Advanced Micro Devices, Inc. (AMD): Business Model Canvas [Dec-2025 Updated]

You're trying to figure out how Advanced Micro Devices, Inc. is actually making its money now that it's a serious contender in the AI server space, and frankly, the numbers from late 2025 tell a clear story. The Data Center segment alone pulled in $4.3 billion in Q3 2025, showing their pivot to EPYC CPUs and Instinct accelerators is paying off big time, all while balancing that high-cost, fabless model dependent on TSMC. To understand the engine driving this growth-from their Zen 5 designs to the ROCm software ecosystem-you need to see the full Business Model Canvas below, so you can map the risks and opportunities clearly.

Advanced Micro Devices, Inc. (AMD) - Canvas Business Model: Key Partnerships

You're looking at the critical external relationships that fuel Advanced Micro Devices, Inc. (AMD)'s growth engine as of late 2025. These aren't just vendor agreements; they are foundational to the company's strategy in AI and computing.

TSMC for leading-edge silicon manufacturing capacity

The relationship with Taiwan Semiconductor Manufacturing Company (TSMC) is about securing the most advanced process nodes. Honestly, capacity is the bottleneck across the industry right now. TSMC Chairman and CEO C. C. Wei estimated in November 2025 that the company's existing advanced-node capacity is still roughly three times short (3X) of what its major customers plan to consume. To manage this, TSMC is aggressively expanding, projecting that AI processor-related revenue will account for a mid-teens percentage of its total revenue in 2025. Currently, TSMC Arizona is producing chips for two Apple products and one AMD product. TSMC is doubling its advanced chip packaging capacity by 2025, focusing on CoWoS technology to keep up with demand from key clients like Advanced Micro Devices, Inc. (AMD).

OpenAI for multi-year, multi-generation AI GPU deployment

The deal with OpenAI is a massive commitment to Advanced Micro Devices, Inc. (AMD)'s Instinct GPU roadmap. OpenAI plans to deploy a total of 6 gigawatts (GW) of Advanced Micro Devices, Inc. (AMD) GPUs over several years. The initial phase involves deploying 1 gigawatt of Instinct MI450 series GPUs in the second half of 2026. To align interests, Advanced Micro Devices, Inc. (AMD) issued OpenAI a performance-based warrant for up to 160 million shares of Advanced Micro Devices, Inc. (AMD) stock. This partnership is expected to generate tens of billions of dollars in annual AI-related revenue for Advanced Micro Devices, Inc. (AMD), potentially exceeding $100 billion in revenue from OpenAI and related customers over four years.

Here's the quick math on the scale of this commitment:

Metric	Value	Notes
Total GPU Deployment Commitment	6 gigawatts (GW)	Total capacity over multiple years.
Initial Deployment Capacity	1 gigawatt (GW)	Using Instinct MI450 series, starting H2 2026.
Stock Incentive	Warrant for up to 160 million shares	Vests based on deployment and share price milestones.
Projected Annual Revenue Impact	Tens of billions of dollars	Annual revenue contribution once deployments ramp.

Hyperscalers (e.g., Microsoft, Google) for cloud instance adoption

The cloud providers are driving significant top-line performance for Advanced Micro Devices, Inc. (AMD). In the third quarter of 2025, Data Center revenues hit $4.34 billion, which was 47% of total revenues. This segment surged 57% year-on-year in Q1 2025. You can see the adoption directly in the cloud offerings:

• Hyperscalers launched over 160 EPYC-powered instances in Q3 2025.
• New Turin offerings from Google and Microsoft Azure were among those launched.
• There are now more than 1,350 public EPYC cloud instances globally available.
• This global instance count is up roughly 50% year-over-year.

Overall, Advanced Micro Devices, Inc. (AMD) expects Data Center revenues to reach $16 billion for the full 2025 fiscal year, out of total expected revenue of $34 billion.

HPE for open rack-scale AI infrastructure (Helios architecture)

Hewlett Packard Enterprise (HPE) is a key design partner for Advanced Micro Devices, Inc. (AMD)'s open rack-scale AI platform, 'Helios.' This architecture combines EPYC CPUs, Instinct GPUs, Pensando advanced networking, and the ROCm open software stack. The system is engineered to deliver up to 2.9 exaFLOPS of FP4 performance per rack. HPE will be one of the first OEMs to adopt this, with worldwide availability planned for 2026. Also, the new 'Herder' supercomputer for the High-Performance Computing Centre Stuttgart is built on the HPE Cray GX5000 platform using Instinct MI430X GPUs and next-generation EPYC "Venice" CPUs.

PC OEMs (Dell, HP, Lenovo) for commercial and consumer systems

The Client segment remains a substantial portion of Advanced Micro Devices, Inc. (AMD)'s business, showing strong rebound. In the third quarter of 2025, Client and Gaming revenues accounted for 43.8% of total revenues. The Client segment top-line figure specifically increased 72.8% year-over-year. This growth is supported by the launch of Ryzen Threadripper 9000WX and Threadripper PRO 9000X processors for high-end desktop performance. Dell Technologies is explicitly listed as part of Advanced Micro Devices, Inc. (AMD)'s rich partner base driving prospects in the data center space, which often includes commercial PC system integration.

Advanced Micro Devices, Inc. (AMD) - Canvas Business Model: Key Activities

High-performance CPU/GPU/NPU architecture design (Zen 5, MI400)

Advanced Micro Devices, Inc. (AMD) is driving architecture design across its product lines. The Zen 5 architecture is currently powering the fifth-gen AMD EPYC processors, which are a key revenue driver in the data center segment. Looking ahead, the Zen 6 CPUs are codenamed "Venice" and "Verano," targeting data center workloads. On the accelerator front, the Instinct MI350 series, based on CDNA 4 architecture, is launching mid-2025, promising a 35x performance bump over the MI300X in inference. The MI350 series is expected to offer 8 TB/s of memory bandwidth. The next generation, the MI400 Series, is planned for 2026. The MI400 is projected to deliver up to 10 times higher inference performance on Mixture of Experts (MoE) models compared to the MI300X generation.

The MI400 series is designed to leverage HBM4 memory, targeting 432 GB capacity, which is a 50% uplift from the 288GB HBM3e in the prior generation, achieving a massive 19.6 TB/s bandwidth. Compute projections for the MI400 include 40 PFLOPS for FP4 and 20 PFLOPS for FP8 performance. For client devices, Ryzen AI chips featuring an integrated NPU have seen notebook sell-through jump 50% quarter-over-quarter since their Q1 launch, with design wins up 80% year-over-year.

Product Family	Architecture/Generation	Target Segment	Key Performance Metric
EPYC CPU	Zen 5 (Turin)	Data Center	Currently driving revenue
Instinct GPU	MI350 (CDNA 4)	AI/HPC	Promised 35x inference performance vs MI300X
Instinct GPU	MI400 (CDNA-Next)	AI/HPC	Expected 40 PFLOPS (FP4)
Ryzen AI CPU	Zen 5 based	Client/PC	Notebook sell-through up 50% Q/Q

Developing the ROCm open-source AI software ecosystem

Advanced Micro Devices, Inc. (AMD) is actively developing its open-source software stack, ROCm, to support its hardware. ROCm version 7 was announced and is scheduled for general accessibility in Q3 2025. This new version delivers significant performance gains over ROCm 6, showing over 3.5x better inference capability and 3x better training prowess in internal tests. The software ecosystem's viability is being bolstered by broad framework support; ROCm 7 supports over 2 million Hugging Face models, including LLaMA 4 and Gemma 3, on day one. For specific models, ROCm 7 showed a 3.2x inference performance increase for Llama 3.1 70B compared to ROCm 6. Furthermore, ROCm support is expanding to client devices, with expected wide availability on Windows and Radeon GPUs in the second half of 2025.

• ROCm 7 Performance Uplift vs ROCm 6: 3.5X Inference
• ROCm 7 Performance Uplift vs ROCm 6: 3X Training
• ROCm 7 Day-One Model Support: Over 2 million Hugging Face models
• ROCm 7 Windows/Radeon Support Availability: 2H 2025

Managing the complex fabless supply chain and foundry capacity

As a fabless company, Advanced Micro Devices, Inc. (AMD) manages a complex supply chain heavily reliant on external foundries. The company is reportedly considering manufacturing partners beyond TSMC to mitigate supply chain risks. TSMC's advanced packaging capacity, including CoWoS and SoIC, for both 2024 and 2025 has been fully booked by major customers including Advanced Micro Devices, Inc. (AMD). TSMC's Arizona fabrication plants are scheduled to begin production in 2025. Near-term challenges included U.S. export restrictions on advanced AI chips destined for China. These restrictions resulted in an inventory write-down of approximately $800 million in Q2 2025. The MI308 series inventory prepared for China in Q1 2025 was valued at over $500 million. The impact on profitability was significant, with non-GAAP gross margins falling to 43% in Q2 2025, compared to a forecasted 54%. If these export licenses are not secured, the company faces a potential recurring annual revenue drag of $1.5 billion.

Scaling data center AI and HPC system-level solutions

Advanced Micro Devices, Inc. (AMD) is aggressively scaling its data center and AI solutions. The Data Center segment reported revenue of $4.3 billion in Q3 2025, marking a 22% year-over-year increase. For the same quarter, total company revenue reached a record $9.2 billion, a 36% year-over-year surge. In Q2 2025, AI GPU sales (excluding China-related shipments) reached $1 billion, representing a 40% year-over-year increase. The company reports that seven of the top ten AI model builders are currently leveraging Advanced Micro Devices, Inc. (AMD)'s Instinct GPUs.

The company is building out rack-scale solutions. The next-generation 'Helios' rack system is designed to integrate up to 72 MI400 Series GPUs and will offer up to 260 TB/s of scale-up bandwidth. Advanced Micro Devices, Inc. (AMD) has a commitment with OpenAI to deploy 6 GW of Advanced Micro Devices, Inc. (AMD) GPUs, starting with 1 GW of MI450 GPUs in H2 2026. Long-term financial targets include a greater than 60% revenue compound annual growth rate (CAGR) for the data center business and an AI revenue CAGR of more than 80%.

• Q3 2025 Data Center Revenue: $4.3 billion
• Q3 2025 Total Revenue: $9.2 billion
• AI GPU Revenue Growth (YoY, Q2 2025 excl. China): 40%
• MI350 Series: Fastest ramping product in company history
• Helios Rack System GPU Capacity: Up to 72 MI400 Series GPUs

Advanced Micro Devices, Inc. (AMD) - Canvas Business Model: Key Resources

You're looking at the foundation that lets Advanced Micro Devices, Inc. (AMD) compete at the highest levels, which really boils down to their silicon DNA and the manufacturing access that brings it to life.

Core Intellectual Property (IP): Zen, RDNA, and Infinity Fabric Architectures

The value here is in the proprietary instruction set and graphics designs that power their major product lines. The success of the Zen 5 architecture, for example, drove Client revenue to a record $2.8 billion in the third quarter of 2025.

The performance of the Data Center segment, which reached $4.3 billion in revenue for the third quarter of 2025, is directly tied to the EPYC processors utilizing the latest Zen iterations and the Instinct GPUs leveraging RDNA/CDNA lineage.

• Infinity Fabric architecture enables chiplet designs, a key differentiator in high core-count server CPUs.
• RDNA architecture underpins the Radeon gaming GPUs and the high-performance Instinct accelerators.
• Zen architecture continues to drive market share gains, with Client CPU market share reaching over 50% of the desktop channel as of Q2 2025.

World-class R&D Talent and Chip Design Expertise

This resource is quantified by the sheer investment Advanced Micro Devices, Inc. (AMD) pours into its engineering teams. The company is clearly prioritizing future product development, as evidenced by the spending levels.

Research and development expenses for the twelve months ending September 30, 2025, totaled $7.473B. More recently, the third quarter of 2025 saw R&D spend rise 31% year-over-year to $2.14 billion, even as the non-GAAP gross margin held steady at 54%.

This sustained investment supports the roadmap, including the planned Instinct MI400 series GPU for 2026 and the EPYC 'Verano' CPU (Zen 7) for 2027.

Advanced Process Node Access via TSMC (e.g., 2nm for Venice EPYC)

Access to Taiwan Semiconductor Manufacturing Company (TSMC)'s leading-edge fabrication is a non-negotiable resource. TSMC is pushing its 2nm mass production to commence in 2025, which is critical for Advanced Micro Devices, Inc. (AMD)'s next-generation EPYC processors, like the rumored 'Venice' generation.

The entire AI accelerator ramp is dependent on this foundry relationship. Advanced Micro Devices, Inc. (AMD) secured an agreement with OpenAI to deploy 6 gigawatts of compute power, with the first 1-gigawatt deployment of MI450 GPUs scheduled for the second half of 2026, all relying on TSMC's advanced nodes.

TSMC's planned capital expenditure for 2025 is between $38 billion and $42 billion, showing the scale of the ecosystem Advanced Micro Devices, Inc. (AMD) taps into.

Xilinx's Adaptive Computing Technology (FPGAs and SoCs)

The integration of Xilinx technology forms the Adaptive and Embedded Computing Group (AECG), providing differentiation beyond traditional CPUs and GPUs. This IP is driving significant design pipeline growth.

Since the acquisition, design wins have exceeded $14 billion in 2024 and are on a path to exceed $16 billion in 2025, representing over $50 billion of new design wins at a 21% CAGR.

The Embedded segment, which primarily houses this technology, reported revenue of $857 million in the third quarter of 2025.

Here's a quick look at the scale supporting these resources as of late 2025:

Metric	Value (Latest Reported Period)	Context
Total Revenue (Q3 2025)	$9.2 billion	Quarterly revenue, up 36% YoY
Data Center Revenue (Q3 2025)	$4.3 billion	Driven by 5th Gen EPYC and MI350 GPUs
LTM R&D Expense (to Sep 30, 2025)	$7.473B	Investment in future IP and talent
Server CPU Revenue Share (Q2 2025)	41%	Market share for EPYC processors
Total New Design Wins (CAGR since Xilinx acquisition)	21%	Reflecting Xilinx IP integration success

The company's ability to secure a multi-year OpenAI agreement for 6 gigawatts of compute power underscores the perceived value of its combined CPU/GPU/Adaptive IP portfolio.

Finance: draft 13-week cash view by Friday.

Advanced Micro Devices, Inc. (AMD) - Canvas Business Model: Value Propositions

You're looking at the core reasons customers choose Advanced Micro Devices, Inc. (AMD) products right now, late in 2025. It's all about performance leadership in key, high-growth areas, even if some parts of the business, like Embedded, are facing a tougher patch.

Data Center: Superior performance-per-watt EPYC CPUs and open AI accelerators

The Data Center segment is where Advanced Micro Devices, Inc. (AMD) is making its biggest strategic play, especially with AI accelerators. You saw the Q3 2025 revenue hit $4.3 billion, which was a solid 22% increase year-over-year, and even better, a 34% jump sequentially, largely thanks to the ramp of the AMD Instinct MI350 Series GPUs. The value here is performance-per-watt and an open software alternative to proprietary stacks. For instance, the fifth-generation EPYC Turin processors hit an all-time high in server CPU revenue, making up nearly half of all EPYC sales in the quarter. Looking out, Advanced Micro Devices, Inc. (AMD) is targeting a CAGR of more than 80% for its Data Center AI revenues over the next three to five years, betting on the total addressable market hitting roughly $200 billion by 2030. Plus, the strategic partnership with OpenAI commits Advanced Micro Devices, Inc. (AMD) to deploying 6 gigawatts (GW) of next-generation AI computing capacity, starting with 1 GW of MI450 GPUs in the second half of 2026. That's a concrete commitment to scale.

Client: AI-enabled computing with Ryzen AI 300 Series processors

For the Client side, the value proposition centers on bringing powerful, AI-enabled computing to the PC user, which is definitely paying off. Client revenue alone reached a record $2.8 billion in Q3 2025, growing 46% year-over-year. The Ryzen AI 300 Series processors, codenamed "Krackan Point," are key here, featuring the XDNA 2 Neural Processing Unit (NPU) delivering 50 to 55 TOPS (trillion operations per second) of AI performance. This NPU capability is what powers the next generation of Microsoft Copilot+ experiences. In benchmarks, the flagship Ryzen AI 7 PRO 350 showed it could be up to 40 percent faster than the Intel Core Ultra 7 165H on the Cinebench R24 n-thread test. It's about delivering workstation-level performance in thin and light laptops, honestly.

Gaming: High-fidelity graphics and semi-custom silicon for major consoles

The Gaming segment is showing explosive growth, primarily driven by the semi-custom silicon deals for major consoles, which you can see in the financial results. Gaming revenue in Q3 2025 was $1.3 billion, a massive 181% increase year-over-year. This huge jump is attributed to higher semi-custom revenue and strong demand for Radeon gaming GPUs. The value proposition is delivering high-fidelity graphics experiences, both through the discrete Radeon GPUs and the guaranteed, high-volume business from console partners. The combined Client and Gaming segment revenue was $4.0 billion, up 73% year-over-year.

Embedded: Flexible, adaptive compute for industrial and aerospace markets

The Embedded segment is the one area that saw a contraction in Q3 2025, with revenue at $857 million, marking an 8% decline year-over-year. However, the value proposition remains strong for specialized, long-lifecycle markets. Advanced Micro Devices, Inc. (AMD) is pushing flexible, adaptive compute solutions here. You're seeing new product introductions like the EPYC Embedded 4005 Series for secure networking and the Versal RF Series for aerospace and defense applications. The goal is providing high performance-per-watt and low-latency compute for automation and machine-vision systems, even if the current demand environment is mixed.

Here's a quick look at the segment financial snapshot from Q3 2025:

Business Segment	Q3 2025 Revenue (USD)	Year-over-Year Growth	Key Driver/Status
Data Center	$4.3 billion	Up 22%	5th Gen EPYC and Instinct MI350 GPU demand
Client	$2.8 billion	Up 46%	Record Ryzen processor sales
Gaming	$1.3 billion	Up 181%	Higher semi-custom revenue
Embedded	$857 million	Down 8%	Mixed demand environment

The Client and Gaming revenue combined for $4.0 billion, showing a 73% year-over-year increase. This portfolio approach helps balance the cyclical nature of the PC market with the massive, sustained demand from AI infrastructure.

Advanced Micro Devices, Inc. (AMD) - Canvas Business Model: Customer Relationships

You're looking at how Advanced Micro Devices, Inc. (AMD) manages the relationships with its diverse customer base, which is critical given its high-stakes positioning in data center AI and console markets. It's not one-size-fits-all; the approach shifts dramatically from a cloud giant to a retail shopper.

Dedicated engineering support for hyperscale and enterprise deployments

For the hyperscale and enterprise segments, the relationship is deep and platform-focused. Advanced Micro Devices, Inc. (AMD) is moving beyond just selling components to offering full-stack solutions, like the 'Helios' AI server rack, to simplify deployment for cloud providers. This dedicated support is essential for winning large, recurring contracts for EPYC processors and Instinct GPUs.

The results of this focus are showing up in the financials. For instance, in the first quarter of 2025, the Data Center segment revenue reached $3.7 billion, marking a 57% year-over-year increase. This growth was fueled by strong demand from these large customers, with EPYC enterprise adoption tripling year-over-year as of early 2025. Looking ahead, Advanced Micro Devices, Inc. (AMD) expects to achieve more than 50% server CPU revenue market share. Furthermore, the company projects its data center division will see a 60% annual growth rate over the next three to five years.

Here's a quick look at the segment performance driving these relationships:

Metric	Value (Q1 2025 or Latest Available)	Context/Driver
Data Center Revenue	$3.7 billion	Q1 2025 revenue, up 57% YOY.
Data Center AI Revenue (FY 2024)	Exceeded $5 billion	Full-year 2024 figure.
Server CPU Market Share Target	More than 50%	Long-term goal for server CPU revenue share.
Commercial Systems Wins (Q1 2025)	80% increase from 2024	Number of AMD-powered commercial systems from major OEMs.

Strategic, long-term semi-custom contracts with console makers

The semi-custom business is built on multi-year, high-volume agreements, primarily with console makers. These contracts provide a predictable revenue base, even when the discrete GPU market fluctuates. You see the impact of these deals showing up in the Gaming segment revenue.

For example, in the third quarter of 2025, Gaming revenue was $1.3 billion, which was up a stunning 181% year-over-year, directly attributed to higher semi-custom revenue. While console channel inventories normalized in the first half of 2025, the long-term pipeline is massive. Advanced Micro Devices, Inc. (AMD) has reported over $45 billion in total custom chip design win content, with revenue expected to start ramping in 2026 across sectors including aerospace, defense, automotive data centers, and communications.

Standardized retail and channel support for consumer products

For the Client and retail side, Advanced Micro Devices, Inc. (AMD) relies on a structured channel to push Ryzen processors through OEMs and system builders. The company recently restructured its global channel strategy into the AMD Partner Network (APN) in October 2025, consolidating sales training and marketing resources. This network engages 520 partners and has a committed 40% year-on-year increase in channel investment.

This channel support is translating into client segment growth. Client revenue reached a record $2.8 billion in the third quarter of 2025. In the first quarter of 2025, Ryzen Pro PC sell-through grew more than 30% year-over-year, supported by that 80% increase in the number of AMD-powered commercial systems from partners like HP, Lenovo, Dell, and ASUS compared to 2024. Advanced Micro Devices, Inc. (AMD) is on a path to exceed 40% client revenue market share.

Developer community engagement for the ROCm software stack

Engaging the developer community through the ROCm software stack is a non-revenue-generating but absolutely crucial relationship for winning the AI workload war. Advanced Micro Devices, Inc. (AMD) is focused on making ROCm an open, accessible alternative to proprietary stacks. The company accelerated its release cadence, with ROCm 7 becoming generally accessible in the third quarter of 2025.

This latest version shows tangible developer benefits:

• ROCm 7 supports over 2 million Hugging Face models, including LLaMA 4 and Gemma 3, from day one.
• It offers 3.5X better inference capability and 3X better training prowess compared to the ROCm 6 release.
• The platform is expanding beyond the data center, with support for PCs using Radeon GPUs and Ryzen AI APUs expected to be widely available in the second half of 2025.

The company is making internal tools available to the community and actively listening to feedback to build this ecosystem. Finance: draft 13-week cash view by Friday.

Advanced Micro Devices, Inc. (AMD) - Canvas Business Model: Channels

You're looking at how Advanced Micro Devices, Inc. (AMD) gets its silicon from the fab to the end-user, which is a complex, multi-pronged approach that shifted heavily toward the data center in late 2025. The channel strategy is defined by high-volume, direct relationships at the top end and broad distribution for the consumer side.

The Data Center segment, which is the primary channel for Hyperscale Cloud Service Providers (CSPs), accounted for 47% of Advanced Micro Devices, Inc. (AMD)'s total revenues in the third quarter of 2025, bringing in $4.3 billion in that period. This direct engagement is critical for selling EPYC CPUs and Instinct GPUs. For instance, in Q3 2025, hyperscalers like Google, Microsoft Azure, and Alibaba launched over 160 EPYC-powered instances, building on a global base of more than 1,350 public EPYC cloud instances. This channel prioritizes large, strategic deals for next-generation accelerators.

Here's a quick look at how the major revenue segments channeled their sales in the latest reported quarter:

Segment	Q3 2025 Revenue (Approximate)	Year-over-Year Growth (Q3 2025 vs Q3 2024)	Percentage of Total Revenue (Q3 2025)
Data Center	$4.3 billion	22.3%	47%
Client & Gaming (Consolidated)	Around $4 billion	72.8%	43.8%
Embedded	Declined 8% (Implied from segment performance)	Not explicitly stated for Q3 2025 YoY	Implied $\approx 9.2\%$

For Original Equipment Manufacturers (OEMs) like Dell and HP, the channel is the Client segment, which saw record performance driven by Ryzen processors. Client revenue alone hit $2.75 billion in Q3 2025, making up 29.7% of total revenue. This channel saw significant momentum, with Client revenue in Q2 2025 already up 67% year-over-year to $2.5 billion. You see this direct OEM relationship materialize in announcements, such as the new strategic collaboration with Dell to offer a full portfolio of commercial PCs powered by Ryzen Pro processors. Furthermore, Advanced Micro Devices, Inc. (AMD) gained another 1% of the client CPU market share in Q3 2025.

The distribution and retail channel for consumer CPUs and GPUs is where the enthusiast market shows its strength. While the consolidated Client & Gaming segment is large, the retail floor tells a specific story, especially for high-end gaming GPUs and X3D processors. The Gaming revenue component in Q3 2025 surged to $1.3 billion, a massive 181% year-over-year increase. The retail preference is stark when looking at specific point-of-sale data; for example, in Week 47, a major German retailer showed Advanced Micro Devices, Inc. (AMD) selling almost 2260 units compared to Intel's roughly 220 units, giving Advanced Micro Devices, Inc. (AMD) a 91% selling share for that period.

Direct engagement with large enterprise and government High-Performance Computing (HPC) customers is a subset of the Data Center channel, but it's worth noting the specific product focus and future commitments:

• The Q3 2025 growth in Data Center was driven by fifth-gen EPYC processors and the Instinct MI350 series GPUs.
• Oracle Cloud Infrastructure (OCI) announced it will offer the first publicly available Instinct MI350 series accelerators.
• Advanced Micro Devices, Inc. (AMD) has a major infrastructure deal with OpenAI, which will deploy 6 gigawatts of Advanced Micro Devices, Inc. (AMD) GPUs, starting with 1 gigawatt of Instinct MI450 GPUs in the second half of 2026.
• The company expects Q4 2025 revenue to be approximately $9.6 billion at the midpoint, implying roughly 25% year-over-year growth, which will rely on continued success in these high-value channels.

Finance: draft Q4 2025 cash flow projection by next Tuesday.

Advanced Micro Devices, Inc. (AMD) - Canvas Business Model: Customer Segments

You're looking at the core buyers for Advanced Micro Devices, Inc. (AMD) as of late 2025. This isn't just about selling chips; it's about powering the biggest shifts in computing, from AI training to the next generation of home entertainment. Honestly, the numbers from the third quarter of 2025 show a clear pivot toward data center and AI workloads.

For Q3 2025, Advanced Micro Devices, Inc. reported a record total revenue of $9.2 billion. This revenue base is carved up across the key customer groups you listed, which the company reports through its formal segments.

Hyperscale Cloud Providers (AI training and inference) and Enterprise and HPC customers for server and data center compute

These two groups are largely captured within the Data Center segment, which is the powerhouse for Advanced Micro Devices, Inc. right now. In Q3 2025, the Data Center segment generated $4.3 billion in revenue, which accounted for 47% of the company's total revenue for the quarter. This was a 22% year-over-year increase. The growth here is fueled by both AMD EPYC processors for general server and High-Performance Computing (HPC) workloads and the AMD Instinct GPUs for AI training and inference.

The AI focus is intense. Advanced Micro Devices, Inc. announced a strategic partnership with OpenAI to deploy 6 gigawatts of AMD GPUs for their next-generation AI infrastructure, with the initial 1-gigawatt deployment of AMD Instinct MI450 GPUs scheduled to start in the second half of 2026. Looking forward, the company has set a long-term goal for its Data Center business to deliver a greater than 35% revenue Compound Annual Growth Rate (CAGR), with the AI-specific revenue targeted for a CAGR of more than 80%. Advanced Micro Devices, Inc. also expects to achieve more than 50% server CPU revenue market share long-term.

PC OEMs and DIY enthusiasts (consumer and commercial)

These customers fall under the Client portion of the Client and Gaming segment. For Q3 2025, the Client revenue specifically hit a record $2.8 billion, marking a 46% year-over-year jump. This is driven by strong demand for the latest AMD Ryzen processors, especially among DIY enthusiasts and Original Equipment Manufacturers (OEMs) building commercial and consumer PCs. The entire Client and Gaming segment revenue was $4 billion in Q3 2025, up 73% year-over-year. The long-term target for Client revenue CAGR is greater than 10%, with Advanced Micro Devices, Inc. aiming to exceed 40% client revenue market share.

Gaming Console Manufacturers (Sony, Microsoft)

Revenue from these manufacturers is reported under the Gaming portion of the Client and Gaming segment, primarily through semi-custom System-on-Chips (SoCs). Gaming revenue in Q3 2025 was $1.3 billion, which was a massive 181% year-over-year increase. This surge is attributed to higher semi-custom revenue as inventory normalized and customers prepared for the holiday season. For context, in Q1 2025, Gaming revenue was down 30% year-over-year, showing the cyclical nature of this business before the Q3 rebound. The long-term CAGR target for the combined Client and Gaming business is greater than 10%.

Industrial, Communications, and Aerospace (Embedded)

The Embedded segment serves these specialized markets with CPUs, GPUs, FPGAs, and adaptive SoCs. This segment saw revenue of $857 million in Q3 2025. This figure represents an 8% year-over-year decline, as end-market demand remained mixed during that quarter. Advanced Micro Devices, Inc. is expanding its offerings here with products like the AMD EPYC Embedded 9005 Series processors and Versal AI Edge XQRVE2302 adaptive SoCs. The long-term growth target for the Embedded segment revenue CAGR is greater than 10%, with a goal to exceed 70% revenue market share in adaptive computing.

Here's a quick math summary of the Q3 2025 segment revenue breakdown:

Customer Segment Group (Reported Segment)	Q3 2025 Revenue (USD)	Year-over-Year Change
Hyperscale/Enterprise/HPC (Data Center)	$4.3 billion	Up 22%
PC/DIY & Gaming Consoles (Client and Gaming)	$4.0 billion	Up 73%
Industrial/Comms/Aerospace (Embedded)	$857 million	Down 8%

The mix clearly shows the company is leaning heavily on the Data Center segment, which is now nearly half of the total business. Still, the massive growth in the Gaming portion of the Client and Gaming segment is a significant near-term driver.

• Data Center Revenue (Q3 2025): $4.3 billion.
• Client Revenue (Q3 2025): Record $2.8 billion.
• Gaming Revenue (Q3 2025): $1.3 billion.
• Embedded Revenue (Q3 2025): $857 million.
• Total Revenue (Q3 2025): $9.246 billion.

Finance: draft 13-week cash view by Friday.

Advanced Micro Devices, Inc. (AMD) - Canvas Business Model: Cost Structure

The Cost Structure for Advanced Micro Devices, Inc. (AMD) is heavily weighted toward the direct costs of manufacturing its complex semiconductor products and the substantial, ongoing investment required to maintain technological leadership, especially in the Artificial Intelligence (AI) space.

High Cost of Goods Sold (COGS) due to reliance on external foundries.

Your reliance on external foundries, like Taiwan Semiconductor Manufacturing Company (TSMC), means that the Cost of Goods Sold (COGS) is a dominant cost component. This is inherent in the fabless model, where you pay for wafer fabrication, packaging, and testing services. For instance, in the first quarter of 2025, GAAP Gross Margin stood at 50% on revenue of $7.438 billion, meaning COGS represented 50% of revenue, or approximately $3.719 billion. By the third quarter of 2025, as product mix shifted and perhaps due to increased complexity or volume, the GAAP Gross Margin improved to 52% on record revenue of $9.246 billion. This high direct cost base means that even small fluctuations in manufacturing costs or pricing pressure from competitors directly impact your bottom line.

Significant R&D investment; Q1 2025 operating expenses were $2.2 billion.

To stay competitive against rivals, R&D spending remains massive and non-negotiable. You are pouring capital into next-generation CPU and GPU architectures. While GAAP Operating Expenses for Q1 2025 were reported at $2,930 million, the management commentary from the Q1 2025 earnings call specifically highlighted operating expenses at $2.2 billion, noting this was an increase of 28% year-over-year, driven by aggressive investment in R&D and go-to-market activities. Looking at the trailing twelve months (TTM) ending September 2025, your R&D expenses peaked at $6.386 billion. This aggressive spending is necessary to fuel the Data Center segment growth, which is key to your valuation.

Capital expenditure for AI infrastructure and system-level integration.

Capital expenditure (CapEx) reflects the investment in equipment, tooling, and potentially internal infrastructure to support system-level integration and future product ramps. You are seeing CapEx increase significantly compared to the prior year. For the quarter ending September 30, 2025, CapEx was reported at $258 million. This compares to $212 million in Q1 2025 and $208 million in Q4 2024. The TTM annual CapEx as of Q3 2025 reached $960 million, a substantial jump from the $636 million spent in the full year 2024. This spending supports the increasing complexity of integrating accelerators and CPUs into complete system solutions for customers.

Go-to-market and sales expenses to gain market share.

Gaining share, particularly in the lucrative Data Center market against entrenched players, requires significant spending on sales, marketing, and channel development. These costs fall under Operating Expenses, alongside R&D. The Q1 2025 commentary explicitly cited investment in go-to-market activities as a driver for the $2.2 billion operating expense figure. Furthermore, the GAAP Operating Expenses for Q3 2025 rose to $3,510 million, up 30% year-over-year, reflecting the ongoing need to support the expanding sales footprint across all segments, including the rapidly scaling AI business.

Here's a quick look at some key cost and expense metrics from the recent quarters:

Metric (GAAP)	Q1 2025 Amount (Millions USD)	Q3 2025 Amount (Millions USD)	TTM Ending Q3 2025 Amount (Millions USD)
Revenue	$7,438	$9,246	N/A
Gross Profit	$3,736	$4,780	N/A
Operating Expenses	$2,930	$3,510	N/A
Capital Expenditure	$212 (Q1 only)	$258 (Q3 only)	$960

The cost profile is characterized by high variable costs tied to manufacturing and high fixed/semi-fixed costs related to R&D and market expansion. You need to manage the foundry relationship carefully to control COGS.

• GAAP Gross Margin ranged from 50% (Q1 2025) to 52% (Q3 2025).
• Non-GAAP Operating Income reached $2.2 billion in Q3 2025.
• R&D expenses hit a 5-year peak in the LTM ending September 2025 at $6.386 billion.
• Q2 2025 results included an inventory charge of approximately $800 million related to export controls, which impacted GAAP margins but was excluded from non-GAAP figures.

Finance: draft 13-week cash view by Friday.

Advanced Micro Devices, Inc. (AMD) - Canvas Business Model: Revenue Streams

You're looking at the hard numbers for Advanced Micro Devices, Inc. (AMD) revenue streams as of the third quarter of 2025. Here's the quick math on where the money came from in that period.

Segment	Q3 2025 Revenue Amount	Year-over-Year Growth
Data Center (EPYC, Instinct)	$4.3 billion	Up 22%
Client (Ryzen CPUs)	$2.8 billion	Up 46%
Gaming (Radeon, Semi-Custom)	$1.3 billion	Up 181%
Embedded (Xilinx)	$857 million	Down 8%

The total reported GAAP revenue for Advanced Micro Devices, Inc. (AMD) in Q3 2025 was a record $9.2 billion, representing a 36% increase year-over-year.

The revenue breakdown by segment for Q3 2025 shows the following contributions:

• Data Center segment (EPYC, Instinct): $4.3 billion.
• Client segment (Ryzen CPUs): $2.8 billion.
• Gaming segment (Radeon, Semi-Custom): $1.3 billion.
• Embedded segment (Xilinx): $857 million.
• Licensing and software services related to the ROCm defintely ecosystem.