Skip to main content
Non-Oil GDP Share: 55% 2025 real GDP |Saudi Unemployment: 7.2% Q4 2025 |PIF AUM: $1.21T 2025 actual |FDI Share of GDP: 2.8% Q1 2026 |Female Participation: 33.9% Q1 2026 |Credit Rating: Aa3/A+/A+ Moody's/Fitch/S&P |GDP Growth: 4.5% 2025 actual |Umrah Pilgrims: 18M+ 2025 foreign |Non-Oil GDP Share: 55% 2025 real GDP |Saudi Unemployment: 7.2% Q4 2025 |PIF AUM: $1.21T 2025 actual |FDI Share of GDP: 2.8% Q1 2026 |Female Participation: 33.9% Q1 2026 |Credit Rating: Aa3/A+/A+ Moody's/Fitch/S&P |GDP Growth: 4.5% 2025 actual |Umrah Pilgrims: 18M+ 2025 foreign |
Home Analysis & Editorial Alat Was Going to Build Saudi Arabia's Electronics Industry. It Dropped Semiconductors and Lost Its CEO.
Layer 2 editorial

Alat Was Going to Build Saudi Arabia's Electronics Industry. It Dropped Semiconductors and Lost Its CEO.

Alat, the PIF company launched in February 2024 with a $100bn budget to make Saudi Arabia a global electronics hub, dropped its semiconductor plans and removed its founding CEO in 2026. A venture-by-venture delivery ledger of what was announced, what was actually signed, and what exists on 31 July 2026.

Donovan Vanderbilt · · 20 min read
Alat Was Going to Build Saudi Arabia's Electronics Industry. It Dropped Semiconductors and Lost Its CEO. — Analysis — Saudi Vision 2030

Alat dropped its semiconductor business. The Public Investment Fund company launched on 1 February 2024 with a $100 billion budget to make Saudi Arabia a global centre for electronics and advanced industries removed founding chief executive Amit Midha in early 2026, reassigned its semiconductor team and abandoned plans to invest in chip production. Semafor reported the decision on 8 April 2026, citing people familiar with the matter, and added the detail that matters most: “several attempts to invest in chipmakers and secure agreements to build chip plants in the kingdom failed” [S1]. Alat confirmed the leadership change and said its “operations, partnerships, and strategic direction remain firmly in place.”

The company’s own website has since made the retreat structural. On 17 May 2026 alat.com listed nine business units, one of them Semiconductors [S2]. Checked on 31 July 2026, the site lists four sectors — AI & Digital Hardware, Building & Heavy Equipment, Electrification, and Home & Medical Equipment — and the address alat.com/en/our-business/semiconductors/ returns a 301 permanent redirect into the AI & Digital Hardware page [S3]. That page’s copy does not contain the word semiconductor. The unit that Alat once described as “responsible for development and manufacturing of semiconductors in three key technology segments: Power, Perception, and Processing” no longer exists as a named part of the company [S4].

What has not collapsed is the rest of it. Two Alat ventures are physically under construction, one of them producing. The gap between those two facts — a chip programme erased from the corporate structure, an elevator factory breaking ground in Dammam on 10 June 2026 — is the subject of this page. For Alat’s full company profile, ownership and business model, see our Alat company profile; this piece audits delivery against the announcements, venture by venture.

Last verified: 31 July 2026.

Announced at launchReality on 31 July 2026
$100bn deployment budget to 2030No cumulative deployment figure ever published; largest disclosed commitment is a $2bn convertible bond
Seven business units, later nineFour sectors; Semiconductors URL 301-redirects into AI & Digital Hardware
39,000 direct jobs by 2030No progress figure published; 28 Saudi engineers named in the most recent Lenovo update
$9.3bn non-oil GDP contribution by 2030No progress figure published
Global CEO Amit MidhaActing CEO Dr Muhammad Nasser Aldawood, unreplaced for roughly four months

What Was Alat Supposed to Build?

Alat was launched by Crown Prince Mohammed bin Salman, who chairs its board, on 1 February 2024 with a mandate to create a global champion in electronics and advanced industrial segments, powered by clean energy. The announced budget was $100bn (SAR375bn) to 2030. The announced outcomes were 39,000 direct jobs and a direct non-oil GDP contribution of $9.3bn, both by 2030 [S5].

The structure at launch was seven business units: advanced industrials, semiconductors, smart devices, smart buildings, smart appliances, smart health, and next-generation infrastructure. Two more — Electrification and AI Infrastructure — were added at the Milken Institute Conference in Los Angeles on 6 May 2024, taking the count to nine [S6]. That May 2024 release is the clearest statement of the original thesis: Alat would manufacture “network and communications equipment, servers, data center networking equipment, data center storage, industrial edge servers, and Industry 4.0 computing,” plus grid transmission and distribution technology.

Three weeks after the launch, on 20 February 2024, Alat announced four partnerships at once — SoftBank Group, Carrier Corporation, Dahua Technology and the Saudi Technology and Security Comprehensive Control Company (Tahakom) [S7]. On 4 March 2024 it signed a semiconductor partnership agreement with King Abdulaziz City for Science and Technology (KACST) at LEAP 2024 [S4]. On 29 May 2024 it announced the Lenovo collaboration. On 24 February 2025 it announced a joint venture with TK Elevator.

That is the complete list of manufacturing partnerships Alat has announced. No new manufacturing joint venture has been announced since 24 February 2025 — twenty-three months into a six-year programme, and seventeen months before the date of this page. Everything Alat has published since is either execution of those existing deals or a memorandum of understanding on talent and human capital.

The Alat Delivery Ledger

This is the audit the announcements never received. Each row states what was announced, the legal form it actually took, the figure attached, and what can be verified to exist on 31 July 2026. Ground-breaking is not production. A memorandum is not a joint venture. A convertible bond is not a factory.

VentureAnnouncedStructure actually announcedFigureStatus on 31 July 2026
Lenovo29 May 2024; closed 8 Jan 2025Three-year zero-coupon convertible bond plus strategic collaboration$2bn bond; first-phase plant above SAR2bn ($533m)Operating in part. Bond issued and outstanding; Riyadh plant ground broken 9 Feb 2025; trial production completed; commercial production targeted 2026; META regional HQ inaugurated 14 Apr 2026
TK Elevator24 Feb 2025; closed 5 Aug 2025Joint venture plus Alat minority stake in TKEEUR160m JV; SAR285m Dammam plantUnder construction. All statutory approvals received 5 Aug 2025; Dammam groundbreaking 10 Jun 2026; 40,000+ sq m; no production
SoftBank Group20 Feb 2024Joint venture / “strategic partnership”Up to $150mUnverified. First factory targeted for Dec 2024; Midha said exports would begin “by May 2025”; no opening, production or shipment has been announced since
Carrier Global20 Feb 2024“Collaboration agreement” — a commitment to buildNot disclosed; ~5,000 jobs claimedNo entity, site or groundbreaking announced. Carrier has never named Alat in an SEC filing
Dahua Technology20 Feb 2024Joint venture — Alat AIVisio Technology Co. Ltd.$200mNo production or groundbreaking announced since the launch release
Tahakom20 Feb 2024Partnership on smart mobilityNot disclosedNo further announcement
KACST4 Mar 2024Partnership agreement — R&D, talent, IC design clusterNone disclosedCounterparty unit dissolved. Alat’s Semiconductor Business Unit no longer exists
Dell TechnologiesNo Alat venture exists. Dell’s Saudi footprint is a merge and fulfilment centre in DammamDell has never named Alat in an SEC filing
CiscoNo Alat venture foundCisco has never named Alat in an SEC filing
Foxconn / Hon HaiNo Alat venture. Foxconn’s Saudi links are Ceer Motors (a PIF joint venture) and an EV-charger venture with Saleh Suleiman Alrajhi & SonsNot an Alat counterparty
Sensata TechnologiesNo Alat venture foundSensata has named neither Alat nor Saudi Arabia in any SEC filing since 2024
Elm / ZhongxinCould not stand upNo announcement located

Two entries in that table are doing most of the work, and both cut against the launch narrative.

The Lenovo $2bn is not factory capital. It is a three-year, zero-coupon convertible bond, completed on 8 January 2025 through the Hong Kong Stock Exchange, convertible at HK$10.42 a share into roughly 1.5 billion shares — about 12.1% of Lenovo’s enlarged capital base [S8]. Lenovo applied the proceeds to repaying debt and general corporate purposes. It is a corporate financing that bought Alat a large minority position in a listed PC maker, and a separate commercial agreement brought a plant to Riyadh. The plant’s own first-phase budget is SAR2bn ($533m) — roughly a quarter of the headline figure it is routinely conflated with [S9]. Both facts are true; only one of them is a Saudi factory.

The Carrier row is the most instructive. In February 2024 Carrier and Alat announced a manufacturing and R&D facility expected to create around 5,000 local jobs. Carrier Global is listed on the New York Stock Exchange. Between 1 January 2024 and 31 July 2026 it filed a continuous stream of 10-Ks, 10-Qs and 8-Ks, eight of which mention Saudi Arabia — and not one of which contains the word Alat [S10]. The same query returns zero Alat mentions for Dell Technologies and Cisco, and for Sensata Technologies it returns zero mentions of Alat and zero of Saudi Arabia. A public company discloses what is material to it. Silence across two and a half years of filings is not proof that nothing happened, but it is evidence about scale, and it is the opposite of what a 5,000-job manufacturing commitment would normally produce.

Who Runs Alat?

Dr Muhammad Nasser Aldawood, in an acting capacity. He heads the industrials and mining section within PIF’s MENA investments team, chairs Dussur, Qassim Cement Company and Arabian Drilling Company, and holds a PhD in management science and engineering from Stanford. Alat’s leadership page described him as Acting Chief Executive Officer when checked on 31 July 2026 [S3] — roughly four months after Midha’s removal was reported, with no permanent appointment announced.

Two details on that page are worth stating precisely. Alat’s newsroom still carries the press release announcing Amit Midha’s appointment as chief executive, dated 6 February 2024. It carries no release announcing his departure, and none announcing Aldawood’s arrival. And the three named sector presidents — Ray Wah for AI & Digital Hardware, Dr Karsten Hoppe for Building & Heavy Equipment, Dr Stefan Hoetzl for Home & Medical Equipment — cover three of the four sectors. There is no named president for Electrification, and no executive anywhere on the page with a semiconductor remit.

Midha’s recruitment was itself the strategy. He spent two decades at Dell Technologies, latterly as president of its Asia-Pacific and Japan business, and was hired to import the supply-chain and contract-manufacturing playbook that builds electronics industries. That is the sense in which Dell matters to the Alat story: through a curriculum vitae, not a joint venture.

Is Saudi Arabia Building a Chip Fab?

No, and it never announced one. This is the point at which most coverage of the Alat Saudi Arabia semiconductor programme loses precision, so it is worth being exact about what was ever on the table.

Alat’s Semiconductor Business Unit was described in March 2024 as covering “Power, Perception, and Processing” [S4] — power semiconductors, sensors, and logic. Power and perception devices are mature-node analogue and discrete products, typically built on 65nm and above, on 200mm wafers, using equipment that is decades old and largely outside the tightest export-control regimes. That was always the realistic tier for a new entrant: not leading-edge logic, but the unglamorous silicon that goes into inverters, motor drives, grid equipment and industrial sensors — which is precisely what an electrification-and-industrials portfolio would consume.

Even that tier requires a fab, and Alat never announced one. Its only semiconductor agreement was the KACST partnership: research collaboration, workforce development, and participation in the National Strategy for Semiconductors through an integrated-circuit design cluster. No site was named. No wafer size, no process node, no equipment order, no construction budget, no partner foundry. Semafor’s reporting that “several attempts to invest in chipmakers and secure agreements to build chip plants in the kingdom failed” is consistent with everything visible in the public record: there was an ambition and a business unit, and there was never a fab project to cancel.

The ladder of semiconductor capability runs design, then assembly, test and packaging, then mature-node fabrication, then leading-edge logic. Saudi Arabia is on the first rung. The tier that gets announced and the tier that gets built are rarely the same, a pattern the site has documented in Ceer’s electric-vehicle manufacturing timeline and in the gap between Oxagon’s industrial-city promises and its port reality.

Does Saudi Arabia Make Semiconductors?

No. There is no operating semiconductor fabrication plant in Saudi Arabia, and no announced assembly, test and packaging facility. What exists is a design programme.

The National Semiconductor Hub, established under KACST with the Ministry of Communications and Information Technology, targets over SAR1bn of investment and the establishment of 50 chip-design companies in the kingdom by 2030, supported by a National Capability Center for Semiconductors spanning some 3,600 square metres across KACST and KAUST [S11]. Its model is to relocate fabless design teams, fund them to a first tapeout, and build a local design workforce. A tapeout is a completed design handed to somebody else’s fab — almost certainly in Taiwan, Korea or China. It is a genuine capability and a reasonable first rung. It is not manufacturing.

The distinction matters because Saudi Arabia’s chip access improved sharply through a different channel entirely. The Biden administration’s AI diffusion rule, which would have capped advanced-compute exports to Saudi Arabia, was rescinded by the Bureau of Industry and Security in May 2025, and Nvidia and AMD were cleared to supply HUMAIN for its data center build-out. The kingdom got much better at buying chips in exactly the period it stopped trying to make them — a sequence traced in detail in our analysis of Nvidia GPU supply and Saudi export-control exposure. Buying, however, remains conditional on Washington: by July 2026 the Emirates had won a rule-based clearance while Riyadh was still licensed case by case, a split set out in our piece on how the two states’ chip access diverged.

What Happened to Alat’s $100 Billion?

Nobody outside PIF knows, because Alat has never published a cumulative deployment figure. The $100bn was always an envelope to 2030 rather than committed capital, and the disclosed commitments do not come close to it: $2bn in the Lenovo convertible bond, EUR160m in the TK Elevator joint venture, $200m with Dahua, up to $150m with SoftBank, SAR2bn in the Riyadh plant’s first phase, SAR285m in the Dammam plant. Add every announced number together and the total is in the low single-digit billions of dollars — under 5% of the headline, at roughly 45% of the elapsed time to 2030.

The absence of a deployment figure is itself the finding. Alat publishes press releases about talent programmes and memoranda; it does not publish capital deployed, jobs created against the 39,000 target, or GDP contribution against the $9.3bn target. Neither does PIF break Alat out in its disclosures. The result is a $100bn commitment with no published denominator, which is a governance characteristic rather than an accident, and it is the same characteristic that made PIF’s 2025 results legible only to readers who went past the headline profit number.

Why Alat Survived the Retrenchment When Sport Did Not

Because advanced manufacturing is in the strategy and sport is not.

PIF’s board, chaired by the Crown Prince, approved the fund’s 2026-2030 strategy on 15 April 2026, built around six integrated ecosystems: tourism, travel and entertainment; urban development and livability; advanced manufacturing and innovation; industrials and logistics; clean energy, water and renewables infrastructure; and NEOM [S12]. Advanced manufacturing made the list. Sport did not. That single fact explains why Alat kept its chairman, its capital line and its remaining sectors while the fund was cutting elsewhere — and why the semiconductor unit, the one limb with no plant, no partner and no path, was the part that went.

The direction of travel around it is unambiguous. On 23 June 2026, at the FII Priority Europe summit in Rome, PIF governor Yasir Al-Rumayyan described the fund as directing 80% of its capital domestically, with foreign allocations cut to 20% from a peak of 30%, and framed the shift as “Now our new strategy is to bring the world back to Saudi” [S13]. A fund pulling capital home has more reason to build factories in Riyadh, not less. It also has less appetite for the one category of industrial project — a fab — where the capital goes out for a decade before anything comes back. Read alongside PIF’s 2026-2030 capital allocation and the $16bn NEOM has budgeted purely to cancel contracts, the Alat semiconductor decision looks less like a reversal and more like triage: keep what has a counterparty and a site, drop what has neither.

The Constraint Nobody Priced: Power and Water

A fab is an electricity and water project wearing a technology costume. A modern logic fab draws hundreds of megawatts continuously — not peak, continuously — and consumes millions of litres of ultrapure water a day, with the ultrapure treatment train itself a substantial additional load. Fabs cannot brown out. A momentary voltage sag ruins a wafer lot worth millions.

Saudi Arabia is already power-constrained for a far less demanding class of industrial load. The kingdom operated roughly 467MW of data center capacity in Q1 2026 against a 6.6GW target for 2034, and grid interconnection, energisation queues and gas-turbine procurement — not chips — set the timeline. That argument is set out in full in our pillar analysis, Saudi Arabia’s AI data center power constraint is the grid, not the chips, and it applies with more force to fabrication than to compute. Data centers can be sited where power exists and phased in modules; a fab is a single indivisible load that must be energised before the first wafer starts and cannot be built in halves.

Electronics assembly of the Lenovo kind is a different proposition. It is labour- and logistics-intensive rather than power-intensive, it can be phased, and it can be sited inside an existing industrial city on existing supply. The two Alat ventures that are physically progressing are exactly the two that do not need a new grid.

Why This Matters for Vision 2030

Industrial localisation is being attempted into a contracting economy. GASTAT’s flash estimate for Q2 2026, published on 30 July 2026, put real GDP down 4.8% year on year, with oil activities down 24.7% and non-oil activities up just 0.6% [S14]. The non-oil deceleration sequence is now 4.9% for full-year 2025, 2.9% in Q1 2026, and 0.6% in Q2 — the series that carries the non-oil GDP growth KPI and, through it, much of the Vision 2030 diversification case. On a seasonally adjusted quarterly basis, non-oil activity actually contracted 0.5%.

Manufacturing sits inside that non-oil number, though the flash estimate does not break it out; the sector detail arrives with the full quarterly national accounts. What the flash does establish is the environment. The National Industrial Development and Logistics Programme targets a manufacturing-led lift in non-oil GDP at precisely the moment non-oil growth has fallen to a rounding error, and PIF is redirecting capital home into that same economy.

Alat was the flagship of the electronics half of that plan — the counterpart to Ceer in vehicles, Lucid in premium EVs and the broader Saudi manufacturing sector build-out. Its trajectory suggests the achievable version of the plan is assembly and systems integration under foreign technology licence, with domestic demand as the anchor. That is a real industry. It is not the industry that was announced, and it does not produce a chip.

Risks, Contradictions and Open Questions

The retreat is reported, not documented. Semafor’s account rests on people familiar with the matter, and Alat’s on-the-record response asserted that its “operations, partnerships, and strategic direction remain firmly in place.” What is documented is the corporate structure: nine business units on 17 May 2026, four sectors and a redirecting semiconductor URL on 31 July 2026. Those are consistent with the reporting but are not a confirmation of it, and Alat has issued no statement about the semiconductor unit either way.

Absence of announcement is not absence of activity. Six of the twelve ledger rows say no further announcement was found. Alat is a private PIF subsidiary with no obligation to publish, and joint ventures can be quietly incorporated, capitalised and built without a press release. The Dahua venture in particular has a named legal entity — Alat AIVisio Technology Co. Ltd. — and may well exist and operate. Saudi commercial-registry records are not comprehensively available for automated verification, and this audit does not claim they were checked.

The SEC finding proves materiality, not non-existence. Carrier, Dell, Cisco and Sensata not naming Alat in any filing means the arrangements are immaterial to companies of their size. It does not mean nothing was signed. The correct reading is that these are commercial agreements small enough to sit below a US disclosure threshold — which is itself informative when one of them was announced with a 5,000-job headline.

The SoftBank robotics venture is the largest unresolved item. It was announced with a December 2024 factory opening, revised by Midha in February 2025 to exports “by May 2025”, and has produced no verifiable milestone since. Either it is operating quietly, or the first of Alat’s four launch partnerships has not delivered a factory in twenty-nine months. Both readings are currently available, and neither party has clarified.

Four of the partnerships in circulation do not appear to exist at all. Dell, Cisco, Foxconn and Sensata are frequently attached to Alat in secondary coverage. This audit could not locate an announcement for any of them. Foxconn’s Saudi activity is real but runs through Ceer and an Alrajhi charger venture, neither of which is an Alat counterparty.

What to Watch Next

  • Whether Alat names a permanent chief executive. Aldawood has been acting for roughly four months and holds a full-time PIF role alongside it. A permanent external hire would signal the mandate is being rebuilt; an extended acting arrangement would signal it is being absorbed into PIF’s industrials portfolio.
  • First commercial shipment from the Riyadh Lenovo plant. Trial production is complete; the stated target is two million units a year within 18 to 24 months. A dated shipment announcement is the first hard delivery milestone the programme will have produced.
  • The Q2 2026 full national accounts, which will break out manufacturing from the 0.6% non-oil aggregate.
  • TKE Alat’s Dammam commissioning, expected to follow the 10 June 2026 groundbreaking on a SAR285m, 40,000-square-metre site.
  • Any Saudi announcement of an assembly, test and packaging facility. ATP is the realistic next rung above design, needs a fraction of a fab’s power, and would be the first genuine evidence that semiconductor manufacturing ambitions have been rebuilt rather than abandoned.
  • PIF’s 2026 annual report, and whether it discloses anything about Alat’s deployed capital against the $100bn envelope.

Sources