Cold Spare Backups: The Underrated Lifeline in Times of Skyrocketing Hardware Prices
Rising prices, empty warehouses, delivery times stretching into months – anyone looking to buy server hardware today quickly feels as though they’re in the midst of a digital gold rush, with everyone mining for it at the same time. SSDs and RAM modules, in particular, are in greater demand than ever before. AI models are devouring resources, hyperscalers are buying up entire production lines, and small businesses are suddenly faced with the question: how are we supposed to keep our systems running when even standard components are becoming scarce? This is precisely where an approach comes into play that was long considered a ‘nice to have’, but is now almost essential for survival: cold spare backups. In other words, spare hardware that isn’t in use but sits on the shelf – ready to step in when the need arises. In the following article, we explain the current ‘component crisis’, why prices for RAM and SSDs are skyrocketing, and how you can arm yourself against this – and the next – hardware crisis with a few smart strategies.
AI boom = high resource consumption & component shortages
Anyone looking to buy server hardware at the moment will quickly realise that the market has changed. Spare parts and components that used to be readily available are suddenly in short supply – delivery times often stretch over weeks or even months. And the prices? Just as overpriced as petrol at the moment – everything is already barely affordable and is still rising merrily…
With high fuel costs, we know what’s behind them – but what about RAM, SSDs and the like? The main driver behind this trend is the global AI boom. Since chatbots, image generators and analytical models have found their way into businesses (and our homes) in record time, the demand for computing power has multiplied.
AI systems are real resource guzzlers – in addition to enormous amounts of water, electricity and raw materials, they also require an incredible amount of RAM lightning-fast SSDspowerful GPUs and scalable server architectures. AI: The Very Hungry Caterpillar. Large data centres therefore don’t just buy individual components – they buy entire production batches. And they do so before these even reach the open market.
AI – the sole culprit?
We live in a world where digitalisation is advancing at full speed. What once began with marvellous gadgets like PCs and smartphones, designed to make our daily lives easier, has long since developed into total dependence on technology. Whether it’s the weather, what to wear, lunch or entertainment – for everything, we reach for the nearest app or ask Dr Google. The fact that the big tech corporations can wage such battles at all is due not least to our unrelenting demand.
Streaming services are expanding their data centres to accommodate growing user numbers, cloud providers are opening new sites worldwide, telecommunications companies are scaling up their infrastructure, and gaming and content platforms require ever more storage and processing power. All these major corporations have one thing in common: they buy gigantic quantities of hardware – and are willing to pay almost any price to realise their projects.
For manufacturers, this means that virtually everything produced is already sold. The cash registers are ringing – but for the rest of the market, namely smaller businesses, this means one thing above all else: a severe shortage. Because, naturally, priority is given to the customers who order the most and most regularly.
SMEs and start-ups are sliding to the bottom of the food chain, and that is disastrous. In most cases, the components are simply no longer affordable, and even if the budget is available, that is by no means a guarantee that you will actually get the hardware. The gap between the overwhelming competition is growing ever wider – projects are delayed, systems are overloaded, outages become a real risk, and the digitalisation of small businesses, which is often already lagging behind, falls even further behind.
For SMEs, this can become a real danger: the market follows the demands of the ‘big players’, and those who fail to plan ahead will be left behind when the crunch comes.
Available, delivery time 1-3 days
Available, delivery time 1-3 days
How to protect yourself: Cold spare backups as IT’s silent life insurance
We are currently seeing that RAM and SSDs, which were still ‘off-the-shelf’ items just a few years ago, have become genuine luxury goods. Whereas 32 or 64 GB of RAM used to be more than enough for many workloads, today we are quickly talking about several hundred gigabytes being required for tasks such as AI training, data analysis or virtualisation. And the modules required for this are currently barely available – or affordable.
The same applies to enterprise SSDs, which are optimally designed for data centres with high write cycles and capacities. For small and medium-sized enterprises, which are already operating on tight budgets, the hardware market is thus becoming a real problem area. Prices are rising faster than one can plan for. Growth and swift action? Almost impossible.
Even supposedly ‘outdated’ hardware is suddenly becoming attractive again (and therefore expensive too). HDDs, which many had already written off, are making – yet another – comeback. Why? Because they are available and, compared to SSDs, are currently almost a bargain.
But how should one deal with the tight supply situation? Waiting and seeing is not an option – not while AI models are growing, cloud providers are expanding and hyperscalers are rapidly expanding their data centres. For small businesses, this means thinking strategically and not relying on spontaneous availability. Anyone who only orders hardware when something breaks down risks long downtimes. Delivery times of several weeks can be critical. It makes more sense to order an extra set of key components straight away – as a cold spare backup that sits on the shelf and can be deployed immediately in an emergency. This ensures you remain operational, even when the market is completely sold out.
Refurbished server parts are particularly cost-effective – and, above all, available for immediate delivery. Used enterprise servers, RAM modules or SSDs are not only significantly cheaper, but are also usually available for immediate delivery. Whilst new parts often take months to arrive, tested refurbished components are usually in stock and offer the same reliability as new goods.
Furthermore, those who rely on their own servers – whether for production or as cold spares – remain more independent than those who rely exclusively on cloud resources. There too, high demand is driving up prices, reducing capacity and prioritising large customers. With your own hardware, you retain control: over costs, availability and performance. And those who factor in a small set of critical components at the time of purchase, rather than reacting only when a failure occurs, remain capable of acting even during tight market conditions.
Forward-thinking and future-proof planning – the be-all and end-all of IT infrastructure
Extreme price surges and supply bottlenecks are no longer unpredictable phenomena – but a stark reality that hits us every few years. During the COVID-19 pandemic, it was the global chip crisis, triggered by lockdowns, production halts and logistical problems. At the same time, demand skyrocketed, primarily due to working from home and home schooling.
Trade wars, tariffs, export bans and sanctions are forcing manufacturers to restructure their supply chains. Prices are rising – this is more relevant than ever, but has also been the case since 2018. In 2017, we experienced GPU shortages for the first time and the longest lead times in over a decade, triggered by crypto mining.
Cold spare backups are therefore more than just a good idea. We also have further tips on how you can prepare for supply bottlenecks and avoid high prices:
15 tips for avoiding supply bottlenecks and high prices
- Extend hardware lifecycles
Replacing all hardware every two to three years just because that’s the norm? Instead, replace specific components to reduce the load on your systems. Also, set up a monitoring system to detect faults early and take action. - Use standardised hardware to remain flexible
The more ‘exotic’ the hardware, the harder it is to source. It’s better to rely on widely used (professional!) hardware, common RAM types and current standards. - Keep an eye on and follow market trends
Always keep an eye on tech developments in the daily news. Are demand and prices rising, or is availability falling? Track market trends (e.g. AI, mining, cloud expansion) – those who stay up to date can make purchases in good time. - Introduce internal hardware rotation
New servers are purchased and then deployed to the production environment. Older devices can be kept as cold spare backups or (if compatible with the new devices) used as a source of spare parts. - Plan for and stock spare parts in good time
Storage is running at 98% capacity, and the virtualisation servers are pushing the RAM modules to their absolute limit – not best practice. Those who do not buy ‘on the cheap’ remain more flexible whilst protecting themselves against failures. - Use the cloud only as a supplement, not as the sole infrastructure
Cloud solutions can help in the short term or serve as a backup. A hybrid approach is often more stable – both in terms of data security and price/availability. - Keep duplicate critical components
Not every server needs to sit unused on a shelf just in case. However, keeping power supplies, RAM, SSDs, fans or RAID controllers for critical systems in stock can be a real game-changer. - Maintain documentation & inventory
If you know what you have, you know what you need. A good overview prevents unnecessary purchases and saves time in an emergency. - Keep firmware & software up to date
Many failures are not caused by the hardware, but by outdated software. That’s why you should keep firmware and software up to date, as fewer failures mean less need for spare parts. - Opt for components with a longer service life
Professional-grade hardware is essential for servers operating under continuous 24/7 load. If the components are designed for a long service life, you can also use HDDs for archive or cold data if SSDs are unaffordable. - Collaborate with other SMEs
Sharing is caring: joint stock-keeping or the exchange of spare parts can reduce costs and bridge shortages. Important: ensure data security! - Opt for modular systems
Servers that can be easily expanded or upgraded are more flexible and durable. - Check maintenance contracts
Spare parts are often included. If necessary, have servers retrofitted with extended warranties. We also do this for third-party devices – find out more in our warranty terms and conditions. - Allow for a budget for ‘market stress’
Supply bottlenecks come in waves. A small financial buffer prevents you from having to make expensive purchases at the wrong moment. If you need new equipment, you can generate a new IT budget from your old hardware. How does it work? Find out here. - Opt for refurbished hardware
Available faster, reliable in use, tested and – even in times of supply shortages – many times cheaper than new equipment.
Not a ‘stopgap’, but the better alternative – especially in critical times
At a time when new server components cost more than many a used car, it is worth considering an option that has long been underestimated: refurbished hardware. What was once seen as a ‘stopgap’ is now, for many companies, the most strategically sound way to remain independent, flexible and cost-effective.
The biggest advantages are obvious: price and availability. Whilst new parts can take months to arrive, tested, refurbished components are usually available immediately – and at a fraction of the new price. For SMEs and start-ups that have to watch every penny, this is a real game-changer. Instead of pouring the budget into overpriced new goods, refurbished hardware allows you to build a stable, high-performance infrastructure without compromising on quality.
In addition to the benefit of greater sustainability – which, in times of rising energy prices and growing environmental demands, makes not only economic sense but also sends a strong message – there is also the advantage of predictability: Those who opt for refurbished hardware for cold spare backups and plan ahead will quickly have a set of spare SSDs or additional RAM modules at the ready, without having to incur significant costs should the worst come to the worst.
Refurbished hardware is therefore by no means the ‘cheap alternative’. It is the sensible alternative in a market that is becoming increasingly unpredictable for small businesses.
👉 Act now and stay competitive. We have large quantities of RAM modules, HDDs & SSDs, server components and complete systems in stock. And in our Sale section, you’ll also find plenty of bargains for your cold spare backup.
Other interesting topics:
