Jetson Thor T4000 VP9 decode capacity
Dataset 2026-09-07 · verified 2026-09-07 · last updated September 2026 · preliminary (NVIDIA)
NVIDIA rates the Jetson Thor T4000 NVDEC at 1 × 8K30 · 3 × 4K60 · 7 × 4K30 · 14 × 1080p60 · 29 × 1080p30 VP9 (1840 MP/s) at 70 W.
Verbatim datasheet figures with their source. Anything not in the table below is an EdgeAIStack estimate and is labelled as such. Jetson Thor T4000 platform page.
Published decode points (class A)
Jetson Thor T4000: 1 NVDEC instance, VP9 profiles Profile 0, Profile 2, 8-bit 4:2:0 unless the datasheet says otherwise. Stream counts are per module (per-instance counts × 1, as the datasheet’s total column does).
70 W mode (datasheet default)
| Profile | Pixels @ fps | Per instance | Streams (module) | MP/s per stream | Streams × MP/s | Evidence |
|---|---|---|---|---|---|---|
| 8K30 | 7680×4320 @ 30 | — | 1 | 995.3 | 995 | A datasheet |
| 4K60 | 3840×2160 @ 60 | — | 3 | 497.7 | 1,493 | A datasheet |
| 4K30 | 3840×2160 @ 30 | — | 7 | 248.8 | 1,742 | A datasheet |
| 1080p60 | 1920×1080 @ 60 | — | 14 | 124.4 | 1,742 | A datasheet |
| 1080p30 | 1920×1080 @ 30 | — | 29 | 62.2 | 1,804 | A datasheet |
| Rated throughput (up to) | 1,840 MP/s | A datasheet | ||||
Source
- NVIDIA · DS-11945-001 v1.5 §2.2 Multi-Standard Video Decoder, Table 2-3 (T4000), 70W Mode · verified 2026-09-07 · Jetson Thor Series Modules Data Sheet · class A
“VP9 Profile 0, Profile 2 29x 14x 7x 3x 1x 1840”
MAXN mode
| Profile | Pixels @ fps | Per instance | Streams (module) | MP/s per stream | Streams × MP/s | Evidence |
|---|---|---|---|---|---|---|
| 8K30 | 7680×4320 @ 30 | — | 1 | 995.3 | 995 | A datasheet |
| 4K60 | 3840×2160 @ 60 | — | 4 | 497.7 | 1,991 | A datasheet |
| 4K30 | 3840×2160 @ 30 | — | 8 | 248.8 | 1,991 | A datasheet |
| 1080p60 | 1920×1080 @ 60 | — | 16 | 124.4 | 1,991 | A datasheet |
| 1080p30 | 1920×1080 @ 30 | — | 32 | 62.2 | 1,991 | A datasheet |
| Rated throughput (up to) | 2,000 MP/s | A datasheet | ||||
Source
- NVIDIA · DS-11945-001 v1.5 §2.2 Multi-Standard Video Decoder, Table 2-3 (T4000), MAXN Mode · verified 2026-09-07 · Jetson Thor Series Modules Data Sheet · class A
“VP9 Profile 0, Profile 2 32x 16x 8x 4x 1x 2000”
v1.5 dated June 2026. Tables are per NVDEC/NVENC instance and per power mode; totals here multiply by the instance count as the datasheet's "Total Mpix/sec (2x)" column does. T4000 rows are marked preliminary by NVIDIA.
Preliminary: Jetson T4000 module encode, decode, and low-speed I/O specifications are preliminary and subject to change (NVIDIA). Every result built on this row is LOW confidence until NVIDIA finalises the datasheet.
Between the published points (class D)
Camera profiles NVIDIA does not print are estimated by comparing the pixel rate against the rated 1840 MP/s (70 W table): max streams = ⌊1840 ÷ (width × height × fps ÷ 10⁶)⌋. These are EdgeAIStack numbers, not NVIDIA's; they assume roughly linear NVDEC throughput inside the published range and are not asserted outside it. Rows that coincide with a published point show the datasheet count instead.
| Profile | Pixels @ fps | MP/s per stream | Max streams (est.) | Evidence | Full pipeline |
|---|---|---|---|---|---|
| 720p30 | 1280×720 @ 30 | 27.6 | 66 | D pixel-rate | check the full pipeline |
| 1080p15 | 1920×1080 @ 15 | 31.1 | 59 | D pixel-rate | check the full pipeline |
| 1080p20 | 1920×1080 @ 20 | 41.5 | 44 | D pixel-rate | check the full pipeline |
| 1080p25 | 1920×1080 @ 25 | 51.8 | 35 | D pixel-rate | check the full pipeline |
| 1440p15 | 2560×1440 @ 15 | 55.3 | 33 | D pixel-rate | check the full pipeline |
| 1440p20 | 2560×1440 @ 20 | 73.7 | 24 | D pixel-rate | check the full pipeline |
| 1440p30 | 2560×1440 @ 30 | 110.6 | 16 | D pixel-rate | check the full pipeline |
| 4K15 | 3840×2160 @ 15 | 124.4 | 14 | D pixel-rate | check the full pipeline |
| 4K20 | 3840×2160 @ 20 | 165.9 | 11 | D pixel-rate | check the full pipeline |
| 4K25 | 3840×2160 @ 25 | 207.4 | 8 | D pixel-rate | check the full pipeline |
Power mode
On Thor the NVDEC/NVENC clock follows the power mode (1.56 GHz at 120 W / 70 W, 1.69 GHz at MAXN) and the datasheet publishes separate decode tables per mode.
Source
- NVIDIA · DS-11945-001 v1.5 §1 Overview, HD Video table · verified 2026-09-07 · Jetson Thor Series Modules Data Sheet · class A
“Jetson T5000: 2x NVDEC @ 120W: 1.56 GHz | MAXN: 1.69 GHz”
Encode
The datasheet publishes no hardware VP9 encode row for Jetson Thor T4000. Published encoder rows: H.264, H.265.
What these numbers do and do not cover
- Hardware limit, not software. JetPack, DeepStream, GStreamer or Frigate may cap concurrent decoder sessions below the datasheet count.
- Decode in isolation. Simultaneous NVENC use shares memory bandwidth; expect real decode + encode totals below the sum of the two tables.
- Profile and chroma. Counts are for Profile 0, Profile 2 at 8-bit 4:2:0; 4:4:4 halves throughput (datasheet note); Main10 shares the row.
- Decode is rarely the bottleneck. Inference, memory and preprocessing usually fail first on a detection pipeline; run the full pipeline check for your workload.
- Method. Formulas, evidence classes and the confidence rule are on the methodology page.
Other codecs on this module
VP9 on other modules
Is decode really the limit?
Run Jetson Thor T4000 with your camera count, model and frame rate: every stage as a constraint, every number with its source, a permanent link.