Why Small Businesses Struggle to Acquire Hardware

BIZTECH:  What's causing the supply chain crunch?

Wade: Increased demand and shifts in tariff policy have created a more uncertain environment around IT product availability. While supply chain disruptions have eased from the peak of pandemic-era challenges, many small businesses are still experiencing issues, particularly when it comes to technology investments. For most, the impact is indirect, showing up as higher costs for equipment, software and other IT purchases. Even if they are not directly tracking factors such as data center expansion or AI demand, those pressures are felt through rising prices on the products they rely on.

Sanders: The basic problem is being caused by a structural shift in how semiconductor capacity is being used. Several forces are stacking on top of each other. First, AI demand is consuming the most lucrative chip capacity. The explosive growth of AI — especially training and running large models — has created massive demand for graphics processing units and AI accelerators. Hyperscalers such as Amazon Web Services, Microsoft Azure and Google Cloud are getting it first. That leaves a shortage for average users, including consumers and small businesses. High-bandwidth memory and advanced chips use the same fabrication resources as other memory products; when those resources prioritize AI, less capacity is left for standard RAM used in PCs, which is what most of us use.

BIZTECH: Is this a short-term problem, or can we expect these challenges to continue indefinitely?

Nguyen: The supply chain pressure is likely to persist for the foreseeable future, driven largely by sustained demand tied to AI infrastructure. Memory constraints remain a key bottleneck, and current manufacturing buildouts are not expected to close the gap until at least the late 2020s, with some projections extending closer to 2030. While there is ongoing debate about whether the AI boom will normalize, the underlying demand for compute and storage continues to outpace supply. Absent a breakthrough that significantly improves efficiency at scale, the pressure on infrastructure and the resulting supply constraints are expected to remain.

Wade: The supply chain crunch is likely to be a medium-term challenge, driven by sustained AI investment and shifting trade policy, with supply and demand expected to take time to rebalance. In the meantime, while most businesses can still access the technology they need, they are facing higher costs for equipment and components, adding to broader cost pressures that must ultimately be absorbed or passed on to customers.

Sanders: This is not indefinite, in my opinion. There are forces that will eventually rebalance things. Massive global investment is underway, and as markets adjust, governments and companies are pouring money into capacity. The U.S. and European Union have created semiconductor incentives, and there are new fabrication resources opening across Asia and North America. This will increase total supply meaningfully by 2028. Another factor is that over time, technology improvements will reduce pressure. AI models will become more efficient, and new architectures will reduce hardware intensity. This helps stretch existing supply further.

REVIEW: Limited-time deals on hardware and peripherals for small businesses.

BIZTECH: What else would potentially ease the crunch?

Sanders: Easing this crunch comes down to either increasing supply meaningfully or reducing how fast AI is consuming it — basically, supply and demand. The biggest levers that could move the needle are things like new fabrication capacity becoming available. Major expansions from TSMC, Samsung and Micron are already underway. This is the single biggest long-term relief valve, but it will arrive slowly because it takes three to five years for them to build.

Another option for which I hold out hope is AI demand becoming more efficient. Right now, AI is extremely hardware- and energy-hungry, but that won’t stay constant. Improvements could include more efficient model architecture, better training techniques, data compression and reducing memory needs.

Wade: Companies are working to increase capacity, particularly in response to AI-driven demand for key components. But progress is constrained by changes in trade policy and the complexity of scaling semiconductor production. Even large domestic investments in chip manufacturing are highly specialized and slow to ramp, meaning they only address a portion of overall demand. How quickly the situation improves will depend on continued investment levels, policy stability and the pace at which new capacity can come online — likely over a multiyear horizon.

BIZTECH: From what you’ve heard, what kinds of challenges are businesses experiencing as a result?

Nguyen: One immediate impact is that many organizations are unable to execute on their AI ambitions because they simply cannot access enough infrastructure. Training and running even smaller, specialized models requires significant computing power, and most businesses are not getting priority access to GPUs or AI servers, which are largely allocated to hyperscalers and large AI providers.

Sanders: The most immediate impact is sticker shock. This is servers, PCs, storage, memory — they are getting more expensive. Even before this obvious crunch, corporate boards have been asking about the ROI. It is now much more acute. I am seeing companies buying directly from vendors that are experiencing budget overruns and delayed deployment cycles.

Wade: Uncertainty is a major challenge, affecting how small businesses plan and invest. Many are delaying technology purchases due to high costs and limited availability, choosing to hold onto cash until conditions improve. Unlike expenses such as wages or insurance, IT investments can be postponed, at least for a little while, leading some businesses to wait for prices to drop or supply to stabilize. At the same time, rising electricity costs tied to broader infrastructure build-outs are adding another layer of pressure on already-tight budgets.

DISCOVER: The benefits of centralized management of peripheral devices.

BIZTECH: Are small businesses especially vulnerable? Why?

Nguyen: Small businesses are especially vulnerable because they lack the purchasing power and long-term procurement agreements larger enterprises rely on to secure pricing and supply. Without that scale, they are more exposed to shortages, price volatility and delays, making it harder to plan investments or guarantee access to critical IT hardware when demand tightens.

Wade: The burden of navigating supply chain disruptions often falls on the owners of small businesses, who have limited time and resources to manage procurement challenges. Unlike larger organizations, they lack dedicated teams to source alternatives or negotiate pricing, making it harder to find affordable technology or secure supply. As a result, many fall behind larger competitors in accessing equipment and adapting to disruptions.

Sanders: Small companies also have limited access to scarce AI infrastructure. High-demand components, especially GPUs from NVIDIA, are often preallocated to the hyperscalers. Then, they are bundled into expensive enterprise solutions. Small businesses often struggle to even obtain the hardware and afford the minimum scale needed to use it effectively. This creates a technology access gap, not just a cost gap.

BIZTECH: What can organizations do to mitigate their risk? Are there strategies that can be helpful here?

Sanders: There’s no silver bullet, but organizations can reduce their exposure if they treat this as a strategic constraint, not just a procurement issue. The most effective response is the focused use of AI so the use itself is efficient, and the use of human ingenuity in a clever way. This means prioritizing high-value use cases. When resources are constrained, not every project deserves equal treatment.

Nguyen: Organizations can mitigate risk by extending the lifecycle of existing hardware and being more flexible about replacement cycles, particularly where cutting-edge performance isn’t required. Some are also turning to the secondary market, purchasing well-maintained used equipment to bridge gaps in supply. Better asset management and transparency around hardware condition can help organizations both buy and resell equipment more effectively, easing pressure on constrained procurement pipelines. 

BIZTECH: For businesses using a reseller, what should they be looking for in terms of their partner’s capabilities?

Sanders: In a constrained market, a reseller isn’t just a middleman anymore. The right partner can materially improve access, pricing and flexibility; the wrong one just passes along costs. This is very much in the center of supply chain management and procurement.

First, I think a strong supplier relationship is key. A good reseller should have deep ties with manufacturers — in this case, companies such as Dell, HP, Lenovo — and ideally also upstream chip ecosystem visibility; for example, awareness of constraints from the NVIDIA supply. Another issue to consider is the resellers’ inventory access, not just order-taking.

Nguyen: Businesses should look for reseller partners that can simplify procurement and provide clear, proactive communication about availability and alternatives. Strong partners help navigate constraints by identifying viable configurations, flagging potential supply risks early and offering flexibility through options like refurbished or leased equipment. That combination of transparency and adaptability becomes critical when standard supply channels are constrained.