ERPC Upgrades RPC and Geyser gRPC Nodes in the Chicago Region to 5th Gen EPYC and DDR5 6400MHz. Enhancing Stability and Catch-up Performance at a DoubleZero-Connected North American Hub

ERPC, operated by ELSOUL LABO B.V. (Headquarters: Amsterdam, the Netherlands; CEO: Fumitake Kawasaki) and Validators DAO, has upgraded the CPU of its Solana RPC (HTTP / WebSocket) and Solana Geyser gRPC nodes in the Chicago region to the latest 5th generation AMD EPYC processors and has begun utilizing DDR5 6400MHz memory.

This upgrade has already been applied to all existing users of the Chicago region, with no configuration changes or migration work required.

The objective of this update is not merely to increase peak performance. As a North American hub where traffic is prone to concentrate, the Chicago region benefits from improved processing headroom under concurrent connections, reduced latency variance, and greater stability under streaming workloads.

Chicago is located in the Midwestern United States and offers efficient access to both the East Coast and the West Coast. For applications and infrastructure targeting North America as a whole, securing a relay point that is not biased toward either coast is important not only for average latency, but also for reducing response variance caused by time-of-day effects and network path fluctuations.

As a city, Chicago also hosts a strong concentration of network carriers and data center-related industries, creating a market environment where the latest generation server hardware tends to become available relatively early. ERPC maintains a policy of applying the newest high-performance models as quickly as possible in a form suitable for production use. This upgrade was implemented first in Chicago because procurement conditions, supply availability, and deployment schedules aligned there earlier than in other regions.

The Chicago region includes a configuration that is directly connected via the Solana-dedicated network DoubleZero to Salt Lake City (SLC) and Los Angeles (LA). This setup suppresses the impact of route changes and congestion that commonly occur on the public internet, enabling low-latency and stable data transmission across the East–West axis.

Despite its proximity to eastern hubs such as New York (NY) and Pittsburgh (Pitt), Chicago also maintains favorable network conditions to western locations like SLC and LA, allowing it to function as an effective relay point that uniformly covers North America.

Based on ERPC’s operational observations, Chicago is positioned as one of the locations in the United States with a scale of validator aggregation second only to New York. In the Solana network, locations with a high concentration of validators tend to become critical points for data propagation, which directly affects catch-up behavior and stability for RPC and Geyser gRPC. By combining DoubleZero-based dedicated network paths with its observation and distribution node placement, ERPC continues to strengthen a configuration that minimizes disadvantages arising from regional and routing factors.

In real-world operation of Solana RPC and Geyser gRPC, workloads such as concurrent connections, short-cycle bursts, increasing filter complexity, and reconnection or resubscription events occur simultaneously. With this generational upgrade, ERPC focuses not only on higher CPU processing capability but also on improved memory access performance, suppressing wait times and variance during load concentration and enhancing overall stability.

Updating compute resources in Chicago, a central North American hub, contributes not only to the region itself but also to increased operational headroom across North America and, by extension, global deployments. ERPC continuously accumulates improvements that raise not just average performance, but also the practical limits and consistency that matter most in production environments.

This upgrade is reflected in the underlying infrastructure for Solana RPC (HTTP / WebSocket) and Solana Geyser gRPC provided in the Chicago region. By combining shared RPC, Unlimited endpoints, standalone Geyser gRPC plans, gRPC Bundle plans, and ERPC VPS within the same network, users can build a consistent environment from proof-of-concept through full production.

ERPC provides components such as RPC, Geyser gRPC, Unlimited endpoints, Shredstream, and VPS consolidated within the same region and the same internal network. This design enables low-latency and stable communication entirely within the internal network, without traversing external internet paths or inter-region traffic.

In particular, configurations that place RPC, Geyser gRPC, and application execution environments (VPS) in close proximity help suppress round-trip latency and variance, resulting in more stable behavior during reconnections and high-frequency access. Through its same-network design, ERPC delivers overall catch-up performance and consistency that cannot be achieved by improving individual components alone.

For consultation on optimal region selection, plan choice, or migration design from existing setups, please contact us via the Validators DAO Official Discord.