ERPC Confirms Improved Speed and Stability for Dedicated Solana Shredstream (HTTP gRPC / UDP Raw Shreds). Network Baseline Reorganization Enhances Shreds Ingestion and Operational Conditions

ERPC, operated by ELSOUL LABO B.V. (Headquarters: Amsterdam, Netherlands; CEO: Fumitake Kawasaki) and Validators DAO, announces that the speed and stability of Shreds ingestion have improved for its Dedicated Shredstream (HTTP gRPC / UDP Raw Shreds) deployed within the ERPC network.

As a result of this improvement, intermittent latency, jitter, and momentary disconnections that previously occurred on rare occasions during Shreds ingestion have been suppressed, stabilizing communication conditions for continuous operation.

With this update, ERPC has reorganized the baseline assumptions for L2/L3 network configuration and bandwidth utilization across its network infrastructure. As a result, communication conditions for Shreds ingestion have improved for Dedicated Shredstream instances deployed within the ERPC network.

Dedicated Shredstream continuously receives data at the earliest layer where Shreds first flow. At this layer, any jitter or momentary disconnection can immediately propagate its impact to downstream processing. Following this improvement, communication conditions at this front-layer ingestion path have shifted toward greater stability, improving long-duration operation of Dedicated Shredstream.

Previously, in some environments, operational practices required periodic process restarts due to unstable communication conditions. After this network improvement, the need for such restarts has decreased, allowing Dedicated Shredstream to remain operational for longer periods without interruption.

If you continue to experience any irregularities in communication conditions, please contact us via a support ticket on the Validators DAO official Discord. We will review each case individually and, if necessary, assist with verification or migration to a newer node environment.

Validators DAO Official Discord: https://discord.gg/C7ZQSrCkYR

In real-world Solana operations, transaction submission, block production, Shreds detection, state synchronization, and peripheral processing occur continuously at short intervals. These processes do not operate independently; instead, they progress while influencing one another.

If detection is delayed, transaction submission decisions are delayed. If submission is delayed, subsequent detection results are affected. Within this chain of dependencies, Shreds are the earliest data to propagate. Because Shreds flow ahead of RPC, gRPC, and WebSocket layers, communication conditions at this stage define the upper bound for all downstream processing.

Even if average latency appears low in short-term tests, variability or momentary disconnections can destabilize results. In real-world Solana operations, such fluctuations affect detection and submission timing, and their impact propagates through the entire processing pipeline. Communication conditions for Shreds ingestion are therefore a foundational determinant of operational outcomes.

With shared endpoints, concurrent connections from multiple users lead to contention for bandwidth and processing resources. While such contention may not be visible under normal conditions, it becomes apparent during critical peak periods.

When market activity intensifies and overall traffic surges, shared architectures are more prone to degraded conditions. In Solana operations, these peak periods are precisely when performance matters most.

Dedicated Shredstream structurally eliminates this contention, stabilizing communication conditions for Shreds ingestion. Its value lies in maintaining stable conditions precisely when peak pressure is highest.

Shreds delivery via HTTP gRPC is based on TCP streams, offering advantages in implementation and operational handling. TCP incorporates retransmission control, congestion control, and ordering guarantees, ensuring reliable communication.

However, under congested conditions, retransmissions and waiting can occur, delaying initial data arrival. This behavior is inherent to TCP’s design.

UDP-based Raw Shreds lack these control mechanisms and receive data in a form closer to the wire format exchanged between Solana validators. As a result, Shreds can arrive at an earlier stage. This approach assumes potential packet loss and reordering, placing responsibility for reconstruction and decoding on the application.

UDP Raw Shreds are not evaluated solely on communication speed. Significant discretion remains on the implementation side, including Shreds reconstruction, block decoding, and front-layer processing design.

While this discretion introduces implementation complexity, it provides headroom for use cases that require extremely fast detection and initial response. In practice, a staged approach that validates UDP alongside existing gRPC streams is realistic.

ERPC’s Dedicated Shredstream supports parallel use of gRPC (HTTP/TCP) and UDP, enabling validation and development without requiring full migration from existing setups.

Pricing differences for Dedicated Shredstream are not based on feature disparity. They reflect differences in data center placement, network location, and resource quality, each resulting in differences of several milliseconds in average latency and stability.

Dedicated Shredstream deployments using large-scale bare-metal resources can leverage more threads and bandwidth, maintaining performance during peak periods. Resistance to disconnections and stability during reconnections are also improved.

When Premium VPS and Dedicated Premium Shreds are placed within the same network, Shreds ingestion and processing can be completed without traversing the public internet. These communication conditions influence operational outcomes as directly as CPU and memory performance.

Solana is a global project with developers and operators distributed worldwide. When high-performance infrastructure is limited to specific central regions, long-distance connections introduce both latency and traffic concentration.

By providing Dedicated Shredstream globally, ERPC enables each project to validate and operate in its nearest environment. This avoids excessive concentration in central regions and stabilizes overall network conditions.

Because Solana leaders rotate globally, block production origins continuously shift. Under this execution model, expanding geographic coverage is more effective than concentrating resources in a single location.

The open pricing for Dedicated Shredstream will end in January 2026. Contracts entered during January 2026 will retain the current open pricing for as long as the contract remains active.

For configuration consultations, region selection, or guidance on gRPC / UDP usage, please contact us via the Validators DAO official Discord.

Validators DAO Official Discord: https://discord.gg/C7ZQSrCkYR
ERPC Official Website: https://erpc.global/en