GPU market news,
curated for professionals.
The GPU aftermarket moves fast. We track technology releases, secondary market developments, enterprise procurement shifts, and ITAD industry news — and surface what matters most to buyers, sellers, and operators.
GPU Remarketing Matters: Why “Asset Hugging” is Redefining Used Enterprise GPU Values
As the industry transitions into the Blackwell (B200) production cycle, a peculiar phenomenon has emerged in the secondary hardware market. Traditionally, the arrival of a new semiconductor architecture triggers a rapid liquidation of the preceding generation. However, the anticipated flood of NVIDIA Hopper (H100) and Ampere (A100) units into the remarketing channel hasn't materialized according...
Read Article →Networking Evolution: 1.6T HSIO Adoption Impacts GPU Cluster Longevity
The rapid transition toward 1.6T High-Speed I/O (HSIO) is fundamentally altering the depreciation curves and operational viability of current-generation AI infrastructure. Recent benchmarking indicates that Blackwell B200 clusters limited to legacy 800G networking are experiencing significant throughput bottlenecks during large-scale Mixture of Experts (MoE) training, where high-bandwidth interconnects are critical for minimizing latency in dynamic...
Read Article →GPU Financing: Lenders Increase Collateral Requirements for H100-Backed Loans
Global data center hub reports indicate a significant tightening in AI infrastructure financing, with lenders now requiring 30% more collateral for H100-backed loans compared to Q2 2025. This shift reflects a maturing secondary market where the initial scarcity-driven "liquidity premium" of NVIDIA’s Hopper architecture is being replaced by standard enterprise depreciation models and more conservative...
Read Article →Blackwell Supply Chain Stabilizes: Yields Hit 90%
Intelligence from the TSMC ecosystem indicates that CoWoS (Chip-on-Wafer-on-Substrate) yields for the B300 series have stabilized at approximately 90%, a critical manufacturing milestone that signals an imminent shift in global compute availability. As fabrication efficiencies at the AP6 facility reach these high-water marks, the persistent scarcity premium associated with Blackwell-generation hardware is projected to evaporate...
Read Article →Looking For Secondary Market GPU Pricing? Here Are 10 Things You Should Know About Current Liquidity
The secondary market for high-performance compute has evolved from a fragmented collection of resellers into a sophisticated B2B ecosystem. As of April 2026, the transition from Hopper-based architectures to mass-scale Blackwell deployment has created a unique liquidity window for enterprise-grade GPUs. For data center operators and institutional investors, understanding the technical and economic nuances of...
Read Article →10 Things to Know About Today’s GPU Market
GPU pricing is no longer driven by age alone. Secondary-market value now moves on technical configuration, deployment fit, and current supply conditions rather than a simple depreciation schedule. HBM capacity and generation materially affect resale value. Memory stack configuration remains a primary pricing variable for accelerators used in training and inference environments with high bandwidth...
Read Article →OpenAI’s $122B Surge
OpenAI’s $122 billion raise and 10GW Project Stargate buildout establish a new infrastructure standard: AI-native facilities built around liquid cooling, high-voltage power architecture, and 800G+ networking now define the performance and valuation baseline, while hardware refresh cycles compress toward 18 months and force faster repricing of GPUs, racks, optics, and supporting gear across the secondary...
Read Article →CoreWeave’s $8.5B Deal
CoreWeave’s $8.5 billion DDTL 4.0 shows that GPUs are now being underwritten as investment-grade infrastructure assets, with a credible recovery floor that matters to lenders such as Blackstone. The core collateral is no longer abstract AI capacity but the physical stack itself: H100 and B200 clusters, plus the 800G and 1.6T networking, optics, and high-speed...
Read Article →Hidden Gold: Networking Stack Value
High-performance networking is no longer a rounding error in AI infrastructure valuations: in dense H100, H200, and B200 clusters, the fabric layer, including 800G optics, InfiniBand switching, and SmartNICs/DPUs, now accounts for roughly 20% to 30% of total cluster value, yet many ITAD and decommissioning workflows still push these assets into bulk disposal streams that...
Read Article →Security Risks in GPU Disposal
In GPU decommissioning, the primary security risk is no longer limited to SSDs. In H100 and B200-class nodes, sensitive data can persist in High Bandwidth Memory (HBM) and adjacent GPU memory structures after standard software wipes or incomplete reset procedures, while VBIOS and BMC firmware layers introduce an additional persistence and supply-chain integrity risk if...
Read Article →Remarketing > Recycling
For data center operators, remarketing mid-life GPUs such as the NVIDIA A100 is increasingly a fiduciary duty rather than an optional recovery exercise: inference demand has pushed secondary-market resale values up by roughly 300%, and treating still-deployable accelerators as zero-value ITAD or recycling output destroys capital that can instead fund next-generation H100 and B200 deployments....
Read Article →The Power Wall
Blackwell-era 120kW racks have created a practical "Power Wall" for operators built around legacy 10kW–20kW infrastructure: once floor space, power delivery, and liquid-cooling capacity are repriced against GB200 and other next-generation deployments, many A100 and V100 clusters no longer justify their footprint even if they remain operational. The issue is no longer just hardware age;...
Read Article →NVIDIA GTC 2026: The Vera Rubin Shift
NVIDIA’s launch of Vera Rubin and the NVL72 platform resets the top end of AI infrastructure pricing, but the bigger takeaway is supply-chain pressure. The move toward tightly integrated rack-scale systems built around Rubin GPUs, the Vera CPU, BlueField-4 DPUs, and higher-speed networking raises the capital threshold for deployment and narrows the list of operators...
Read Article →The 120kW Barrier
Blackwell-class deployments have pushed rack density into the 120kW range, shifting the primary constraint in AI buildouts away from GPU availability and toward electrical and thermal infrastructure. The limiting factors are now transformer lead times, switchgear availability, high-density power distribution, and the cooling retrofits required to support NVL72-scale systems with 800G and 1.6T networking. In...
Read Article →The Era of Agentic Infrastructure
Market Shift: From Blackwell to Vera Rubin March 31’s core market signal is that NVIDIA’s roadmap is moving the center of competition from accelerator selection to full-rack architecture. Blackwell already pushed the market toward integrated GPU, CPU, memory, networking, and thermal design decisions at the system level. Vera Rubin extends that shift further. The practical...
Read Article →The Shift to AI-Native Infrastructure
The first week of April 2026 has solidified a fundamental transition in the digital infrastructure supply chain. We are moving past the "experimental" phase of AI cluster deployment into a regime defined by institutional-grade asset management and rigorous hardware-layer economics. For businesses financing, using, and remarketing GPU-based systems, the primary directive is no longer just...
Read Article →The 120kW Barrier: How Infrastructure Bottlenecks are Inflating Existing GPU Capacity Value
Date: March 26, 2026Category: Market and Pricing Blackwell-class deployments have pushed rack density into the 120kW range, shifting the primary constraint in AI buildouts away from GPU availability and toward electrical and thermal infrastructure. The limiting factors are now transformer lead times, switchgear availability, high-density power distribution, and the cooling retrofits required to support NVL72-scale...
Read Article →From Bare Metal to VMs: Liquidity Shifts in the Enterprise GPU Secondary Market
Date: March 31, 2026Category: Market and Pricing Liquid cooling is becoming a gating factor in the secondary GPU market, not just a deployment preference. As H100- and B200-class systems move into higher-density racks with 800G networking, resale value is increasingly tied to whether a platform can be redeployed into modern thermal and power envelopes rather...
Read Article →NVIDIA GTC 2026: Vera Rubin and the Next Supply Constraint
NVIDIA’s GTC 2026 launch of Vera Rubin and the NVL72 platform resets the top end of AI infrastructure pricing, but the bigger takeaway is supply-chain pressure. The move toward tightly integrated rack-scale systems built around Rubin GPUs, the Vera CPU, BlueField-4 DPUs, and higher-speed networking raises the capital threshold for deployment and narrows the list...
Read Article →Hidden Gold: Why Your Networking Stack is Worth as Much as Your GPUs
Category: Market and Pricing The Infrastructure Value Gap In the current AI hardware supercycle, the primary focus remains on the H100 and B200 compute modules. However, capital recovery analysis frequently overlooks the "passive" layer of the stack. High-speed interconnect (HSIO) components: specifically 800G optics, InfiniBand switching, and SmartNICs: now represent a massive percentage of total...
Read Article →Security Risks in GPU Disposal: Beyond the SSD
In the standard IT Asset Disposition (ITAD) workflow, the focus is almost exclusively on NAND flash and mechanical storage. However, as the industry transitions into the AI hardware supercycle, the security perimeter has shifted to the compute layer. Enterprise-grade GPUs, specifically the NVIDIA H100 and upcoming B200 platforms, represent a significant data persistence risk that...
Read Article →Remarketing > Recycling: The New Fiduciary Duty for Data Centers
Category: Market and Pricing The traditional approach to data center decommissioning: relying on zero-value IT Asset Disposition (ITAD) or simple recycling: is no longer financially defensible in the AI hardware supercycle. As compute requirements shift from massive training clusters to high-volume inference production, the capital recovery potential of mid-life GPUs has transformed asset disposition into...
Read Article →The “Power Wall”: Why 120kW Racks are Killing Legacy Clusters
The compute landscape is hitting a physical limit. As Blackwell (GB200) architectures and 120kW+ rack densities become the baseline for frontier model training, the "Power Wall" is forcing a radical re-evaluation of data center floor space. For enterprise operators and hyperscalers, the bottleneck is no longer just the silicon supply: it is the power delivery...
Read Article →OpenAI’s $122B Surge: Why AI-Native Infrastructure is the New Standard
The recent news that OpenAI has secured $122 billion in funding at an $852 billion valuation marks a definitive shift in the global compute supply chain. This is no longer merely a software play; it is a massive capital surge into physical infrastructure. Behind the headlines of valuation multiples lies a concrete reality: the rapid...
Read Article →CoreWeave’s $8.5B Deal: The Institutional Reclassification of GPU Assets
The close of CoreWeave’s $8.5 billion delayed-draw term loan (DDTL 4.0) on March 31, 2026, represents a structural shift in the capitalization of AI infrastructure. While the headline figure is significant, the actual narrative lies in the credit rating. For the first time in the history of specialized compute, a facility secured by GPU hardware...
Read Article →Looking for Better GPU Valuation? Here Are 10 Things You Should Know About Today’s Market
The acceleration of artificial intelligence has fundamentally altered the lifecycle of data center hardware. For years, IT Asset Disposition (ITAD) was a logistical exercise in secure destruction and commodity recovery. Today, as enterprise GPUs become the primary drivers of data center revenue, the recovery and remarketing of these assets have transitioned from a back-office task...
Read Article →The $600 Billion Sovereignty Shift: Why Regulated Industries Are Bringing AI In-House
We’re seeing a massive shift in how the biggest players think about AI infrastructure. McKinsey is estimating that 30% to 40% of AI spending — representing a $600 billion market by 2030 — is going to be influenced by sovereignty requirements. That is a lot of gravitational pull away from the public cloud. For a...
Read Article →The Digital Lettuce Problem: Why H100 Depreciation Is Accelerating
If you’re modeling H100s like traditional server hardware, you’re probably overestimating residual value. I call it the “Digital Lettuce” problem: the asset looks great on Day 1 — then it wilts fast. Why H100 depreciation is accelerating New GPU generations and platform refreshes compress resale windows Demand shifts quickly — workloads, power/cooling constraints, export rules,...
Read Article →GPU End-of-Life as an ESG Advantage: Turning Decommissioned Hardware Into Sustainability Data
ESG reporting is becoming more than just a box to tick. It’s a core metric that influences investor confidence and corporate reputation. For companies dealing with high-performance computing, managing GPU end-of-life is a critical piece of that puzzle. GPU recycling isn’t just about clearing out old server racks. It’s about turning decommissioned hardware into verifiable...
Read Article →Navigating ITAD in the AI Era: A Practical Guide to Secure GPU Decommissioning
The AI boom isn’t just a software revolution — it’s a massive hardware cycle that is fundamentally changing how we manage infrastructure. As businesses scale up to meet the demands of generative AI, the lifecycle of GPU-based servers and high-density networking gear is moving faster than ever. This brings a new set of challenges for...
Read Article →The 5-Year Hardware Lifecycle Is Dead: Inside the New 18-Month GPU Refresh Cycle
The 5-year hardware lifecycle is dead for AI infrastructure. In traditional enterprise IT, 60 months was the gold standard for ROI. With GPU clusters, that has changed. Rapid innovation cycles — moving from H100s to Blackwell architectures — have compressed the refresh cycle to just 18–36 months. Performance isn’t the only factor. Power efficiency and...
Read Article →