SoftBank Group completed its acquisition of UK AI-chip designer Graphcore on July 11, 2024. Graphcore became a wholly owned SoftBank subsidiary, but kept its name, Bristol headquarters and distinct corporate identity. The companies did not disclose the purchase price. Contemporary reports estimated the deal at roughly $400 million to $500 million, but those figures were not confirmed and Graphcore’s chief executive reportedly disputed the widely circulated $500 million figure.
As of August 18, 2026, Graphcore remains part of SoftBank’s AI Computing Segment alongside Arm and Ampere. The acquisition gave SoftBank specialist AI-accelerator expertise and engineering talent, but it did not establish that Graphcore had overtaken Nvidia or achieved mass-market dominance.
As an Amazon Associate I earn from qualifying purchases.
What happened in the SoftBank–Graphcore deal?
| Item | Confirmed information |
|---|---|
| Buyer | SoftBank Group Corp., the Japanese technology-investment group |
| Target | Graphcore Limited, a UK semiconductor company founded in Bristol in 2016 |
| Announcement and completion | July 11, 2024 |
| Ownership | Graphcore became a wholly owned SoftBank subsidiary |
| Purchase price | Not disclosed by either company |
| Headquarters after the deal | Bristol, UK |
| Operating identity | Graphcore continued operating under the Graphcore name |
Graphcore announced that it had joined SoftBank Group to work on next-generation AI computing. This was a completed acquisition, not merely a proposal to buy the company. SoftBank subsequently recorded Graphcore as a subsidiary.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errorsThe transaction also did not amount to Graphcore being absorbed into Arm. SoftBank places Graphcore, Arm and Ampere in a wider AI-computing grouping, but public materials continue to identify Graphcore as a separate subsidiary.
#1 Best Overall
- NVIDIA Volta GV100 Architecture — 4,608 CUDA Cores, 640 1st-Gen Tensor Cores delivering 14 TFLOPS FP32 and 112 TFLOPS deep learning performance for AI training, inference, HPC, and scientific computing workloads
- 32GB HBM2 ECC Memory — 900 GB/s Bandwidth — High-bandwidth memory on a 4096-bit bus with ECC error correction provides the memory capacity and throughput required for the largest AI models, simulations, and datasets
- PCIe 3.0 x16 Interface — 250W TDP — Standard PCIe Gen3 connectivity with passive cooling designed for enterprise rack server deployment in HPE ProLiant, Dell PowerEdge, and Supermicro platforms with adequate chassis airflow
- NVLink — Scale to 96GB Unified Memory — Connect two V100 GPUs via NVLink at 300 GB/s bi-directional bandwidth to scale GPU memory from 32GB to 96GB for larger AI training and HPC workloads
- Multi-Precision Computing — Supports FP64 (7 TFLOPS), FP32 (14 TFLOPS), FP16 (112 TFLOPS) and INT8 precision modes for flexible deployment across training, inference, and scientific simulation workloads
Graphcore’s operations have remained associated with Bristol, with offices also listed in Cambridge, London, Gdańsk and Hsinchu. The company said it would continue investing in high-skilled roles across multiple disciplines. Graphcore’s acquisition announcement described the company’s continuing role within SoftBank.
How much did SoftBank pay for Graphcore?
The confirmed answer is that the financial terms were undisclosed. Neither SoftBank nor Graphcore published the total consideration paid.
Contemporary media reports placed the deal at approximately $400 million to $500 million. EE Times cited an industry estimate of about $400 million, while TechCrunch reported that the commonly repeated $500 million figure was inaccurate according to Graphcore co-founder Nigel Toon. Those numbers should therefore be treated as estimates, not the purchase price.
The reported figures were also materially below Graphcore’s earlier private valuation. That comparison needs care: a previous fundraising valuation reflects the price investors accepted for a particular minority stake at a particular time. It does not guarantee what a buyer will pay later for the whole company.
A disclosure from Schroder British Opportunities Trust provides additional context but does not reveal the transaction’s total value. The trust said the sale of its Graphcore holding represented a 22% uplift over its latest first-quarter 2024 valuation. Its regulatory announcement does not establish the price SoftBank paid for all Graphcore shares.
What did Graphcore make?
Graphcore developed AI accelerators called intelligence processing units, or IPUs. It also built Poplar, the software stack used to program and manage its hardware.
An accelerator is a processor designed to handle particular workloads more efficiently than a general-purpose CPU. A GPU is a graphics-processing architecture that has been extensively adapted for parallel AI calculations. Graphcore’s IPU was a different, proprietary architecture designed specifically around machine-learning workloads.
Quick wins for a faster PC:
Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →Repair Windows errors before they cause bigger problemsFix Now →Graphcore’s proposition was not simply to sell a chip. Its hardware and Poplar software were intended to form an alternative AI-computing platform. That distinction matters because AI-chip adoption depends on much more than the processor’s theoretical capabilities. Customers also need compatible frameworks, libraries, developer tools, manageable model-porting requirements, cloud access, production references, reliable supply and technical support.
Rank #2
- Professional AI & Creator Workstation: AMD Radeon AI PRO R9700 GPU with 32GB GDDR6 is engineered for AI development, professional content creation, and compute-intensive workloads.
- Massive 32GB Memory Capacity: 32GB of GDDR6 memory on a 256-bit bus provides ample bandwidth for large AI models, 8K video editing, and complex 3D rendering.
- Advanced RDNA 4 with AI Accelerators: 64 Compute Units with 3rd Gen Ray Tracing and dedicated 2nd Gen AI Accelerators for groundbreaking AI performance and visual computing.
- Professional Blower Cooling: Efficient single blower design exhausts heat directly out of the chassis, ideal for multi-GPU workstation and server configurations.
- Enterprise-Grade Thermal Solution: Vapor chamber heatsink with industrial Honeywell PTM7950 thermal interface material ensures reliable cooling under sustained professional loads.
An IPU is not automatically better or worse than a GPU. The result depends on the workload, memory architecture, interconnects, software support, system design, availability and total cost. Graphcore sought to offer an alternative to GPU-based systems, particularly those built around Nvidia’s powerful hardware and software ecosystem, but that ambition did not make it an equal competitor to Nvidia by market scale.
TechCrunch’s transaction coverage and EE Times’ report describe Graphcore’s IPU and Poplar approach.
Why did Graphcore need a buyer?
The acquisition is best understood as a scale-and-capital transaction as well as a technology deal.
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Graphcore had raised substantial venture funding and reached a multibillion-dollar private valuation during an earlier period. But building a competitive AI-chip business requires sustained spending on chip design, software, manufacturing relationships, systems, customer support and sales. It also requires enough real-world deployments to persuade developers and enterprises to adopt a platform outside the market leader’s ecosystem.
Nvidia’s advantage extended beyond individual chips. Its software tools, libraries, developer familiarity, cloud availability and customer relationships made it difficult for challengers to convert technical differentiation into broad commercial adoption. Contemporary coverage described Graphcore as needing additional capital, scale and customer access to compete globally.
That does not mean Graphcore’s technology lacked value. A specialist company can possess valuable architecture, intellectual property and engineering talent while still finding it difficult to finance and commercialize a worldwide platform independently. A parent with greater resources may provide more time for product development, access to partners and customers, manufacturing relationships, and the ability to preserve specialist teams through a longer market cycle.
Why did SoftBank buy Graphcore?
SoftBank’s stated rationale
SoftBank said the acquisition would strengthen its AI-semiconductor capabilities and support the development of next-generation AI-compute systems.
Free tools Windows power users keep installed
One-click scans. No signup required.
SoftBank executive Yoshimitsu Goto highlighted Graphcore’s people and engineering resources as a fit with SoftBank’s effort to develop AI-chip businesses centered on Arm. SoftBank investor materials described Graphcore as specializing in the design, development and commercialization of AI and machine-learning chips.
In practical terms, the deal added an established specialist in AI acceleration to a portfolio that already included major semiconductor capabilities. SoftBank’s FY2024 conference-call materials and earnings presentation set out the group’s rationale.
What the strategy may imply
A reasonable interpretation is that SoftBank was buying both technology and talent at a potentially lower price than Graphcore’s earlier private valuation. It could also use the acquisition to build a broader semiconductor portfolio covering CPU architecture, AI acceleration and related semiconductor IP.
Another possible objective is to reduce dependence on outside suppliers for parts of the AI-compute stack and gain greater exposure to the infrastructure behind AI growth. That is an analytical interpretation rather than a separately disclosed breakdown of SoftBank’s acquisition rationale.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →How does Graphcore fit with Arm and Ampere?
SoftBank’s current AI Computing Segment groups three relevant businesses:
- Arm: Semiconductor IP and compute platforms, including CPU-related technology and proprietary silicon.
- Ampere: High-performance, energy-efficient Arm-based CPUs for data-centre and AI-oriented workloads.
- Graphcore: Specialist AI-chip design and semiconductor expertise.
SoftBank says these businesses are working together toward future AI-compute platforms delivered through semiconductor IP, chips, software and related technologies. SoftBank’s AI Computing Segment page confirms Graphcore’s continuing place in that structure as of August 18, 2026.
However, the public information does not establish a fully integrated Graphcore–Arm–Ampere product stack. Nor does it provide a detailed technical roadmap showing that Graphcore IP has been incorporated into Arm products. Common ownership and strategic coordination should not be confused with a specific announced product integration.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What happened to Graphcore after the acquisition?
Graphcore remains a wholly owned SoftBank subsidiary and continues to operate under the Graphcore name. It remains headquartered in Bristol, and SoftBank’s 2026 materials continue to describe it as part of the AI Computing Segment and as a business with AI-chip design expertise.
Outdated Drivers Are Slowing You Down
One free scan finds every outdated or missing driver and matches the right update for your exact hardware.Free scan · exact hardware matchWindows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallGraphcore co-founder Nigel Toon remained involved after the transaction. Because executive roles can change, current leadership titles should be checked against Graphcore’s leadership information rather than inferred from the 2024 announcement, which identified Toon as CEO at the time.
Rank #4
- Powered by Radeon AI PRO R9700 - Supercharge you workflow with the cutting-edge RDNA 4 Architecture and 2nd-gen AI Accelerators.
- 32GB GDDR6 with 256-bit memory bus - Tackle larger, more complex projects without limits.
- PCIe Gen 5 - Unlock lightning-fast data transfers with PCIe Gen 5 support.
- GIGABYTE TURBO Fan Cooling System - Indented metal cover and blower fan increase airflow intake, while the vapor chamber, all copper heat sink, and metal frame offer efficient heat dissipation. Optimized airflow design allows for easy multi-GPU scalability.
- Double Ball Bearing Fan - Delivers superior heat resistance and rotational efficiency for better performance and a longer lifespan compared to conventional sleeve fans.
The available official materials do not verify that Graphcore has launched a specific post-acquisition chip, placed its chips in a named SoftBank data centre, achieved profitability, or displaced Nvidia in a major market. They also do not confirm that Graphcore technology is inside Arm’s 2026 AGI CPU.
What does the deal mean for the UK?
The acquisition preserved a Bristol-founded, UK-headquartered semiconductor company within the UK technology ecosystem rather than ending Graphcore as a recognizable business. SoftBank’s capital may give the company greater capacity to fund engineering and pursue customers than it would have had as a standalone venture-backed company.
There is also a broader industrial-policy tension. The transaction illustrates both the international value of UK semiconductor talent and the difficulty of scaling globally competitive chip companies from the UK without much larger pools of capital, manufacturing access and customer volume. That is an analysis of the deal’s significance, not a disclosed term or promise made by SoftBank.
For the UK, the important distinction is between retaining a company and retaining full economic ownership. Graphcore remains a UK-founded and UK-headquartered business, but after July 11, 2024 it became owned by Japan-based SoftBank Group.
What should investors and technology readers watch next?
The acquisition’s ultimate success will depend on execution rather than the ownership announcement alone. The most useful indicators are:
- Whether Graphcore announces new products or meaningful updates to its IPU and Poplar platform.
- Whether its technology is integrated into specific Arm or Ampere products, rather than merely grouped within the same SoftBank segment.
- Whether SoftBank discloses additional investment, revenue contribution or commercial milestones.
- Whether Graphcore expands beyond specialist deployments into broader enterprise, cloud or data-centre use.
- Whether the company can reduce the software and ecosystem friction that makes alternative AI accelerators difficult to adopt.
Until those details are disclosed, the acquisition should be viewed as a strategic option for SoftBank: it secured AI-chip expertise and talent, but still faces the technical, commercial and ecosystem challenges that confronted Graphcore as an independent company.
The bottom line
SoftBank acquired Graphcore on July 11, 2024, for an undisclosed amount. Graphcore remains a separate, wholly owned subsidiary headquartered in Bristol, and SoftBank still places it alongside Arm and Ampere in its AI Computing Segment in 2026. The deal is best understood as SoftBank’s attempt to build AI-semiconductor capability and secure specialist talent—not as proof that Graphcore had defeated Nvidia or that a specific combined product roadmap had already been delivered.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
