Reducing Compute Inventory

Most compute services now rely upon servers organized into resource groups.  The cost of compute services can be substantially reduced using four basic inventory management strategies.

  • Add capacity to groups incrementally and more frequently.
  • Maximize use of groups (i.e., optimize workload placement and reservations).
  • Appropriately compose each group's configuration.
  • Optimize server refresh times and rates of each group.

The first two strategies broadly serve to increase server utilization, whereas the next two minimize a group's cost basis.  Ravello minimizes cost basis by solving the optimal rate of refresh in real-time — every time capacity is modified, planned, or forecast. This keeps server inventory balances at minimally necessary levels on an ongoing basis, which reduces far more expensive compute inventory and operating costs.

Let Ravello show you a simple, scalable way to use better analytics to align your supply-chain, reduce unnecessary over-provisioning, lower costs, and increase your team's capacity to innovate.

Predictably Lowest Total Costs

The basic building block of digital services starts with a semiconductor.  That drives server inventory, which then drives even more, higher cost compute inventory across the enterprise.

The quantity of servers forecast for tomorrow determines today’s compute inventory cost — from datacenter compute space, power requirements, rack counts, software licenses, and support agreements.  Yet tomorrow’s semiconductors will be significantly “smaller, faster, cheaper.”

Ravello enables your team to "look forward" and quickly determine today's and tomorrow's optimal server inventory levels, appropriately align your investments, and free up limited resources your team can better use to deliver more innovative services.

Capturing Cloud Economics

Chips remain the de facto source for reliable productivity gains, and “smaller, faster” (not cheaper) is the big engine driving cloud economics.

Server performance generally increases at higher rates than price reduction — and is more valuable.   "Smaller, faster" can reduce far larger costs in facilities, power, and software fees for enterprise IT, enabling more innovation.  How long this utility lasts is a server’s economic-life.

That differs from utility indicators like functional-life (wear & tear), technological-life (obsolescence) or accounting-life (depreciation) that turn into fixed use periods and feed current TCO models.

In contrast, inventory management using economic-life is dynamic, market-oriented, and derived from mathematical optimization – as many industries already manage physical inventory, or how other emerging IT tools for "virtual inventory" solve noisy neighbor and workload placement.

Let Ravello show you how it works.

Dynamic Digital Factories

Just as virtualization ushered in new practices and opportunities, so too advances in semiconductors are ushering in fundamental changes to physical inventory management.

Now, "smaller, faster, cheaper" actually shortens server economic-life.  Even modest increases in chip performance can drive significant savings and innovation across the compute inventory estate.

From plant & equipment, power usage, and distinct server configurations, to the vast and increasing stock of software licenses and support contracts, which are correlated to chip performance and span DevOps and business units — the cost and utility of different compute services varies greatly.

Ravello provides the tools so your team can easily plan and coordinate the quickening pace of change — managing server refresh at dynamic times and rates — and  capturing more savings for each type of compute service you manage.

Let Ravello demonstrate how your team can quickly improve upon your current fixed-time server refresh practices, and generate even more value from your DevOps initiatives.