Alternatives to Core42 — Saudi Compute

Compare alternatives to Core42 in the Saudi compute landscape. 8 ranked alternatives by SCS, sector fit, and execution status.

When you’d compare alternatives to Core42

Core42 represents Abu Dhabi’s most direct bet on sovereign GPU cloud infrastructure, and understanding it requires recognizing that the UAE and Saudi Arabia are now engaged in the GCC’s defining infrastructure competition of the 2020s. For Saudi compute strategists, Core42 is not just a vendor option—it is the benchmark against which HUMAIN, the SDAIA cloud program, and every PIF-backed compute investment is measured. When procurement teams, sovereign fund analysts, and AI program managers in Riyadh begin evaluating their options, Core42 surfaces immediately as both a cautionary tale and a capability target.

The comparison to alternatives arises in several specific contexts. Saudi enterprises evaluating where to run large AI workloads in the GCC must decide whether to use UAE-based infrastructure like Core42 or wait for Saudi-sovereign alternatives to reach operational scale. The latency between Abu Dhabi and Riyadh—roughly 8–12ms for direct fiber paths—is acceptable for most inference workloads but not for real-time AI applications that need sub-5ms round trips. This geographic constraint makes Core42 a genuine operational choice for some Saudi organizations, not merely a theoretical competitor.

The funding and governance context matters enormously. Core42 operates under G42, Abu Dhabi’s technology conglomerate, with backing from Mubadala Investment Company—the UAE sovereign wealth fund. The connection to Mubadala is not incidental: it means Core42 has access to sovereign capital at a scale and cost that commercial cloud operators cannot match, and it means Core42’s strategic decisions are aligned with Abu Dhabi’s geopolitical objectives, not just commercial returns. Saudi buyers working with Core42 are, in an indirect but meaningful sense, paying rent to a competitor sovereign’s technology infrastructure. This is not always a dealbreaker, but it is a strategic consideration that Saudi planners should make explicit.

The NVIDIA relationship is central to Core42’s current market position. Core42 operates one of the largest H100 GPU clusters in the GCC region, enabled by NVIDIA’s early commitment to Abu Dhabi as a distribution channel for AI compute into the Middle East. This silicon access advantage is real but not permanent—Saudi Arabia’s own procurement relationships with NVIDIA, AMD, and emerging alternatives are rapidly closing the gap. The critical question for any Saudi buyer is whether the timing advantage Core42 offers today is worth the strategic dependency it creates over the multi-year horizon of major AI programs. Core42’s silicon posture is also more diversified than a pure NVIDIA story: it is a major Cerebras customer, anchoring the wafer-scale vendor’s largest regional deployment, which gives Abu Dhabi a second-source accelerator relationship that Saudi Arabia’s program—anchored on NVIDIA, AMD, Qualcomm, and Groq—has pursued through different vendors.

The G42 governance dimension adds a further layer of complexity. G42 has navigated public scrutiny over its relationships with Chinese technology companies, which led to a strategic realignment and investment by Microsoft. For Saudi buyers evaluating Core42 specifically from a US regulatory compliance perspective, the G42-Microsoft relationship provides some assurance of Western technology governance standards, but it also introduces Microsoft’s own commercial interests into the governance equation.

How to read the alternative rankings

Core42 scores 7.1 on the Saudi Compute Score, with its GCC Competitors sector classification reflecting that it is evaluated from a Saudi strategic perspective rather than as a neutral infrastructure option. The three alternatives—Mubadala at 8.3, EDGE at 7.8, and UAE AI Office at 7.8—all score higher than Core42 itself, which might initially seem counterintuitive since all four are UAE-linked entities. The distinction lies in what each entity represents from Saudi Arabia’s strategic standpoint.

Capacity (18%) weights heavily in Core42’s evaluation because GPU cloud infrastructure is ultimately a capacity question: how many H100s can be accessed, at what utilization rate, with what reservation horizon? Core42’s cluster, while substantial, is finite and subject to competing demand from UAE government programs, international research institutions, and commercial customers. Saudi buyers are not Core42’s priority customer segment—that priority belongs to Abu Dhabi’s own AI programs first and international strategic partners second.

Capital (16%) reflects Core42’s access to Mubadala-backed financing, which gives it a strong balance sheet for infrastructure expansion. However, the capital ultimately serves UAE objectives, not Saudi ones. When Saudi Arabia’s own sovereign capital—through PIF, SDAIA, or direct ministry budgets—is available for comparable investment, the Capital dimension works in favor of Saudi alternatives that deploy capital domestically.

Silicon Access (16%) is where Core42 holds a genuine current advantage. Its established NVIDIA relationship and existing H100 cluster means that Saudi organizations with immediate GPU requirements can access compute today without waiting for Saudi-sovereign procurement cycles to complete. This advantage is eroding as Saudi Arabia secures its own NVIDIA allocations through HUMAIN and bilateral government agreements, but it remains real in the current 6–18 month window.

Sovereignty (13%) is where Core42 loses significant ground from a Saudi perspective. Running sensitive Saudi AI workloads on UAE sovereign infrastructure creates jurisdictional dependencies that Saudi data residency regulations may not permit and that Saudi strategic planners should avoid for anything touching government systems, defense-adjacent applications, or Aramco operational data.

Geopolitical Resilience (13%) for Core42 reflects UAE’s political stability and its robust data center power and connectivity infrastructure, but it scores below Saudi alternatives because Abu Dhabi’s strategic environment involves its own geopolitical dynamics—including UAE-Iran tensions and the complex post-Abraham Accords regional architecture—that Saudi AI planners should factor into long-term dependency analysis.

Velocity (12%) and Execution (12%) are Core42’s stronger suits—it is an operational facility with real GPU capacity available now, which scores well on both dimensions compared to Saudi programs still in buildout phases.

When the alternatives become preferable

Saudi sovereign AI programs require in-kingdom compute. SDAIA’s national AI strategy explicitly prioritizes compute infrastructure within Saudi Arabia’s borders. Any workload tied to Vision 2030 government programs, NEOM’s digital twin infrastructure, or Saudi smart city initiatives cannot be run on Core42’s Abu Dhabi infrastructure without creating compliance exposure under Saudi data governance frameworks. For these workloads, Saudi-domestic alternatives are required regardless of Core42’s performance advantages.
Mubadala’s Saudi investment partnerships create more aligned incentives. Mubadala scores 8.3 SCS because it operates both as UAE sovereign fund and as an active co-investor in Saudi infrastructure through PIF partnership arrangements. When Mubadala and PIF co-invest in a Saudi compute facility, the sovereign alignment is fundamentally different from Core42 simply offering cloud services to Saudi customers. The Mubadala alternative reflects Saudi-UAE partnership rather than Saudi dependence on UAE infrastructure.
Long-term capacity expansion favors Saudi-sovereign build. Core42’s expansion capacity is constrained by Abu Dhabi’s data center land availability, UAE power infrastructure, and Abu Dhabi government funding priorities for its own AI programs. Saudi Arabia’s planned 1.5 GW of AI compute capacity across multiple campuses dwarfs what Core42 can realistically offer to Saudi buyers specifically. As Saudi build programs reach operational scale—expected in the 2026–2028 window for major facilities—the case for using Core42 capacity weakens proportionally.
EDGE’s defense and security specialization fills a gap Core42 cannot. For workloads at the intersection of AI and defense—which is a significant category given Saudi Arabia’s Vision 2030 security industrialization programs—EDGE Group’s AI capabilities and defense-sector specialization make it a more appropriate partner than Core42’s commercial cloud model. EDGE’s SCS 7.8 reflects its value for this specific niche even though its general-purpose GPU cloud capability is smaller than Core42’s cluster.
UAE AI Office policy alignment creates regulatory architecture Core42 doesn’t provide. The UAE AI Office’s SCS 7.8 reflects its ability to shape the operating environment for AI across the UAE and, through GCC coordination mechanisms, to influence Saudi AI policy as well. Organizations deploying AI models across both countries benefit from understanding and engaging with UAE AI Office frameworks—something that procuring Core42 cloud services does not provide.

The competitive tier breakdown

Mubadala Investment Company (SCS 8.3)

Mubadala is the highest-scoring alternative to Core42 because it operates at a fundamentally different level of the infrastructure stack. Core42 is a GPU cloud operator; Mubadala is the sovereign capital vehicle that has co-invested in data centers, semiconductor companies, AI startups, and infrastructure projects across multiple continents. When Saudi Arabia’s PIF and Mubadala co-invest in GCC AI infrastructure, the result is not a UAE-versus-Saudi competition but a sovereign wealth fund alliance that pools capital for mutual benefit.

Mubadala’s portfolio includes investments in GlobalFoundries, which provides chipmaking capacity relevant to semiconductor supply chain diversification. It has direct relationships with NVIDIA, Arm Holdings, and multiple AI frontier labs through its venture portfolio. Its infrastructure arm has developed data center campuses in Abu Dhabi and has actively explored Saudi co-investment opportunities through its partnership frameworks with PIF. The SCS 8.3 reflects this breadth—Mubadala scores well across Capacity, Capital, Silicon Access, and Sovereignty because it is not just a compute provider but a capital ecosystem that can be structured to serve Saudi interests.

The practical implication for Saudi AI planners is that engaging Mubadala at the investment partnership level, rather than procuring Core42 services at the vendor level, changes the entire strategic relationship. Saudi organizations that can structure co-investment arrangements through PIF gain infrastructure assets and governance participation rather than cloud service dependency. This is the key analytical insight: Core42 is a vendor relationship; Mubadala is a partner relationship. The SCS difference reflects this structural distinction.

EDGE Group (SCS 7.8)

EDGE Group, Abu Dhabi’s defense technology conglomerate, occupies a distinct niche. Its AI capabilities are specifically calibrated for defense, border security, surveillance, and autonomous systems applications—exactly the areas where Core42’s commercial cloud model is least appropriate and where Saudi Arabia’s own Vision 2030 security industrialization programs are most active.

EDGE’s SCS 7.8 reflects strong Sovereignty scores for defense-grade security architecture, strong Execution scores from established programs with sovereign clients, and moderate Capacity scores reflecting its specialized rather than general-purpose compute footprint. For Saudi defense-adjacent AI applications—computer vision for border security, AI-assisted threat detection, autonomous logistics—EDGE is a more appropriate partner than Core42 even if its GPU cluster is smaller in absolute terms.

The EDGE-Saudi relationship is also supported by established defense cooperation frameworks between UAE and Saudi Arabia that provide government-to-government legal structures for sensitive technology transfer. These bilateral frameworks reduce the jurisdictional uncertainty that commercial Core42 engagements create for classified or sensitive workloads.

UAE AI Office (SCS 7.8)

The UAE Artificial Intelligence Office represents the regulatory and policy layer above both Core42 and EDGE. Its SCS 7.8 reflects its ability to shape the operating environment for AI across the UAE and, through GCC coordination mechanisms, to influence how Saudi AI policy develops. Organizations deploying AI models across both countries benefit from understanding UAE AI Office frameworks, which provide regulatory clarity and cross-border deployment guidance that Core42 as a compute provider cannot offer.

The rest of the field

Qatar Investment Authority (SCS 7.8) is the GCC’s third sovereign capital pole. With $475 billion in assets under management as of 2024—the seventh-largest sovereign wealth fund globally—QIA holds technology positions adjacent to the AI buildout, including semiconductor supply chain and global cloud investments, without direct Saudi-domiciled compute commitments. Since the 2021 Al-Ula Declaration normalized Qatar-Saudi relations, QIA has functioned as a potential co-investor in GCC compute rather than an infrastructure operator; it is an alternative to Core42 only at the capital layer, never the hosting layer.

Stargate (SCS 7.2) reframes the entire comparison at global scale. The OpenAI-SoftBank-Oracle joint venture has committed $500 billion to US AI infrastructure with an initial $100 billion phase underway, and its UAE-anchored regional partnerships give Abu Dhabi a connection into the largest compute program ever announced. For Saudi planners, Stargate’s relevance to the Core42 question is competitive context: the UAE’s compute ecosystem is being woven into US mega-infrastructure in ways that compound Abu Dhabi’s first-mover advantages.

MGX (SCS 7.2) is the investment vehicle layer of the same Abu Dhabi architecture that Core42 occupies at the operating layer. With reported assets exceeding $30 billion and an AI-specific mandate, MGX co-invests in Stargate and global AI infrastructure deals. Mapping MGX alongside Mubadala, EDGE, and Core42 reveals the UAE’s coordinated stack: sovereign fund, focused AI vehicle, defense conglomerate, and GPU cloud operator, each reinforcing the others.

Bahrain Cloud (SCS 7.1) ties Core42 on the composite and offers the alternative form of GCC compute maturity: multi-hyperscaler cloud regions—AWS operating in Bahrain since 2019, plus Azure and Google Cloud coverage—rather than a single sovereign GPU cloud. For Saudi enterprises, the Bahrain-versus-Core42 choice is a choice between hyperscaler service depth on a small neighboring island and sovereign-scale GPU density in Abu Dhabi; both carry the same fundamental limitation of hosting Saudi workloads under another jurisdiction.

Oman Compute (SCS 6.1) is the emerging entrant, building sovereign compute at a scale well below the Saudi and UAE programs. Its significance is directional—every GCC state now treats compute as sovereign infrastructure—rather than practical for near-term Saudi workload placement.

Core42’s structural position

Core42 holds SCS 7.1 because it is currently the most operationally ready GPU cloud in the GCC with a substantial H100 cluster accessible to non-US organizations. This is a real and meaningful advantage for Saudi buyers with immediate compute needs that cannot wait for Saudi domestic build programs. The structural weakness is that Core42’s strategic alignment runs to Abu Dhabi’s interests, not Saudi Arabia’s, and that Saudi sovereign compute alternatives are being built at a scale that will eventually make the Core42 capacity advantage irrelevant.

The counter-build is specific and funded: Humain’s Riyadh campus is under construction with 18,000 NVIDIA GB300 GPUs targeting 2026 and a 600,000-unit supply trajectory over three years, SDAIA’s sovereign AI factory is deploying 5,000 Blackwell GPUs, the Hexagon facility adds 480 MW of government capacity, and DataVolt’s 1.5 GW Oxagon project targets 2028. Each milestone that lands on schedule narrows the window in which Core42’s operational readiness commands a premium.

The window where Core42 is genuinely useful for Saudi AI programs is approximately 2024–2027, while Saudi domestic compute reaches full operational scale. After that window, Core42 transitions from a capacity provider to a competitor in the GCC AI market. Saudi organizations using it as a primary compute path should plan now for the migration to Saudi-sovereign infrastructure rather than discovering the need reactively when compliance requirements or capacity preferences shift.