2 Gbps guaranteed bandwidth

2Gbps Dedicated Servers: The Second Gigabit as Reserve

Nobody architects a service around two gigabits. You buy the second one because the first is peaking and the next real design tier costs about double the whole machine.

X-ZoneServers 2 Gbps dedicated servers start at €259 per month for a guaranteed 2 Gbps port — 7 bare-metal configurations with up to 648 TB of monthly transfer across 12 EU and US datacenter locations. The port is effectively unmetered: hold the full committed rate around the clock, with no overage charges and no fair-use throttling inside your tier.

259
starting price per month
129.5
per Gbps per month
7
configurations
648 TB
transfer ceiling per month

Ports rarely fail gracefully. They fill, and then everything behind them slows at once: page loads, backup windows, even the health checks your monitoring relies on to tell you what is wrong. Once a gigabit link is running near its ceiling through the evening, the question stops being which capacity you need and becomes when you will act on it, on an afternoon of your choosing or at two in the morning on somebody else's. That is what this step changes. It is not a bigger design, it is the same design with the pressure taken off, and its real product is that it converts an urgent decision into a scheduled one.

What 250 MB/s Buys, and Why the Number Is Exactly Twice

The port moves 250 MB/s sustained and 648 TB over a thirty-day month, committed rather than best effort, with no overage charges inside the tier. The more useful way to read those figures is structural. Two gigabit links running flat out fit inside this port exactly, with nothing spare, and this is the smallest absolute step anywhere in the range. If you are arriving from one saturated port, the practical reading is room to roughly double before the same pressure returns. Measured against the tiers above, that is not a lot of runway, and it is not meant to be. What matters is that none of it requires changing how the service is built: the same application, the same deployment, the same operational habits, on a link that is no longer the first thing to give way.

648 TB
Transfer per month
Sustained volume at full port utilisation across a 30-day month.
400
1080p viewers
Simultaneous streams at 5 Mbps each.
80
4K viewers
Simultaneous streams at 25 Mbps each.
1 h 6 min
Per terabyte
Wall-clock time to move 1 TB off the server at line rate.

Figures use decimal units (1 Gbps = 125 MB/s) across a 30-day month, at 5 Mbps per 1080p stream and 25 Mbps per 4K stream. They describe what the port can carry, not a guarantee of application throughput: disk, CPU and the far end all apply their own limits.

7 configurations at 2Gbps

Seven builds carry this port, and they do not all pay the same for it. The three Dual Xeon E5 machines take €50 for the second gigabit: €259 for the E5-2630v3, €279 for the E5-2690v4 that most customers choose, €349 for the E5-2699v4. The four EPYC builds take €80: €529 for the single EPYC 9254, €759 for the dual 9254, €1039 for the dual EPYC 9554 and €1599 for the dual EPYC 9754. The two lines are priced differently for the same extra gigabit at every step of this ladder, but the direction is not constant, and this is the step where it favours Xeon. Here the Xeon builds take €50 against the EPYC €80. One rung further, from two gigabits to five, that reverses to €260 on Xeon against €240 on EPYC. The practical reading at this step is simple: the Xeon line is the cheaper place to buy the reserve, while the EPYC line is where you buy it because the processor was already going to be EPYC.

Dual Xeon E5-2630v3

Entry Level

259
/month
16c / 32t @ 2.4GHz128GB DDR4 @ 2133MHz
6x SSD Slots2Gbps Guaranteed
AI Optimized Routing Included
Order Now
POPULAR

Dual Xeon E5-2690v4

Popular Choice

279
/month
28c / 56t @ 2.6GHz128GB DDR4 @ 2133MHz
6x SSD Slots2Gbps Guaranteed
AI Optimized Routing Included
Order Now

Dual Xeon E5-2699v4

High Performance

349
/month
44c / 88t @ 2.2GHz256GB DDR4 @ 2133MHz
6x SSD Slots2Gbps Guaranteed
AI Optimized Routing Included
Order Now

Single EPYC 9254

Entry Level

529
/month
24c / 48t @ 2.9GHz128GB DDR5 @ 4800MHz
10x NVMe Slots2Gbps Guaranteed
AI Optimized Routing Included
Order Now

Dual EPYC 9254

High Performance

759
/month
48c / 96t @ 2.9GHz256GB DDR5 @ 4800MHz
10x NVMe Slots2Gbps Guaranteed
AI Optimized Routing Included
Order Now

Dual EPYC 9554

Ultimate Performance

1039
/month
128c / 256t @ 3.1GHz512GB DDR5 @ 4800MHz
24x NVMe Slots2Gbps Guaranteed
AI Optimized Routing Included
Order Now
ULTIMATE

Dual EPYC 9754

Maximum Performance

1599
/month
256c / 512t @ 2.3GHz1024GB DDR5 @ 4800MHz
24x NVMe Slots2Gbps Guaranteed
AI Optimized Routing Included
Order Now

What a 2Gbps server is for

Streaming catalogues where the 4K share is the real variable

The ceiling is 400 concurrent 1080p streams or 80 at 4K, and almost nobody sits at either end. What decides your number is the mix: each viewer who moves up to 4K costs five of the 1080p ones. The input to size on is therefore the share of your audience on the larger profile, not the size of the audience.

Restore windows that have to close before the working day

Wire time for a terabyte is 1 h 6 min at this speed, and the figure that matters is not the average night but the worst one. Halving transfer time does not make a backup better in any technical sense. It decides whether a restore finishes inside the maintenance window or is still running when the first users arrive.

Cache origins that must not become the bottleneck behind a CDN

An origin is sized for misses, not for hits, and miss rates move for reasons outside your control: a purge, a new edge region, one badly set header on a popular object. A doubled uplink means an unexpected miss storm is served slowly rather than refused outright, which is the behaviour an origin actually gets judged on.

Release days and patch days on a machine sized for normal ones

Distribution traffic is bursty in a way that monthly averages conceal: eleven quiet days, then everybody downloads within the same hour. Sizing the link for the median leaves the launch window rationed, and this step is usually bought so that the spike day is merely busy instead of a queue of failed downloads.

A Line Item, Not a Budget Conversation

The step from one gigabit to two costs €50 to €80 a month depending on platform. Set against the machine underneath it, that is an adjustment rather than a decision, which is precisely why it tends to be approved quickly, and why it is worth approving before the port is already full rather than after. On the entry build the per-gigabit figure improves as well, to €129.50, though that is a by-product of the step and not a reason for taking it. Nobody buys a second gigabit for the unit economics of a second gigabit. They buy it because the alternative on the table sits at a price point that needs a different approval, and because a modest recurring cost is far easier to defend than a larger recurring line.

All guaranteed bandwidth tiers

The same hardware range is sold at nine committed port speeds. Compare the entry price of each.

2Gbps dedicated server questions

I have two machines that each peak at a gigabit. Will one 2Gbps server really hold both?
Only if their peaks do not land on the same minute. The port has room for both by arithmetic, which is not the same as having room for both in practice: two services that are busy on the same clock will collide at exactly the moment you need them not to. Compare the traffic curves before merging anything, and remember that CPU, memory and disk have to absorb both workloads as well, which is usually the harder half of the move.
If nobody designs a service around two gigabits, what is this tier actually for?
For the gap between a port that is full and a redesign you are not ready to run. The second gigabit is headroom you hope stays idle, and that is a legitimate thing to pay for: it behaves less like capacity you consume and more like a spare you keep on the vehicle. If it is never called on, the tier did its job quietly. If it is called on, the failure mode becomes a busy port instead of a broken service.
We peak at a gigabit for two hours a night and sit near idle otherwise. Is a second gigabit wasted?
That shape is the strongest case for this tier rather than an argument against it. A port is sized for its busiest moment, not its average, so the hours you spend near idle are irrelevant to the decision. What matters is that the two-hour peak currently has nothing above it: at one gigabit a peak that reaches the ceiling has no way to signal how much further it would have gone. The second gigabit is bought to stay unused, and the first thing it gives you is a truthful measurement of what your peak actually wants.
Is buying two gigabits just a way of putting off the 5Gbps decision?
Yes, deliberately, and the honest test is whether the delay is being used for something. Doubling the port is an adjustment to an existing line item; the tier above is a different budget conversation, usually with somebody else's signature on it. Paying a small monthly premium to keep serving customers while that case is built is sound. It stops being sound the moment you are confident the larger port is coming anyway, because then you are funding the same transition twice.
Two saturated gigabits fit this port exactly. Should I plan to use all of it?
No, and that arithmetic is the warning rather than the invitation. A port whose expected peak equals its own ceiling has no reserve left, which is exactly the condition you are trying to leave behind. Treat the doubled figure as a boundary you do not intend to reach rather than a target you are sizing towards, and start looking upward once routine evening peaks settle into the upper part of the port instead of touching it occasionally.

2 Gbps dedicated servers by location

Same hardware and guaranteed port, deployed from the datacenter closest to your users.

Order a 2Gbps Server, or Have One Built to Order

Every build ships with up to 1Tbps of DDoS protection, a 99.9% uptime SLA, AI-optimized BGP routing and 24/7 support from our own system administrators. CPU, RAM to 1TB DDR5, storage, bandwidth tier and operating system are all adjustable, and anything outside the standard line-up is custom quoted on request.

Up to 1Tbps DDoS protectionDeployed in about 1 hour99.9% uptime SLA