ERPC Launches Overcommit-Free SUPER VPS in Amsterdam — Delivering Top-Tier Peak Performance and Stability for Solana

ERPC, operated by ELSOUL LABO B.V. (Headquarters: Amsterdam, the Netherlands; CEO: Fumitake Kawasaki) together with Validators DAO, announces the expansion of its high-performance VPS offering for Solana, “SUPER VPS,” with a new availability zone in Amsterdam, in addition to the existing Frankfurt location.

SUPER VPS is designed with no compromises on peak performance, featuring high-clock CPUs capable of reaching up to the 5.7 GHz class, representing one of the highest performance tiers available today. In addition, to ensure that this peak performance remains stable under real-world workloads, SUPER VPS is operated with a strict no-overcommit policy. The service is optimized from the ground up—both in design and operations—for Solana use cases that demand peak performance and stability at the same time.

All SUPER VPS instances currently offered—both in Frankfurt and Amsterdam—are operated under this no-overcommit configuration, ensuring consistent performance characteristics across regions.

ERPC’s VPS and bare-metal products are not designed under the same assumptions as general-purpose cloud services. In real-world Solana operations, bottlenecks are often caused not only by CPU or memory specifications, but also by data center operational policies, network topology, and behavior under congestion.

To address this, ERPC has leased and tested data centers worldwide while running actual Solana workloads. Factors such as congestion behavior, tolerance for sustained turbo CPU operation, suitability of the facility as a network hub, and routing quality cannot be determined from marketing materials or specification sheets alone. Through this process, ERPC has continuously eliminated locations unsuitable for high-performance Solana workloads, narrowing its infrastructure to facilities that consistently deliver stable, high performance. SUPER VPS is deployed exclusively on this carefully selected foundation.

Solana applications and bots often operate in situations where peak processing capability directly determines outcomes. This is particularly true for transaction submission, stream processing, and real-time analytics, where single-thread CPU performance and instantaneous compute capacity translate directly into results.

ERPC’s approach begins by maximizing peak performance through the use of high-clock CPUs. At the same time, the design ensures that this peak performance does not degrade due to time-of-day effects or neighboring workloads. Unlike many general-purpose VPS offerings that rely on CPU overcommit—assigning more virtual cores than physically available—SUPER VPS completely avoids overcommitment, structurally eliminating CPU contention, performance degradation, and latency jitter.

With this release, SUPER VPS becomes available in a new Amsterdam region. Amsterdam is one of the locations with the highest concentration of Solana validators and staked SOL, following Frankfurt.

For high-performance Solana workloads, physical proximity alone is not sufficient. What matters more is placement within a network hub where a large number of validators and stake are concentrated, as this directly impacts latency and stability. ERPC selects its deployment regions based on validator density and stake concentration as key decision criteria.

Following Frankfurt as the first location, Amsterdam was chosen as the second region due to these network-level advantages. Looking ahead, ERPC plans to expand SUPER VPS availability to New York next year, continuing to focus exclusively on major Solana network hubs.

To evaluate the real-world performance of SUPER VPS, comparative benchmarks were conducted under identical conditions using the same measurement tools. SUPER VPS (4 vCPU / 16 GB RAM) was compared against a similarly positioned Google Cloud configuration (4 vCPU / 15 GB RAM).

The benchmarks focused on fundamental metrics that directly affect Solana operations, including CPU processing performance, disk I/O, and latency. In addition to peak values, variability and stability were also examined to reflect actual production usage.

While the magnitude of differences varies by metric, the results show that CPU, memory, and storage performance demonstrate numerical gaps ranging from approximately 2× up to around 25× when compared with an equivalently sized cloud configuration. These differences reflect architectural choices such as CPU clock characteristics, the absence of overcommitment, and storage configuration.

As a result, SUPER VPS demonstrates not only higher peak performance, but also more consistent stability and responsiveness under real-world workloads.

The benchmarking tools and metrics used for these comparisons are published as open source in the following GitHub repository:

Rather than presenting isolated benchmark results, ERPC emphasizes transparency by publishing the measurement methodology itself. CPU, memory, and disk performance can vary significantly depending on tools and conditions, making reproducibility under identical conditions essential.

This repository includes both the tools used for the comparison and the actual benchmark results. Additional test categories—such as network performance metrics that are increasingly critical for blockchain-era workloads—will be added over time. The structure also allows for third-party verification and submission of results from other environments via pull requests.

The latest configurations and pricing for SUPER VPS are available in the pricing table below.

SUPER VPS is not designed as a general-purpose VPS. Its pricing reflects the accumulation of design choices that are meaningful for real-world Solana operations, including data center selection, CPU operating policies, and a strict no-overcommit architecture.

ERPC will continue to refine its infrastructure with a focus on real-world Solana operations, prioritizing the coexistence of peak performance and stability. Across VPS, bare-metal servers, and streaming products, ERPC will keep investing in and validating each layer—from CPU and network design to storage and data center selection.

For configuration details, availability, or consultations tailored to specific use cases, please contact us via the official Validators DAO Discord.