System Requirements Document for windows-11

1. Introduction

This document specifies the system requirements for windows-11, a bilingual (Arabic/English) deep-research report and complete optimization plan for maximizing gaming performance on Windows 11 25H2, with a specific focus on reducing RAM consumption by the system and its services while a game is launching and running, and on machines that rely on an integrated GPU (iGPU).

The product intent is derived directly from the authoritative user requirement thread: the user asked for deep research across internet articles, YouTube content, and other sources in both Arabic and English, covering tips for improving gaming performance on Windows 11 25H2; a specific emphasis on reducing RAM consumption by the system and services during game launch; coverage of integrated-GPU systems; identification of services and features that consume device resources while games are running; and a comprehensive report containing a complete plan to achieve maximum system performance while launching a game.

The audience is Arabic-speaking PC gamers running Windows 11 25H2 on integrated-GPU machines who are technical, self-reliant, and skeptical of generic "boost your FPS" content. They want measured evidence, service-by-service breakdowns, and RAM accounting, and they want to apply a complete, ordered plan on their own machine.

2. System Overview

The system is a first-party, publicly readable, bilingual research and planning surface. It delivers:

• A Landing entry that explains the research, its Windows 11 25H2 focus, its RAM/services/features/iGPU scope, and how to review the report and plan.
• A Research Findings surface that organizes deep gaming-performance findings gathered from internet, YouTube, and Arabic and English sources for Windows 11 25H2.
• A Resource Audit surface that presents Windows services and features that consume system resources during game launch, including integrated-GPU considerations and RAM usage.
• An Optimization Plan surface that provides the comprehensive, readable, step-by-step plan for achieving maximum Windows 11 25H2 gaming performance while launching a game.

All four surfaces are application-owned custom pages with access_requirement: none — the content is publicly readable and no account, sign-in, or identity continuity is required by the accepted scope. There is no user-generated state, no private per-user data, and no commitment, entitlement, or value transfer that must remain bound to a specific participant, so no application-owned identity is introduced.

The accepted active human personas are exactly two: Windows 11 Gaming Optimizer and Research Report Reader. No other human personas are introduced.

2a. Product Interpretation and Delivery Boundary

The product is a read-and-apply deliverable, not a system-tuning utility. It does not modify the reader's Windows installation, does not install software, does not run as a background service, and does not connect to the reader's machine. Every optimization action described in the report and plan is performed by the reader on their own Windows 11 25H2 device, outside this product, using the guidance the product presents.

Delivery is first-party custom UI with generated documents as the content form: the research findings, the resource audit, and the optimization plan are authored report content rendered as readable, navigable surfaces. The product owns the presentation, the bilingual source filtering, the service-by-service audit table, and the ordered plan; it does not own the reader's operating system, the games, the drivers, or any third-party tuning tool.

Current scope is limited to Windows 11 25H2 gaming performance, RAM and service consumption during game launch, integrated-GPU systems, and the identification of resource-consuming services and features. Anything outside that boundary — general Windows tuning unrelated to gaming, hardware purchasing advice, overclocking, or non-Windows platforms — is out of scope and is not part of the current product.

2b. Source Content Inventory

The reference directives declare content_source for two source families, and their verified factual content is inventoried below at the level the directives support. The directives name the source families and their language coverage but do not supply individually verified article titles, video titles, channel names, URLs, publication dates, or per-source numeric measurements. Accordingly, this inventory records the verified source families, their declared language coverage, and the declared subject matter, and does not fabricate specific titles, links, dates, or figures that were not verified.

Source family 1 — Internet articles and guides on Windows 11 25H2 gaming performance optimization

• Declared uses: content_source, domain_context.
• Declared instruction: gather tips from Arabic and English web sources.
• Verified content supplied: tips for improving gaming performance on Windows 11 25H2, drawn from Arabic-language and English-language web sources.
• Verified subject coverage: gaming performance optimization on Windows 11 25H2; RAM consumption reduction by the system and services during game launch; services and features that consume device resources while games run; integrated-GPU (iGPU) systems.

Source family 2 — YouTube videos on Windows 11 gaming performance and RAM reduction

• Declared uses: content_source, domain_context.
• Declared instruction: gather tips from Arabic and English YouTube content.
• Verified content supplied: tips for Windows 11 gaming performance and RAM reduction, drawn from Arabic-language and English-language YouTube content.
• Verified subject coverage: gaming performance optimization on Windows 11 25H2; RAM reduction; services and features consuming device resources during gameplay; integrated-GPU (iGPU) systems.

Language coverage (verified): Arabic and English, for both source families.

Target platform (verified): Windows 11 25H2.

Focus areas (verified): reducing RAM consumption by the system and services during game launch; identifying services and features that consume device resources while games are running; integrated-GPU (iGPU) systems; a complete plan for maximum system performance while launching a game.

No specific article titles, video titles, channel names, URLs, publication dates, or per-source numeric measurements were verified by the directives, and none are invented here. Where the product presents findings, each finding is attributed to its source family and language so the reader can distinguish Arabic-sourced from English-sourced guidance.

2c. Page Content and Component Coverage

Landing

• Information and state: the product's purpose — a deep-research report and complete optimization plan for gaming performance on Windows 11 25H2; the specific focus on reducing RAM consumption by the system and services during game launch; the integrated-GPU (iGPU) case; the identification of services and features that consume device resources while games run; the bilingual Arabic/English source basis; and how to review the report and the plan.
• Primary action: begin the full plan (navigate to the Optimization Plan).
• Supporting actions: open the Research Findings; open the Resource Audit.
• Domain entities: research scope, target platform (Windows 11 25H2), focus areas (RAM, services, features, iGPU), source languages (Arabic, English), report, plan.
• Component responsibilities:
  • Kinetic bilingual headline spanning the viewport, with the Arabic line carrying the Windows 11 25H2 subject and the second line carrying the integrated-GPU maximum-performance subject.
  • Live RAM bar-field canvas behind and beside the headline, driven by a plausible RAM-usage series, signalling that the numbers are measured.
  • A full-width hairline rule beneath the headline with the primary capsule CTA pinned under its left end, plus a secondary muted text link.
  • A scrolling service ticker band naming Windows services and features relevant to game-launch resource consumption.
  • A three-column measured-metrics strip summarizing the report's headline measurements (for example RAM at idle vs. RAM during game launch, and iGPU shared-memory impact).
  • A one-line scope statement naming Windows 11 25H2, RAM, services, features, and integrated GPU.
• States:
  • Loading: the bar-field canvas initializes and the headline lines settle into their composed first frame.
  • Empty: not applicable — the Landing always carries the accepted scope statement and entry actions.
  • Success: the reader understands the research scope and can reach the report and the plan.
  • Error/recovery: if the bar-field canvas cannot render, the headline, scope statement, ticker, metrics strip, and both entry actions remain fully legible and operable as static content.
  • Reduced motion: the bar-field holds a static composed frame, headline reveals are not staggered, and no continuous loop runs.

Research Findings

• Information and state: deep gaming-performance findings for Windows 11 25H2, gathered from internet articles and guides and from YouTube content, in both Arabic and English; each finding presented as a numbered entry with its source-language attribution and its subject area (RAM reduction during game launch, services and features consuming device resources during gameplay, integrated-GPU considerations, general gaming performance).
• Primary action: read and filter findings by source language.
• Supporting actions: follow a finding through to the Resource Audit or the Optimization Plan where it maps to a service or a step.
• Domain entities: finding, finding number, source family (internet article/guide, YouTube), source language (Arabic, English), subject area, target platform (Windows 11 25H2).
• Component responsibilities:
  • Sticky left rail carrying the source-language filter (العربية / English / both).
  • Wide right column of numbered findings with inline source chips indicating source family and language.
  • Asymmetric editorial grid with a faintly visible 12-column baseline grid behind the content.
  • In-place re-sorting and re-weighting of the findings list when the language filter changes.
• States:
  • Loading: the findings list and the language filter rail render; the filter defaults to showing both languages.
  • Empty: if a selected language filter yields no findings for a subject area, the surface states plainly that no findings were gathered for that language in that area, and the reader can switch the filter back to both languages.
  • Success: the reader can read findings in either language or both, and can see which language each finding came from.
  • Error/recovery: if the filter interaction fails, the full unfiltered findings list remains readable.
  • Reduced motion: filter changes apply without staggered reveals; the list re-sorts instantly and remains fully legible.

Resource Audit

• Information and state: a dense instrument table of Windows services and features that consume system resources during game launch on Windows 11 25H2, with columns for service/feature name, what it does, RAM at idle, RAM during game launch, iGPU shared-memory impact, and a verdict field (Keep / Tune / Disable with caution).
• Primary action: inspect a service or feature row and expand it in place.
• Supporting actions: read the verdict for each row; identify rows carrying risk; move from an audited service to its corresponding step in the Optimization Plan.
• Domain entities: Windows service, Windows feature, service description, RAM at idle, RAM during game launch, iGPU shared-memory impact, verdict (Keep / Tune / Disable with caution), risk marker.
• Component responsibilities:
  • Dense instrument table with hairline row borders and tabular numerals on every RAM figure and service count.
  • In-place row expansion revealing RAM-at-idle, RAM-during-launch, iGPU shared-memory impact, and the verdict.
  • Lime left-edge sweep on row hover.
  • Vermilion risk marker on any row that carries risk, including "do not disable" cautions.
  • Faintly visible 12-column baseline grid behind the table.
• States:
  • Loading: the table renders with its column headers and rows; expanded detail loads in place.
  • Empty: if no services or features are audited for a given category, the surface states that plainly rather than showing an empty table body.
  • Success: the reader can see, per service or feature, what it does, its RAM at idle, its RAM during game launch, its iGPU shared-memory impact, and its verdict.
  • Error/recovery: if a row's expanded detail cannot load, the row's summary values and verdict remain visible and the row can be collapsed and retried.
  • Reduced motion: the hover sweep and any row animation are suppressed; rows remain fully legible and expandable.

Optimization Plan

• Information and state: the comprehensive, readable, step-by-step plan for achieving maximum Windows 11 25H2 gaming performance while launching a game, organized as a numbered vertical sequence of phases, with each phase carrying its concrete actions and its expected effect on RAM, services, features, and integrated-GPU shared memory.
• Primary action: follow the numbered phases in order.
• Supporting actions: track reading progress along the plan; move from a phase back to the audited service or finding it derives from.
• Domain entities: phase, phase number, step, expected effect (RAM, services, features, iGPU shared memory), target platform (Windows 11 25H2), progress position.
• Component responsibilities:
  • Numbered vertical sequence of phases.
  • Fixed progress spine on the left that fills as the reader scrolls, mirroring the numbered phases.
  • RTL/LTR support: the spine flips side and the grid mirrors, while numerals stay tabular.
  • Per-phase expected-effect statements tied to the audited services and findings.
• States:
  • Loading: the phase sequence and the progress spine render from the top of the plan.
  • Empty: not applicable — the plan always carries its accepted numbered phases.
  • Success: the reader can follow the plan as one continuous procedure from first phase to last.
  • Error/recovery: if the progress spine cannot track scroll, the numbered phases remain fully readable and ordered.
  • Reduced motion: the spine shows a static position indicator instead of a scroll-filling animation; phases remain fully legible.

3. Functional Requirements

Each requirement below is a distinct story point with its provenance, lifecycle facts, and observable acceptance.

FR-1 — Deep research across internet, YouTube, and other sources in Arabic and English As a Windows 11 Gaming Optimizer, I should be able to read deep research on improving gaming performance on Windows 11 25H2 that was gathered from internet articles and guides, from YouTube content, and from other sources, in both Arabic and English, so that I can draw on a broad, bilingual evidence base rather than a single-language or single-format set of tips.

• Provenance: explicit.
• Actor: Windows 11 Gaming Optimizer. Indispensable participant: Research Report Reader, who consumes the same findings.
• Trigger/input: the reader opens Research Findings.
• Observable result: numbered findings are presented, each attributed to its source family (internet article/guide, YouTube, other) and its source language (Arabic, English).
• Access state: publicly readable, no identity required.
• Failure/recovery: if a language filter yields no findings for an area, the surface states that plainly and the reader can return the filter to both languages.
• Continuation: the reader can move from a finding to the Resource Audit or the Optimization Plan.
• Acceptance: findings from both Arabic and English sources are present and each finding shows its source family and language.

FR-2 — Focus the research on reducing RAM consumption by the system and services during game launch As a Windows 11 Gaming Optimizer, I should be able to read research specifically focused on reducing RAM consumption by the Windows system and its services while a game is launching and running, so that I can free memory for the game itself.

• Provenance: explicit.
• Actor: Windows 11 Gaming Optimizer. Indispensable participant: Research Report Reader.
• Trigger/input: the reader opens Research Findings and the Resource Audit.
• Observable result: findings and audited rows carry RAM-at-idle and RAM-during-game-launch figures for the system and its services.
• Access state: publicly readable, no identity required.
• Failure/recovery: if a RAM figure is unavailable for a service, the row states that plainly rather than showing a fabricated value.
• Continuation: the reader can carry the RAM-reduction guidance into the Optimization Plan.
• Acceptance: RAM consumption by the system and services during game launch is addressed explicitly in the findings and the audit.

FR-3 — Cover the integrated-GPU (iGPU) case As a Windows 11 Gaming Optimizer running an integrated GPU, I should be able to read guidance that specifically addresses integrated-GPU systems, so that the plan applies to my machine rather than assuming a discrete graphics card.

• Provenance: explicit.
• Actor: Windows 11 Gaming Optimizer. Indispensable participant: Research Report Reader.
• Trigger/input: the reader opens Research Findings, the Resource Audit, and the Optimization Plan.
• Observable result: iGPU shared-memory impact is a column in the Resource Audit, iGPU considerations appear among the findings, and the plan's phases state their integrated-GPU effect.
• Access state: publicly readable, no identity required.
• Failure/recovery: where an action's iGPU effect is not established, the surface says so rather than implying a benefit.
• Continuation: the reader applies the iGPU-specific guidance on their own machine.
• Acceptance: integrated-GPU systems are addressed in findings, audit, and plan.

FR-4 — Identify services and features that consume device resources while games are running As a Windows 11 Gaming Optimizer, I should be able to see which Windows services and features consume device resources while games are running, so that I can decide what to keep, tune, or disable with caution.

• Provenance: explicit.
• Actor: Windows 11 Gaming Optimizer. Indispensable participant: Research Report Reader.
• Trigger/input: the reader opens the Resource Audit.
• Observable result: a dense instrument table lists each service or feature with what it does, RAM at idle, RAM during game launch, iGPU shared-memory impact, and a verdict of Keep / Tune / Disable with caution.
• Access state: publicly readable, no identity required.
• Failure/recovery: rows carrying risk are marked in vermilion, including "do not disable" cautions, so the reader is warned before acting.
• Continuation: the reader expands a row in place and then moves to the corresponding plan phase.
• Acceptance: every audited service or feature shows its resource consumption and its verdict, and risky rows are visibly marked.

FR-5 — Produce a comprehensive report containing a complete plan for maximum performance while launching a game As a Research Report Reader, I should be able to read a comprehensive report that contains a complete, ordered plan for achieving maximum system performance while launching a game on Windows 11 25H2, so that I can follow one continuous procedure instead of assembling tips myself.

• Provenance: explicit.
• Actor: Research Report Reader. Indispensable participant: Windows 11 Gaming Optimizer, who applies the plan.
• Trigger/input: the reader opens the Optimization Plan, optionally from the Landing's primary action.
• Observable result: a numbered vertical sequence of phases with a progress spine that fills as the reader scrolls, each phase carrying its concrete actions and expected effect.
• Access state: publicly readable, no identity required.
• Failure/recovery: if the progress spine cannot track scroll, the numbered phases remain fully readable and ordered.
• Continuation: the reader applies the phases in order on their own Windows 11 25H2 machine.
• Acceptance: the plan is complete, ordered, and readable as a single continuous procedure.

FR-6 — Bilingual source coverage is a reader-operated control As a Research Report Reader, I should be able to filter the findings by source language — العربية, English, or both — so that I can read in my preferred language or compare both.

• Provenance: required_inference — required to make the accepted bilingual-source requirement operable for the reader.
• Actor: Research Report Reader. Indispensable participant: Windows 11 Gaming Optimizer.
• Trigger/input: the reader operates the sticky language-filter rail on Research Findings.
• Observable result: the findings list re-sorts and re-weights in place, and each finding continues to show its source language.
• Access state: publicly readable, no identity required.
• Failure/recovery: if the filter interaction fails, the full unfiltered list remains readable.
• Continuation: the reader continues reading or switches the filter back to both languages.
• Acceptance: the filter changes the presented findings without hiding the fact that both languages were researched.

FR-7 — Target Windows 11 25H2 specifically As a Windows 11 Gaming Optimizer, I should be able to rely on the research, audit, and plan being specific to Windows 11 25H2, so that the guidance matches my operating system version.

• Provenance: required_inference — required to make the accepted platform-specific research executable.
• Actor: Windows 11 Gaming Optimizer. Indispensable participant: Research Report Reader.
• Trigger/input: the reader reads any surface.
• Observable result: Windows 11 25H2 is named as the target platform on the Landing and carried through the findings, the audit, and the plan.
• Access state: publicly readable, no identity required.
• Failure/recovery: where guidance is version-sensitive, the surface states the version it applies to rather than generalizing.
• Continuation: the reader applies version-appropriate guidance.
• Acceptance: Windows 11 25H2 is the stated target throughout.

FR-8 — Reach the report and the plan from a public entry As a Windows 11 Gaming Optimizer, I should be able to arrive at a public entry that explains the research scope and lets me begin the full plan or open the findings and the audit, so that I can start without an account.

• Provenance: required_inference — required to make the accepted public delivery reachable.
• Actor: Windows 11 Gaming Optimizer. Indispensable participant: Research Report Reader.
• Trigger/input: the reader opens the Landing.
• Observable result: the scope statement, the primary capsule CTA to begin the full plan, and secondary links to the findings and the audit are present and operable.
• Access state: anonymous, no identity required.
• Failure/recovery: if the hero's live bar-field cannot render, the headline, scope statement, ticker, metrics strip, and both entry actions remain fully legible and operable.
• Continuation: the reader enters the plan or the report.
• Acceptance: all three destinations are reachable from the Landing without signing in.

4. User Personas

Windows 11 Gaming Optimizer

• Product context: an Arabic-speaking PC gamer running Windows 11 25H2 on a machine that relies on an integrated GPU. They are technical and self-reliant, and they are skeptical of generic "boost your FPS" content. They watch Task Manager during a game launch and want to know exactly which services and features are taking RAM and device resources away from the game.
• Primary goal: reduce RAM and resource consumption by the Windows system and its services while a game is launching and running, and reach maximum gaming performance on their integrated-GPU machine.
• Distinct accepted responsibilities: researching tips across internet articles, YouTube, and other sources in both Arabic and English; focusing that research on RAM reduction during game launch; covering the integrated-GPU case; identifying the services and features that consume device resources while games run; and applying the complete plan on their own machine.
• Relevant inputs or decisions: which source language to read; which audited services to keep, tune, or disable with caution; which plan phases to apply and in what order.
• Interactions with other accepted participants: shares the same findings, audit, and plan with the Research Report Reader; the Optimizer is the participant who acts on the machine, while the Reader is the participant who consumes the compiled report.
• Observable success: the reader can see, per service, its RAM at idle and during game launch and its iGPU shared-memory impact, and can follow an ordered plan whose phases state their expected effect.

Research Report Reader

• Product context: a reader who consumes the comprehensive report and the complete optimization plan, using the compiled Arabic and English findings to apply the recommended steps on their own Windows 11 25H2 machine. They may not be the person who originally gathered the tips; they arrive to use the compiled result.
• Primary goal: read a comprehensive, trustworthy report and a complete plan, and apply the recommended steps on their own Windows 11 25H2 machine.
• Distinct accepted responsibilities: consuming the report and the plan; using the compiled Arabic and English findings; applying the recommended steps on their own machine.
• Relevant inputs or decisions: which language to read in; whether to compare Arabic and English findings; which phases to apply.
• Interactions with other accepted participants: relies on the same findings, audit, and plan as the Windows 11 Gaming Optimizer; the Reader's distinct work is comprehension and application rather than the Optimizer's machine-level tuning decisions.
• Observable success: the reader can read the plan as one continuous procedure, filter findings by language, and see each audited service's resource consumption and verdict.

5. Core User Flows

Flow A — Windows 11 Gaming Optimizer: research, audit, and apply the plan

1. The Optimizer opens the Landing anonymously. The kinetic bilingual headline names Windows 11 25H2 and the integrated-GPU maximum-performance subject; the live RAM bar-field runs behind and beside the type; the service ticker band and the three-column measured-metrics strip are visible; the scope statement names Windows 11 25H2, RAM, services, features, and integrated GPU.
2. The Optimizer reads the scope statement and confirms the research matches their machine — Windows 11 25H2, integrated GPU, RAM and service consumption during game launch.
3. The Optimizer selects the primary capsule CTA to begin the full plan, or a secondary link to open the findings or the audit.
4. On Research Findings, the Optimizer operates the sticky language-filter rail, choosing العربية, English, or both. The findings list re-sorts and re-weights in place, and each numbered finding keeps its inline source chip showing its source family and language.
5. The Optimizer reads findings on RAM reduction during game launch, on services and features that consume device resources during gameplay, and on integrated-GPU considerations.
6. The Optimizer moves to the Resource Audit. The dense instrument table lists each Windows service and feature with what it does, RAM at idle, RAM during game launch, iGPU shared-memory impact, and a verdict of Keep / Tune / Disable with caution.
7. The Optimizer hovers a row; a lime left-edge sweep marks it. The Optimizer expands the row in place to read its RAM-at-idle, RAM-during-launch, and iGPU shared-memory detail alongside its verdict.
8. The Optimizer notices a row carrying a vermilion risk marker and a "do not disable" caution, and decides to keep that service rather than disable it.
9. The Optimizer moves to the Optimization Plan. The numbered vertical sequence of phases is presented with the progress spine on the left.
10. The Optimizer scrolls; the spine fills with scroll progress, mirroring the numbered phases, and the plan reads as one continuous procedure.
11. The Optimizer reads each phase's concrete actions and its expected effect on RAM, services, features, and iGPU shared memory, and applies the phases in order on their own Windows 11 25H2 machine.
12. Failure/recovery: if the hero's live bar-field cannot render on the Landing, the headline, scope statement, ticker, metrics strip, and both entry actions remain fully legible and operable, and the Optimizer continues. If a Resource Audit row's expanded detail cannot load, the row's summary values and verdict remain visible and the row can be collapsed and retried. If the plan's progress spine cannot track scroll, the numbered phases remain fully readable and ordered.
13. Continuation: the Optimizer returns to the Resource Audit to re-check a service, or to Research Findings to read the other language's findings on the same subject.

Flow B — Research Report Reader: consume the report and the plan

1. The Reader opens the Landing anonymously and reads the one-line scope statement naming Windows 11 25H2, RAM, services, features, and integrated GPU.
2. The Reader opens Research Findings and sets the language filter to their preferred language. The list re-sorts in place and each finding shows its source language.
3. The Reader reads the numbered findings, noting which came from internet articles and guides and which from YouTube content, and in which language.
4. The Reader opens the Resource Audit and reads the instrument table, comparing RAM at idle against RAM during game launch for each service and feature, and reading each verdict.
5. The Reader expands a row in place to read its iGPU shared-memory impact, and notes a vermilion risk marker on a row that must not be disabled.
6. The Reader opens the Optimization Plan and reads the numbered phases as one continuous procedure, following the progress spine as it fills with scroll.
7. The Reader applies the recommended steps on their own Windows 11 25H2 machine, in the order the plan presents.
8. Failure/recovery: if a language filter yields no findings for an area, the surface states that plainly and the Reader switches the filter back to both languages. If the progress spine cannot track scroll, the numbered phases remain fully readable and ordered.
9. Continuation: the Reader returns to the Resource Audit to confirm a service's verdict before applying the corresponding phase.

Flow C — Windows 11 Gaming Optimizer: compare Arabic and English findings on one subject

1. The Optimizer opens Research Findings and sets the language filter to العربية, reading the Arabic-sourced findings on RAM reduction during game launch.
2. The Optimizer switches the filter to English and reads the English-sourced findings on the same subject.
3. The Optimizer switches the filter to both and compares the two sets side by side, using each finding's inline source chip to keep the languages distinct.
4. The Optimizer carries the combined guidance into the Resource Audit to check the specific services named, and then into the Optimization Plan to apply the corresponding phases.
5. Failure/recovery: if the filter interaction fails, the full unfiltered findings list remains readable and the Optimizer continues reading.
6. Continuation: the Optimizer returns to the plan to apply the next phase.

6. Visuals Colors and Theme

The creative direction is authoritative for this section. The muse is Yugo Nakamura; the headline idea is interaction as instrument — kinetic type and cursor-reactive systems that turn invisible system telemetry into something the reader can feel and manipulate. The surface must read as a live diagnostic playground, not a blog.

Mode: dark mode.

Colour tokens by role:

Colour is functional: lime means measured/active, vermilion means risk, grey means reference. No blue anywhere — no #2563EB, #4F46E5, #6366F1 or neighbours, even for links or focus rings; focus is lime. No gradients on text, no glass.

Typography:

• Headings (Latin): Space Grotesk, weights 500–700, tight tracking -0.02em to -0.04em, set large and often as full-width kinetic lines.
• Headings (Arabic): IBM Plex Sans Arabic at 700, with slightly looser tracking than Latin and never letter-spaced, since letter-spacing damages Arabic joining.
• Body: IBM Plex Sans Arabic.
• Type scale: 1.25 modular with a deliberate jump for display — 128 / 72 / 44 / 28 / 20 / 16 / 13.
• Display sizes use clamp(40px, 9vw, 128px) for hero lines and clamp(28px, 4.5vw, 56px) for section openers, so no headline is ever a single fixed desktop size.
• Body copy: 17px/1.75 for Latin, 18px/1.9 for Arabic (Arabic needs more leading).
• Micro-labels: uppercase Space Grotesk 500 at 11px with 0.16em tracking — HUD-style, always paired with a lime rule.
• Tabular numerals via font-variant-numeric: tabular-nums on every RAM figure, FPS value, and service count.

Shape language: hard-edged and procedural. No rounded cards, no pills, no soft shadows — rectangles with 1px hairline borders (#242830) and sharp corners. The one deliberate exception: live-value chips and the primary CTA are full capsules, because they behave like physical controls. Section boundaries are thin lime or vermilion rules that extend past the content column and get clipped by the viewport. A persistent 12-column baseline grid is faintly visible as vertical hairlines behind content on Research Findings and Resource Audit.

Spacing rhythm: full-viewport playground sections rather than stacked cards; a consistent 12-column grid with a sticky left rail on Research Findings and a fixed progress spine on Optimization Plan; generous vertical rhythm between phases so the plan reads as one continuous procedure.

Imagery style: generative and procedural rather than photographic. The hero's visual is a canvas of thin vertical bars whose heights are driven by a plausible RAM-usage series — it looks like a system monitor, because it is one. Section openers use simple geometric diagrams: a bar chart of RAM before/after, a shared-memory split diagram for iGPU, a service-dependency line graph. No stock photos of gaming setups, no RGB-lit PC towers, no 3D renders. Where a screenshot is genuinely needed (Task Manager, Services.msc, Windows settings panels), it is shown as a flat, high-contrast capture inside a hairline frame with a lime callout rule pointing at the relevant row.

RTL/LTR: both supported — the spine flips side, the grid mirrors, numerals stay tabular.

7. Signature Design Concept

The public entry — the Landing — is a full-viewport dark instrument panel, not a centered SaaS hero.

The dominant element is a single kinetic headline in Space Grotesk that spans the viewport width: «ويندوز 11 25H2» set at clamp(48px, 11vw, 150px) in near-white #F2F4F6, with the second line «أقصى أداء للألعاب على كرت مدمج» sliding in beneath it in acid lime #C6FF3D at roughly 60% of that size. Behind and to the right of the type, off the left edge of the text column, a live bar-field canvas animates continuously: thin vertical bars, 2px wide, heights driven by a RAM series, in lime at 30% opacity with the tallest bars at full opacity. A single hairline rule runs the full viewport width beneath the headline, and pinned directly under its left end sits the primary capsule CTA in lime with black text — «ابدأ الخطة الكاملة» — plus a secondary text link in muted grey #7A828C.

The composition is deliberately asymmetric: type occupies the left 7 of 12 columns, and the live bar-field bleeds off the right and bottom edges. There is no centered stack, no gradient blob, and no subheadline paragraph longer than one line. The page's ornament is literally a system monitor — the same instrument the reader watches during a game launch.

This concept recomposes only accepted content, states, and controls: the Windows 11 25H2 subject, the integrated-GPU maximum-performance subject, the RAM focus, the entry actions to the plan, the findings, and the audit. It introduces no new behaviour, page, or destination.

8. Interaction Model & Motion Direction

Interaction Model: Animated Motion Tempo: cinematic Hero Dimensionality: layered_2d

Landing Hero Motion Brief

• Focal subject: the kinetic bilingual headline — «ويندوز 11 25H2» in near-white over «أقصى أداء للألعاب على كرت مدمج» in acid lime — with the live RAM bar-field canvas behind and beside it.
• Input → transformation → outcome thesis: as the reader's pointer moves near the headline, the headline lines shift a few pixels and shift weight, and the live bar-field's bars respond in amplitude — so the reader's own movement over the type produces a visible, measured response in the instrument behind it. The outcome is that the reader feels the page is a live diagnostic surface reading system behaviour, not a static article.
• Motion vocabulary: cursor-reactive proximity response on the headline; a continuously animating RAM bar-field at low amplitude; scroll-triggered staggered word-by-word reveals on section openers (60ms stagger, 400ms each, ease-out); a lime left-edge sweep on service table row hover; a scroll-filling progress spine on the Optimization Plan.
• Composed first frame: the headline lines are already in place at their composed positions, the bar-field holds a plausible RAM series with the tallest bars at full opacity, the hairline rule runs the full viewport width, and the lime capsule CTA sits under its left end. Nothing is mid-animation on first paint.
• Reduced-motion state: all motion collapses to static, fully legible states — the bar-field holds a single composed frame with no continuous loop, headline reveals are not staggered, the row hover sweep is suppressed, and the plan's spine shows a static position indicator instead of a scroll-filling animation.

Landing Hero 3D Scene Brief — DIRECTION-DERIVED

The direction's hero dimensionality is layered_2d, not webgl, so no Canvas/R3F/Drei scene is required. The hero's depth is composed in layered 2D: the bar-field canvas sits behind the type, the type sits above it, and the hairline rule and CTA sit above the type — three flat layers that read as an instrument panel. A generated 3D hero is permitted but not required, and the layered 2D composition is the intended realization.

9. Non-Functional Requirements

• NFR-1 — Bilingual Arabic and English presentation. The product must present its findings, audit, and plan in both Arabic and English, with correct RTL and LTR layout support: the spine flips side, the grid mirrors, and numerals stay tabular. Provenance: explicit (research must cover both Arabic and English sources). Rationale: the accepted research is bilingual, so the surface must render both languages correctly.
• NFR-2 — Arabic typography integrity. Arabic text must never receive letter-spacing, and Arabic headings must use IBM Plex Sans Arabic at 700 with slightly looser tracking than Latin. Provenance: creative direction. Rationale: letter-spacing damages Arabic joining.
• NFR-3 — Numeric legibility. Every RAM figure, FPS value, and service count must use tabular numerals. Provenance: creative direction. Rationale: the product is a data-driven report and the numbers must align for comparison.
• NFR-4 — Readable text and controls stay whole. Headlines, wordmarks, labels, numbers, cards' text, and controls must stay entirely inside the viewport and their container at 375px, 768px, and 1280px, wrapping or scaling (for example font-size: clamp(...) with its mobile value) to fit, and no other element may cover any part of them. Provenance: creative direction. Rationale: the report is dense and must remain readable at every viewport.
• NFR-5 — Reduced-motion compliance. Under prefers-reduced-motion, all motion must collapse to static, fully legible states with no staggered reveals and no continuous loops; moving and scrollable content must stop and show whole items, wrapping into rows or sitting in a horizontally scrollable row (overflow-x: auto). Provenance: creative direction. Rationale: accessibility and legibility.
• NFR-6 — No blue in the palette. No blue or indigo may appear anywhere, including links and focus rings; focus is lime. Provenance: creative direction. Rationale: the palette is functional — lime means measured/active, vermilion means risk, grey means reference.
• NFR-7 — Public readability without identity. All four surfaces must be readable without an account, sign-in, or identity continuity. Provenance: explicit via the planning boundary's access_requirement: none on every surface. Rationale: the accepted scope has no private per-user state, commitment, entitlement, or value transfer that must remain bound to a participant.
• NFR-8 — Windows 11 25H2 specificity. The product must name and target Windows 11 25H2 specifically, and must state the version where guidance is version-sensitive. Provenance: explicit. Rationale: the accepted research targets that version.
• NFR-9 — No fabricated measurements. Where a RAM figure, iGPU shared-memory impact, or expected effect is not established, the surface must state that plainly rather than presenting a fabricated value. Provenance: required_inference. Rationale: the audience is skeptical of generic performance content and the product's credibility depends on measured, attributed evidence.

10. Tech Stack

• Frontend: React, with RTL/LTR support for Arabic and English.
• Backend: Python / FastAPI, serving the research findings, the resource audit, and the optimization plan content.
• Storage: appropriate storage for the report content — findings, audited services and features with their RAM-at-idle, RAM-during-launch, iGPU shared-memory impact, and verdicts, and the ordered plan phases.
• Containerization: Docker / docker-compose for local and deployment packaging.
• Kubernetes: only if deployment requires it.

No source-specified technology choices beyond the above were stated by the user; the remaining stack items are the minimal set needed to deliver the accepted custom UI and generated documents.

11. Assumptions and Constraints

Constraints (binding):

• Research must cover both Arabic and English sources. explicit
• Research must target Windows 11 25H2 specifically. explicit
• Optimization focus must include reducing RAM consumption by the system and services during game launch. explicit
• Must address integrated GPU (iGPU) systems. explicit
• All four surfaces are publicly readable with no identity requirement. explicit via the planning boundary.
• No blue or indigo in the palette; focus is lime. Creative direction.
• Arabic text must never be letter-spaced. Creative direction.
• Readable text and controls must stay whole at 375px, 768px, and 1280px. Creative direction.

Assumptions (narrow and labeled):

• The product is a read-and-apply deliverable: it does not modify the reader's Windows installation, install software, run as a background service, or connect to the reader's machine. [Assumption — derived from the accepted scope, which is research and a plan, not a tuning utility]
• The reader performs every optimization action on their own Windows 11 25H2 device, outside this product. [Assumption — derived from the accepted scope]
• The verified source content is the two declared source families (internet articles and guides; YouTube content) in Arabic and English; specific article titles, video titles, channel names, URLs, publication dates, and per-source numeric measurements were not verified by the directives and are not invented. [Assumption — bounded by the reference directives]
• Where a RAM figure or iGPU impact is not established by the gathered sources, the surface states that plainly rather than presenting a fabricated value. [Assumption — required for the product's credibility with a skeptical technical audience]

Out of scope (current):

• General Windows tuning unrelated to gaming performance.
• Hardware purchasing advice, overclocking, and non-Windows platforms.
• Any capability that modifies the reader's operating system, games, drivers, or third-party tuning tools.

12. Glossary

• Windows 11 25H2: the specific Windows 11 release targeted by the research, audit, and plan.
• RAM: random-access memory; the resource whose consumption by the system and services during game launch is the product's central focus.
• RAM at idle: a service's or feature's memory consumption when no game is running.
• RAM during game launch: a service's or feature's memory consumption while a game is launching and running.
• iGPU (integrated GPU): a graphics processor integrated into the CPU, which shares system memory rather than having dedicated video memory.
• iGPU shared-memory impact: the effect a service or feature has on the memory shared with the integrated GPU.
• Service: a Windows background service that consumes device resources while games are running.
• Feature: a Windows feature that consumes device resources while games are running.
• Verdict: the audit's recommendation for a service or feature — Keep, Tune, or Disable with caution.
• Resource Audit: the surface presenting Windows services and features that consume system resources during game launch, with RAM and iGPU columns and a verdict field.
• Research Findings: the surface organizing deep gaming-performance findings gathered from internet, YouTube, and Arabic and English sources for Windows 11 25H2.
• Optimization Plan: the surface providing the comprehensive, readable, step-by-step plan for achieving maximum Windows 11 25H2 gaming performance while launching a game.
• Progress spine: the fixed element on the Optimization Plan that fills as the reader scrolls, mirroring the numbered phases.
• Tabular numerals: numerals of equal width, used for every RAM figure, FPS value, and service count so values align for comparison.