giants. Augmented reality and virtual reality (AR/VR), autonomous vehicles, and large 8K content delivery will not just require high bandwidth, but low latency to serve objects and feed edge computation and decision making.

Alongside latency problems, the large file sizes at play in content-rich experiences mean that high per-GB storage costs and egress fees will become unsustainable for anyone looking to build an experience with large content demands.

Telnyx Cloud Storage saves you money on cloud object storage

Storing objects with Telnyx Cloud Storage costs $0.006 per GiB, per month. That's more than a 60% cost saving over S3's standard entry-level storage costs. We'll never charge egress fees, so you can export your data without worry. Plus, our operations costs are lower than S3's so you can build, test, and scale your applications for less.

Learn more about Telnyx Storage pricing.

Telnyx Cloud Storage gives your apps lower-latency access to objects

We're building Telnyx Cloud Storage by deploying modern, distributed infrastructure around the US. By storing objects closer to end-users, Telnyx Cloud Storage customers can access and serve their objects to end-users faster, shaving milliseconds off latency.

To illustrate how low latency can transform an end-user's experience and enable low-latency, content-rich use cases, let's use an example. In autonomous driving applications serving small files to vehicles moving at highway speeds, 80 milliseconds of latency (typically seen in centralized cloud storage services) means vehicles will move four feet between requesting and receiving a response. Cutting that latency to 5 milliseconds with edge data centers means those same vehicles will move less than four inches between request and response. Reducing latency in object retrieval could enable latency-sensitive applications like this one to run almost entirely in the cloud.

Telnyx Cloud Storage gives you security, compatibility and data sovereignty

Telnyx APIs to store, manage, and access objects are fully compatible with S3, so you can simply point your S3-centric applications and services at Telnyx Storage API endpoints. No more wasted development cycles during painstaking migrations.

Our S3-compatible API supports native encryption, so you can encrypt objects when adding them to a Telnyx Storage bucket.

Telnyx Cloud Storage is for anyone building next-gen connected experiences

Telnyx Cloud Storage is for businesses that are building backup and restore applications, autonomous vehicles, 8K content delivery, or any field that requires data storage at low cost, with low latency.

Get in touch with a member of our team today to see how Telnyx can help you with your Cloud Storage needs.

FAQ

What is low latency storage? Low latency storage is a system optimized to return reads and writes in microseconds, prioritizing fast response over peak throughput. Latency is affected by media type, interface, queue depth, and payload size, similar to how larger content in MMS messaging takes longer to transmit than plain text.

Should I turn low latency on or off? Turn it on for interactive workloads like OLTP databases, real-time analytics, or user-facing APIs where tail latency shapes user experience. Leave it off for batch or large sequential jobs where throughput dominates, much like choosing compact texts in SMS vs. MMS when speed is the only goal.

Which memory storage technology has the lowest latency? Static RAM has the lowest latency among common memory types, followed by DRAM and high-bandwidth memory on accelerators. For persistent storage, storage-class memory once neared DRAM latency but is uncommon now, with NVMe SSDs the fastest mainstream option.

Which SSD has the lowest latency? Enterprise NVMe SSDs on PCIe 4.0 or 5.0 with efficient controllers and firmware deliver the lowest latency at low queue depths. Look for drives tuned for consistent p99 response times, power-loss protection, and SLC or low-latency TLC configurations rather than peak IOPS claims.

Is latency or throughput more important for my workload? Prioritize latency for small, random I/O with user impact, like key-value lookups and transaction logs, where p95 and p99 response times govern experience. Favor throughput for large sequential transfers like backups and media processing, where payload characteristics defined by messaging types map to predictable transfer patterns.

How do I measure storage latency? Use tools like FIO, iostat, or Windows DiskSpd to benchmark with realistic block sizes, concurrency, and read write mixes. Track median, p95, and p99 latency under sustained load, and watch for jitter when device caches or write buffers fill.

What configurations help reduce storage latency? Use NVMe over PCIe, pin critical threads, align I/O with NUMA, keep queue depths shallow for small requests, and choose I/O schedulers that favor minimal queuing. Control concurrency at the application layer to smooth tail latency, similar to how bulk MMS sending shapes parallel deliveries to avoid spikes.