Menu

New to You, Phison Pascari Enterprise SSD Range for Hardcore AI Workloads

New to You, Phison Pascari Enterprise SSD Range for Hardcore AI Workloads

Analyst(s): Alastair Cooke
Publication Date: June 20, 2025

What is Covered in this Article:

  • Phison’s history of building SSDs with custom controllers for special uses
  • The new challenges of AI in data centres, particularly data storage and transfer challenges
  • The breadth and depth of the Phison Pascari Enterprise SSD range

The Event – Major Themes & Vendor Moves: AI Infrastructure Field Day is a semiannual, invitation-only event held in Santa Clara, organized by Tech Field Day. Independent industry experts join presenting companies to learn about product innovations. The event is live-streamed, and videos are published on the Tech Field Day YouTube channel.

New to You, Phison Pascari Enterprise SSD Range for Hardcore AI Workloads

Analyst Take: We recognize that data volumes continue to grow, and both the volume of data and the need for speed have increased accordingly. AI applications are the latest category to push the boundaries of storage capabilities. You may not be familiar with Phison, or be unaware that you likely have their products in your possession. Phison builds SSDs with its in-house developed controllers and custom firmware for specific uses. Until recently, these SSDs were all sold with another company’s label. Phison has introduced direct branding and sales for the Pascari range of SSDs. Customers can purchase individual drives from Newegg or find them inside server and storage products, such as the recently released VDura V5000 all-flash 1U storage server. Placing 12 of Phison’s 122TB Pascari drives in a 1-U server allows VDura a raw capacity of 1.5 PB per rack unit.

AI Data Challenges

Data governance is a crucial aspect of AI data, particularly for inference, where production data may include personally identifiable information (PII) or sensitive health data subject to HIPAA requirements. These requirements may prevent an application from utilizing a cloud-based inference service and may require AI infrastructure to be on-premises. As a data center workload, AI is characterised by high compute load and large amounts of data movement. The compute load is typically satisfied by GPUs connected to a 200 Gb Ethernet network with NVMe SSDs in storage servers attached to that network. For training and fine-tuning generative AI models, large numbers of GPUs work in parallel to process large amounts of training data, resulting in massively parallel access to NVMe SSDs. NVMe SSDs are essential as they have multiple parallel queues, providing lower latency and faster throughput to the GPUs. Model checkpoints and results must also be stored on these SSDs, so the storage load alternates between sequential reads and writes. The same storage may be used for inference, where the trained AI delivers answers into an application. Inference is a random read and write workload, resulting in a highly mixed load on the storage. AI data centers demand large amounts of high-performance NVMe SSDs to enable these new access patterns, and the Phison Pascari Enterprise SSD range meets these requirements.

Phison Pacari Enterprise SSD range

The Pascari D-series SSDs are designed for dense data center use. The D205V is a 122.88TB monster in U.2 or E3.L form factor, designed to integrate into storage servers with a PCIe Gen 5 x 4 dual-port interface, delivering up to 14 GB/s of sequential read and three million random read IOPS per drive. The D-series is designed for more capacity and read-intensive workloads, or as the capacity tier in a storage system. The X-series is the speed demon for writes, making an ideal cache tier in a storage system; nearly a million random write IOPS, and endurance for 60 Drive writes per day (DWPD). These drives can handle a significant amount of random I/O without experiencing performance problems or failures. The Pascari range also includes SATA SSDs for NAS devices, SSDs suitable for operating system boot disks, and some SSDs optimized for use with Phison aiDAPTIV+ technologies to expand GPU memory for large models.

Phison has long led customized solid-state drive (SSD) solutions. With the release of the Phison Pascari enterprise SSD range, those innovative products are now available with a Phison label. The Pascari range encompasses SDDs for desktop and server deployment, offering a variety of price, performance, and capacity options, as well as different packaging options.

What to Watch:

  • Several vendors offer their ranges of SSDs to the market, many of which overlap with the capabilities of the Pascari range. These incumbents, some of which are Phison’s clients, will pose the most significant challenge for Phison as a new entrant in the enterprise market.
  • Enterprise SSDs, especially the largest and fastest ones, are primarily shipped to enterprises within storage servers and arrays, or engineered solutions. You may still not be aware that you have Phison Pascari Enterprise SSDs in your storage system.
  • Phison’s aiDAPTIV+ is an exciting product in its own right. It enables GPUs to work with AI models and applications that would otherwise be too large for the GPU memory.

The Phison presentations at AI Infrastructure Field Day are available on their appearance page. You can watch all the presentations from the four days of AI Infrastructure Field Day on the Tech Field Day website.

Disclosure: Futurum is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.

Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of Futurum as a whole.

Other insights from Futurum:

Six Five Connected: The Need For High-Density Storage in The Age of AI

The Latest from GTC: Announcements on Data Infrastructure and How the AI Markets are Shaping Up – Six Five Webcast – Infrastructure Matters

The Futurum Group Report on Flash Memory Summit

Author Information

Alastair has made a twenty-year career out of helping people understand complex IT infrastructure and how to build solutions that fulfil business needs. Much of his career has included teaching official training courses for vendors, including HPE, VMware, and AWS. Alastair has written hundreds of analyst articles and papers exploring products and topics around on-premises infrastructure and virtualization and getting the most out of public cloud and hybrid infrastructure. Alastair has also been involved in community-driven, practitioner-led education through the vBrownBag podcast and the vBrownBag TechTalks.

Related Insights
Elastic Q3 FY 2026 Strong Quarter, but Reacceleration Thesis Unproven
March 3, 2026

Elastic Q3 FY 2026: Strong Quarter, but Reacceleration Thesis Unproven

Nick Patience, VP and Practice Lead for AI Platforms at Futurum reviews Elastic Q3 FY 2026 earnings, highlighting sales-led subscription momentum, AI context engineering adoption, and agentic workflow expansion across...
CoreWeave Q4 FY 2025 Results Highlight Backlog Growth And Capacity Expansion
March 3, 2026

CoreWeave Q4 FY 2025 Results Highlight Backlog Growth And Capacity Expansion

Futurum Research reviews CoreWeave’s Q4 FY 2025 earnings, focusing on backlog-driven capacity expansion, platform monetization beyond GPUs, and execution cadence shaping AI infrastructure supply....
Snowflake Q4 FY 2026 Results Highlight AI-Led Consumption and Platform Expansion
March 2, 2026

Snowflake Q4 FY 2026 Results Highlight AI-Led Consumption and Platform Expansion

Brad Shimmin, Vice President & Practice Lead at Futurum analyzes Snowflake’s Q4 FY 2026 earnings, highlighting AI-driven consumption growth, expanding platform scope, and guidance shaping expectations for FY 2027....
Collapsing the Stack VAST Data’s Bid to Own the AI Data Loop
February 27, 2026

Collapsing the Stack: VAST Data’s Bid to Own the AI Data Loop

Brad Shimmin, Vice President at Futurum, analyzes the VAST Data platform updates from VAST Forward, detailing how the new Policy Engine, Tuning Engine, and Polaris architectures are simplifying the AI...
Are Enterprises Ready for the Virtualization Reset, or Just Swapping Out One Complexity for Another
February 27, 2026

Are Enterprises Ready for the Virtualization Reset, or Just Swapping Out One Complexity for Another?

Futurum’s Alastair Cooke shares his insights on new HPE research that finds that only 5% of enterprises are fully prepared for the so-called Great Virtualization Reset, even as two-thirds plan...
NVIDIA Q4 FY 2026 Earnings Highlight Durable AI Infrastructure Demand
February 27, 2026

NVIDIA Q4 FY 2026 Earnings Highlight Durable AI Infrastructure Demand

Futurum’s Nick Patience analyzes NVIDIA’s Q4 FY 2026 earnings, highlighting data center scale, networking expansion, and agentic AI adoption shaping AI infrastructure demand....

Book a Demo

Newsletter Sign-up Form

Get important insights straight to your inbox, receive first looks at eBooks, exclusive event invitations, custom content, and more. We promise not to spam you or sell your name to anyone. You can always unsubscribe at any time.

All fields are required






Thank you, we received your request, a member of our team will be in contact with you.