🔍 Read the full analysis: Singapore Data Centers Find New Capacity In Idle SAP HANA Memory on Rymvard

Prime Big Deal Days · Oct 6–7Offer from Amazon

Get the latest gadgets delivered free — and shop member deals

• Fast, free delivery on millions of items
• Access to Prime Big Deal Days deals on October 6–7
• Prime Video, Amazon Music and more included

Start your free Prime trial →Free trial for eligible customers · Cancel anytime

As an affiliate, we earn on qualifying purchases.Hide

TL;DR

Rymvard says its running early-access product can identify SAP HANA memory that has been allocated but lightly used, helping hosting providers assess capacity inside existing data centers. Its Singapore scenario is illustrative, not a report of a customer deployment or verified capacity gains; the company says recovery checks remain part of right-sizing.

Rymvard says its early-access product can help SAP hosting providers find capacity in existing Singapore data centers by identifying SAP HANA memory allocated but used sparingly. The company’s example describes how operators could right-size allocations while retaining memory needed for system recovery, but it does not document a customer deployment or measured capacity gains.

Rymvard’s tool compares each HANA system’s configured allocation limit with its peak memory use over a 30-day period. It flags hosts that are overbooked or above 90 percent of capacity, and identifies systems that used less than 10 percent of their allocation throughout the period. Those signals are intended to help operators investigate whether allocations can be reduced.

The company says its host-level view separates physical memory into operating-system reserve, HANA systems’ peak use, allocated but unused memory, memory reserved for replicas at other sites, and genuinely free capacity. The default operating-system reserve is 10 percent of the first 64 GiB and 3 percent of remaining memory, unless a host declares a different reserve. Allocations exceeding physical memory appear as overbooking rather than being represented as negative free memory.

Rymvard says each replicated system pair retains a takeover check during right-sizing. That is intended to keep recovery needs visible rather than treating memory reserved for failover as available to reclaim. The product is running in early access, but the screens and Singapore regional hosting-hub scenario described by the company use an illustrative example estate, not a named customer environment.

Finding Capacity Without New Halls

Singapore’s data-center supply is subject to government allocation, making better use of installed infrastructure a potential alternative to adding physical space. The government’s 2023 pilot Data Centre Call for Application selected four proposals totaling about 80 MW, and a second call opened in December 2025. In that setting, identifying memory that is reserved but rarely used could help SAP hosting operators examine whether existing facilities can support more workload.

The practical value remains unproven in the material provided. Rymvard describes a method for exposing memory commitments and recovery reserves; it does not provide customer results, a quantified amount of memory recovered, or evidence that identified memory can always be safely reassigned. For operators, the distinction matters: memory that appears idle in ordinary use may still be needed for a workload surge or a recovery event.

Rymvard’s treatment of replica memory addresses that operational concern by keeping recovery checks in the assessment. Singapore banks, under MAS Notice 644 on Technology Risk Management, must set a recovery time objective of no more than four hours for each critical system. That regulatory requirement does not itself establish that any particular memory reduction is safe, but it makes recovery capacity a relevant constraint when operators consider changes.

Singapore’s Selective Data-Center Supply

Singapore has used application calls to allocate new data-center capacity. In the 2023 pilot, four proposals were selected for about 80 MW in total; a second call opened in December 2025. The supplied details do not state how much capacity that later call will allocate or what decisions have been made.

Unlike adding a new data-center hall, reviewing software memory allocations focuses on resources already present in a hosting environment. HANA systems may have allocation limits substantially above recent usage, according to Rymvard’s description. A 30-day peak-use comparison can surface candidates for review, but it is not by itself proof that the unused portion can be removed without affecting service.

Rymvard is an early-access product, and the company says pricing is agreed with early-access partners rather than publicly listed. Its Singapore hub scenario is illustrative: no customer, site, or operational outcome is identified or implied.

“Systems that used less than a tenth of their allocation for 30 days”

— Rymvard

Customer Results Remain Unreported

Rymvard has not provided a named customer deployment, a measured amount of memory recovered, or results showing how often flagged allocations can be reduced safely. The company characterizes the Singapore hosting-hub case as illustrative and says the product screens come from a running example estate rather than a customer.

It is also unclear how many providers are participating in early access, what workloads or HANA configurations have been evaluated, and whether the tool’s alerts have led to changes in production environments. No public pricing is available; Rymvard says terms are agreed with early-access partners. The status and outcome of Singapore’s December 2025 data-center capacity call are not specified here.

Early Access and Capacity Decisions

Rymvard says the product is currently running in early access. The company has not announced a general-availability date or published a schedule for customer results. The next evidence needed to assess its practical value would include deployments in real hosting environments, quantified changes to memory allocations, and details on how operators validated failover capacity after right-sizing.

For Singapore providers, any decision to reclaim memory remains an operational review, not an automatic consequence of a dashboard flag. Operators would need to assess workload behavior, physical limits, replica requirements, and recovery objectives before changing allocations. Future decisions on government-allocated data-center capacity may also clarify how much room providers have to expand through new facilities.

Primary source: IMDA · via Rymvard

Key Questions

What is Rymvard offering?

Rymvard says its early-access product analyzes SAP HANA memory allocations against peak use over 30 days, highlighting overbooked hosts and systems with low utilization for review.

Has the tool freed capacity at a Singapore customer site?

No customer deployment or measured capacity gain is reported. Rymvard describes its Singapore hosting-hub example as illustrative and says the screens are from an example estate, not a customer.

How does the product account for disaster recovery?

Rymvard says replicated system pairs retain a takeover check, and memory reserved for replicas of other sites is shown separately. The company says this is intended to keep recovery needs visible during right-sizing.

Why does the four-hour recovery objective matter?

Singapore banks must set a recovery time objective of no more than four hours for each critical system under MAS Notice 644 on Technology Risk Management. It makes recovery capacity a relevant consideration, but does not establish that a particular memory reduction is safe.

Is Rymvard generally available, and what does it cost?

The product is in early access. Rymvard has not published prices and says pricing is agreed with early-access partners.

Primary source: IMDA · via Rymvard