The narrative is everywhere. Cloud vendors say it. Serverless platforms echo it: "Print servers are legacy. Eliminate them."
It's compelling. Clean. A clear direction. And for most enterprises, it's catastrophically wrong.
The problem isn't that print servers are bad. The problem is that the advice treats all print workloads as identical—as if Epic EMR printing, telehealth printing, field office printing, and compliance printing all have the same requirements. They don't.
This is what honest enterprise architecture looks like: mission-critical print needs what servers provide. Distributed print needs what direct IP provides. Both exist in your organization simultaneously.
Forcing one approach on both is negligence.
Vendors want you to choose. Cloud: "Servers are dead." Serverless: "No infrastructure to manage." Both are selling a lie.
The lie is that one approach works for everything.
The truth: Different workloads need different infrastructure.
Mission-critical print—Epic EMR, patient wristbands, classified documents, financial compliance—needs:
Redundancy you control (not a vendor SLA)
Offline capability (internet down ≠ printing down)
Auditability (chain of custody for compliance)
Local failover (<30 second recovery)
Distributed print—telehealth, field offices, remote workers—need:
Simplicity (minimal overhead)
Cost optimization (no licensing sprawl)
Speed (low latency)
Ease of deployment
Stop asking "servers or serverless?" Ask: "What infrastructure does each workflow actually require?"
The answer is: both.
When you outsource print infrastructure to serverless vendors, you lose three critical things:
When you own infrastructure:
You control security policy
You control patch schedules
You control access and auditability
You decide your uptime standard
When you outsource:
Vendor controls everything
You wait for their patches (sometimes 3+ years for critical vulnerabilities)
One vendor outage = everyone on that platform fails simultaneously
You audit their attestations, not their actual infrastructure
Real example: One major serverless print vendor had 83 documented vulnerabilities (2021-2024), including cross-tenant data breaches and hardcoded AWS keys. Critical patches took 3+ years. You can't force faster remediation when the vendor owns it.
When you own infrastructure:
ISP down? Printing still works
Emergency operations continue
Hospitals can print patient wristbands
Manufacturing can print shift handoff documents
Government can print classified material
When you outsource:
No internet = no printing
Mission-critical workflows fail completely
You're hostage to your ISP's uptime
This isn't theoretical. Gartner and Microsoft both explicitly validated this in 2025: Organizations need both approaches, managed together. Microsoft even updated Azure Virtual Desktop to support local, on-premises deployment. Translation: Cloud-only was a failed promise.
When you own infrastructure:
Complete chain of custody for PHI, financial data, classified material
You demonstrate control to regulators
Full forensic investigation capability
When you outsource:
PHI traverses public cloud (HIPAA violation)
You depend on vendor forensics
You carry the compliance risk; they carry the data
FedRAMP requires you to control the infrastructure—you can't outsource that
Here's where the math gets interesting. The serverless pricing model creates a structural disadvantage that gets worse as your organization grows.
Serverless vendors typically license by print queue, not by user. This means: Organizations start with an attractive per-queue price ($X per queue/month). But as they grow—adding branch offices, new departments, remote printing—they add more queues. Each new queue adds to the bill.
Concurrent user licensing charges based on simultaneous users, not infrastructure components. This means:
Whether you have 10 queues or 100 queues, the cost stays the same
Adding new print queues = $0 additional cost
Growth doesn't trigger licensing escalation
Queue consolidation provides cost savings (fewer users managing print = lower concurrent user count)
| Cost Factor | Serverless (Per-Queue) | Hybrid (Concurrent User) | Impact |
|---|---|---|---|
| Year 1 Appeal | Lower upfront cost per queue | Higher base cost | Serverless looks cheaper |
| Year 2: Add queues | Cost increases with each new queue | No increase | Hybrid advantage grows |
| Year 3: More growth | Escalating per-queue fees accumulate | Still flat | Gap widens significantly |
| Year 4-5: Mature org | Per-queue costs compound; subscription creep adds 15-25%/year | Predictable increases | Hybrid is dramatically cheaper |
| Vendor lock-in | Expensive to migrate away; locked into per-queue model | Easy to scale independently; no lock-in | Risk compounds |
Scenario: Organization grows from 50 queues to 80+ queues over 5 years
| Serverless (Per-Queue) | Concurrent User (Hybrid) |
|---|---|
| Year 1: 50 queues = baseline cost | Year 1: Fixed concurrent user cost |
| Year 2: +10 queues = baseline + increment | Year 2: Same cost (10 new queues = no charge) |
| Year 3: +15 queues = baseline + larger increment | Year 3: Same cost (15 new queues = no charge) |
| Year 4: +8 queues = baseline + additional increment | Year 4: Same cost (8 new queues = no charge) |
| Year 5: Total cost has escalated 40-60% | Year 5: Total cost unchanged |
| Outcome: Dramatically higher TCO, vendor lock-in difficult | Outcome: Predictable costs, scalable, no lock-in |
Serverless vendors know per-queue licensing creates escalating costs. So they add other charges:
Subscription creep: Features you didn't budget for, 20-25% annual increases (sometimes even more)
Transaction fees: Per-print or per-job charges
Storage/archive fees: Document retention and retrieval
API/integration fees: Connecting to other systems
Migration fees: When you finally try to leave
Traditional servers have similar overhead, but at least you own the infrastructure. Serverless locks you in while costs rise.
EMR/EHR printing isn't optional. When a clinician prints a medication order, a lab result, or a patient wristband, it has to print. Network down? ISP outage? Vendor cloud failure? Doesn't matter. Patient safety depends on it printing.
Why direct IP/serverless fails here:
No redundancy if the local connection drops
No offline capability if ISP is down
No failover if the print server is compromised
Compliance audit trail is fragmented
Vendor outage = ER can't print wristbands
No offline operations for emergency scenarios
PHI traverses public cloud (HIPAA violation)
You have zero control when it breaks
What works: Fortified, redundant on-premises infrastructure with automatic failover. You control uptime. You control security. You control compliance.
Classified networks require on-premises, air-gap capable infrastructure. Cloud vendors literally cannot serve this market. Serverless vendors can't even claim compliance.
You need servers. Hardened, controlled, on your network. This isn't debatable.
Audit trails for financial transactions require centralized, auditable infrastructure under your control. You can't outsource control of compliance documentation to a vendor and pass an audit.
To be fair: there are workloads where serverless or direct IP approaches work well.
Telehealth printing: Remote clinicians printing intake forms don't need Epic-grade redundancy. Direct IP is great.
Field office printing: Branches printing local materials don't need centralized control. Distributed infrastructure is the right call.
Remote worker printing: Home office users don't need mission-critical redundancy. Simplicity wins.
In these cases, direct IP works because:
Downtime is inconvenient, not dangerous
No compliance requirements
No offline operations needed
Users can retry or find alternatives
But here's the critical point: These workloads coexist with mission-critical ones in most enterprises.
You don't get to choose "eliminate servers." You need both, managed together.
Instead of "How do we eliminate servers?", ask:
1. Which workflows are mission-critical?
What happens if they fail?
Do they have compliance requirements?
Do they need offline capability?
2. Which workflows are distributed?
Can they tolerate occasional downtime?
Do they need simplicity over control?
3. Can one approach serve both?
If not, what's the cost of fragmentation across tools?
4. What's the real TCO?
Include subscription creep, vendor lock-in, downtime costs
Don't just compare year-1 numbers
5. What happens in an emergency?
ISP outage: Can you still print?
Vendor infrastructure failure: Are you down?
Security breach: Can you recover?
If the answer to any of these is "it depends on which workflow," then you need multiple approaches, unified.
Microsoft's November 2025 announcement: Organizations can now run Azure Virtual Desktop locally. Their message: "This new hybrid approach gives additional options for organizations wanting to bridge the gap between cloud and on-premises environments."
Translation: Cloud-only failed. Organizations need both.
Gartner's 2025 findings:
91% of organizations planning hybrid strategies
80% of print management initiatives include hybrid approaches
67% of CIOs prioritize cost optimization (which hybrid delivers)
Stop asking "How do we eliminate servers?"
Start asking "What's the right infrastructure for each workflow?"
Map your actual workloads
Mission-critical (needs centralized, redundant infrastructure)
Distributed (needs direct IP simplicity)
Calculate honest TCO
Don't just compare year-1 subscription costs
Include admin overhead, downtime, vendor lock-in, switching costs
Question the "eliminate servers" pitch
Ask which workflows they're optimizing for
Ask what happens to other workloads
Ask who bears the cost when their approach fails
Build for your actual requirements
Not vendor narratives
Not industry dogma
Your enterprise's real needs
"Eliminate print servers" is good marketing. It's simple. It's a clear direction.
But enterprise IT doesn't work that way. Mission-critical systems need what servers provide: control, redundancy, offline capability, auditability. Distributed systems need what direct IP provides: simplicity, cost-effectiveness, speed.
The question isn't whether to keep servers. It's whether you architect print infrastructure to match your actual workloads, or whether you force your workloads to fit a vendor's narrative.
Most enterprises are still doing the latter.
It's time to ask better questions.