A Contabo VPS may be able to transcode 1080p video for a 24/7 YouTube stream, but its advertised vCPU count cannot tell you whether it will keep up. The answer depends on the source, encoder settings and CPU capacity available continuously on the particular instance.
First check whether you need full video re-encoding at all. Sending an already encoded file by stream copy or relay avoids the video encoding workload; it is not transcoding, and it should be tested as a different kind of job.
What 1080p transcoding requires
Full transcoding means decoding the source video and encoding it again in the format and settings you intend to send to YouTube. A file's 1080p resolution is only one part of that workload. Frame rate, codec, encoder, preset, filters and any scaling or other processing also matter. A source with complex motion can be harder to encode than a static image, even at the same resolution and frame rate.
For a 24/7 stream, the encoder has to process media at least as quickly as it is played. If it falls behind, latency can grow, frames can be dropped, or the stream can become unstable. An encode that briefly looks fine is not proof that it will remain real-time after hours of continuous work.
YouTube's live encoder settings guidance recommends H.264 at 5 Mbps for 1080p30 and 6 Mbps for 1080p60. It also recommends constant bitrate (CBR), a two-second keyframe interval, and says not to exceed four seconds. These are ingest settings, not a CPU sizing formula: the bitrate target does not show whether a particular VPS can encode your source quickly enough.
YouTube also converts an incoming live stream into multiple output formats for viewers. Your encoder's task is therefore to produce a suitable ingest stream; you do not need to generate every viewer rendition on your VPS. That distinction can prevent you from sizing the server for work YouTube performs after ingest, but it does not reduce the need to measure the encoding your own source actually requires.
Transcoding is not stream copy
If you decode and re-encode video, you are transcoding it. If you send an already encoded video track unchanged, you are using stream copy for that track. A relay may forward a stream without re-encoding it. These workflows can look similar to viewers—a continuous video on YouTube—but they place very different demands on a VPS.
Stream copy can be useful when the existing media already meets the output requirements. It does not make an incompatible codec, frame rate, resolution or other property compatible by magic. If the video must be changed, that track needs to be encoded; changing audio alone does not make the video a transcode, but it can add its own processing work.
Before choosing a plan, inspect what your playback and streaming process will actually do. Confirm whether it copies or encodes the video, and whether filters, scaling, subtitles or compositing force a decode-and-encode path. If you are preparing a playlist-based channel, this guide to running a 24/7 lo-fi playlist stream is useful context for thinking through the media workflow. Do not assume that a playlist automatically means either stream copy or transcoding; the command and source format decide.
The difference matters when you interpret a test. A VPS that relays an already encoded file successfully has not demonstrated that it can encode a 1080p source in real time. Equally, a CPU-heavy transcode test says little about a simple relay. Test the workflow you plan to leave running.
Why shared vCPU counts are not proof
Contabo documents its Core VPS and Performance VPS families as using shared vCPUs. A displayed vCPU count tells you how many virtual processors are assigned in the plan description; it does not establish the uninterrupted CPU time available to your job, or the encoding throughput of your specific source and encoder. On a shared host, other activity and scheduling can affect the CPU time available to a virtual machine.
A higher advertised count is not a benchmark. Nor do RAM capacity, storage type or a processor generation by themselves prove sustained 1080p encoding performance. Those specifications may be relevant to choosing a machine, but none answers the practical question: can this exact workload remain at real time for a long run without accumulating dropped frames or interruptions?
Contabo's VPS documentation describes its current product families and resource allocations. The Core VPS range shown in the documentation spans 4–18 vCPUs and 8–96 GB RAM, but those figures are product specifications, not measured encoding results. Product names and specifications can change, so consult the live documentation when comparing plans rather than treating these figures as permanent.
The sources reviewed for this article do not provide a controlled benchmark for 1080p transcoding on a named Contabo plan. There is no responsible way to infer that result from a vCPU count or promise that a particular VPS will sustain the work overnight. Use the plan sheet to identify candidates, then use a test of your own workload to make the decision.
When dedicated CPU resources may be relevant
Contabo's Max Performance VPS, also described as VDS, is documented with dedicated CPU cores and reserved RAM. That makes it more relevant to evaluate when continuous full re-encoding is CPU-heavy and predictable CPU access matters. It still does not prove that any particular 1080p source, encoder and preset will run in real time.
The VDS documentation lists AMD EPYC 7282 at 2.8 GHz for the plans shown. Treat this as a published specification, not an FFmpeg benchmark or a promise of capacity. A dedicated-core product may be worth comparing with a shared-vCPU VPS under the same test, especially if contention on a shared machine is a concern. The only useful comparison for your channel is the sustained result you observe with the same input and settings.
Dedicated resources are not automatically the right choice. If your source can be relayed or stream-copied, a full CPU allocation may solve a problem you do not have. If you must re-encode, test whether a dedicated-core option improves consistency enough to justify its cost in your region. Review Contabo's Max Performance VPS documentation for current resource descriptions and confirm current prices and terms directly with the vendor before purchasing.
Test the exact source and encoder settings
Write down the workload before ordering a server. Record the source resolution and frame rate, the video and audio codecs, the intended output resolution and frame rate, the encoder and preset, any filters, and the target bitrate. If you will loop a file, test the actual loop and its transitions, not a short, unusually simple excerpt. The goal is to reproduce the work the VPS will perform day and night.
Run the intended encoder on the instance you are considering. Confirm that its processing rate remains at or above real time. Watch for dropped frames, encoding lag and CPU contention, and check whether the process remains stable rather than merely reporting a successful start. Repeat with the actual media if the channel rotates between different types of content; a single representative file may not expose the hardest section of the programme.
For a useful comparison, change one factor at a time. You might compare the intended preset with a faster preset, or test a shared-vCPU candidate against a dedicated-core VDS. Keep the input and output settings the same so that a result has meaning. If you change resolution, frame rate, filters and preset all at once, you will not know which change made the workload manageable.
Then send a representative test stream to YouTube and review its health messages as well as local encoder behaviour. YouTube advises testing with content similar to the real stream and monitoring stream health during the event. A local encode that keeps up does not by itself confirm that the outbound connection is stable or that YouTube receives the intended signal. For practical background on ingest settings, the 720p pre-recorded loop bitrate guide can help you think through the relationship between resolution and stream settings; use YouTube's current guidance for the actual configuration you choose.
Monitor sustained performance before relying on it
A brief test is a screening step, not evidence for a 24/7 commitment. Run the complete workload continuously for a representative sustained interval, long enough to expose the conditions that a quick launch test can miss. There is no single duration that proves every deployment safe: the relevant point is to observe the exact process for long enough to decide whether its performance stays stable under your operating conditions.
Track whether encoding remains at real time, whether dropped frames accumulate, and whether CPU use or contention changes. Watch memory and disk activity too, particularly if the workflow reads large files or writes temporary outputs. A server can have enough average CPU capacity but still produce poor results if the process is interrupted, a resource becomes constrained, or a restart does not recover the stream cleanly.
Observe the outbound connection separately from encoding. YouTube's suggested H.264 ingest bitrates are targets for the video feed, but your measured upload stability must support the stream with operating margin. A listed port rate is not a guarantee of sustained throughput for a particular VPS. Contabo says there is no default bandwidth cap, subject to its fair-use policy, and notes that exceptionally high or disruptive usage may be throttled; see its bandwidth and traffic limits guidance. Check the current terms and monitor your own route rather than treating a port figure or the word “unlimited” as a performance guarantee.
Finally, test recovery. Decide what happens after a process exits, a connection drops, or the virtual machine restarts. Add process supervision, alerts and a fallback stream or source if losing the channel for an extended period would matter. For a troubleshooting checklist when YouTube reports an ingest problem, see what to check when a YouTube stream shows an error in Live Control Room. Monitoring and recovery make failures visible; they do not guarantee uninterrupted service.
Choose a workload-appropriate plan
Choose from evidence, not a headline specification. If the channel only needs to relay compatible pre-encoded media, test that lightweight workflow and avoid paying for CPU capacity intended for full video encoding. If it must re-encode continuously, compare candidate plans with the exact source and settings, including shared-vCPU VPS and dedicated-core VDS options where relevant.
| What to compare | What it tells you | What it does not prove |
|---|---|---|
| Shared or dedicated CPU allocation | Whether CPU resources are described as shared or dedicated | Sustained encoding speed for your workload |
| CPU model and published specifications | Which hardware specification the vendor documents | A real-time 1080p benchmark |
| RAM and storage | Whether the plan suits the surrounding workflow and files | That encoding can keep pace |
| Sustained test result | Whether your chosen workload stayed real-time during the test | That future conditions can never change |
| Upload stability and policy | Whether the outbound feed appears suitable under observed conditions | Guaranteed throughput from a listed port rate |
| Current regional cost and terms | The purchase trade-off at the time you check | That a cheaper plan is adequate or a dearer one is necessary |
For a CPU-heavy transcode, prefer the candidate that passes the sustained test with a useful margin and stable output, not the one with the largest advertised count. If neither candidate passes, revise the workload—perhaps a simpler preset or a suitable source format—or choose a different way to run the channel. Retest after material changes to media, encoder version, filters or stream settings.
If the key issue is keeping a file-based channel running without leaving your own computer on, StreamNeo removes that specific always-on-computer task by turning an uploaded video into a YouTube live stream; it does not change YouTube's ingest settings or make a Contabo VPS benchmark unnecessary for workloads that still need one. It is YouTube-only, so it is not a fit if you need another platform.
Before committing, compare the operating options on the pricing page. When the file and channel are ready, start free — 24-hour trial, no card.
FAQ
Can I tell from Contabo's vCPU count whether 1080p will work?
No. Shared vCPU counts describe a virtual plan, not sustained encoding throughput for your source, encoder and settings. Test the exact workload on the intended instance and check that it remains at real time.
Is sending a video by stream copy the same as transcoding?
No. Transcoding decodes and re-encodes video; stream copy sends an already encoded track without doing that video encoding step. A successful relay test does not demonstrate that the same VPS can re-encode 1080p in real time.
Does a dedicated-core VDS guarantee 24/7 1080p encoding?
No. Contabo documents dedicated CPU cores and reserved RAM for its Max Performance VPS/VDS, which makes it relevant to test for CPU-heavy work, but that is not a workload benchmark or guarantee. Compare it with other candidates using the same sustained test.
What should I monitor during a YouTube test?
Check that encoding stays at or above real time, dropped frames do not accumulate, the process remains stable, and the outbound connection holds up. Review YouTube's stream health messages and test recovery after a process or connection interruption before relying on the setup around the clock.