Calculate the TPS of the B300 on local AI models
Every model in our catalogue assessed against this card at the context length and minimum quality you choose. Speed is an estimate for a single request, calculated from this card's memory bandwidth and the size of each model once compressed.
Calculated for this card
721 models in our catalogue altogether
Largest model it holds
Nemotron 3 Ultra
550B · Q3_K_M · 92.4 tok/s
Fastest model
Gemma 3 QAT 1B
3,388 tok/s · 1B
Which AI models can run on a B300?
Set the inputs, read the answer
More context means more memory for the conversation cache. Speed is for a fresh conversation and does not change with this setting.
Hides models that would only fit by being compressed below this point.
689 models match
Calculating| Quantisation | Fit | ||||||
|---|---|---|---|---|---|---|---|
|
3,388
tok/s
2,880–4,066 |
Gemma 3 1B | 1B | Mar 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
3,388
tok/s
2,880–4,066 |
Gemma 3 QAT 1B | 1B | Apr 2025 | 1.8 GB | 33k tokens | Q8_0 | Comfortable |
|
3,388
tok/s
2,033–5,421 · low confidence |
HGRN 1B (WT 103) ≈ | 1B | Nov 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
3,388
tok/s
2,033–5,421 · low confidence |
LLama 3..2 Typhoon 2 1B ≈ | 1B | Dec 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
3,388
tok/s
2,033–5,421 · low confidence |
OLMo-1B ≈ | 1B | Feb 2024 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
3,388
tok/s
2,033–5,421 · low confidence |
Pythia-1b ≈ | 1B | Apr 2023 | 1.8 GB | 131k tokens ? | Q8_0 | Comfortable |
|
3,137
tok/s
1,882–5,020 · low confidence |
OpenELM-1.1B ≈ | 1.1B | May 2024 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
3,080
tok/s
1,848–4,928 · low confidence |
DeciCoder-1B ≈ | 1.1B | Aug 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
3,080
tok/s
1,848–4,928 · low confidence |
SantaCoder ≈ | 1.1B | Jan 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
3,080
tok/s
1,848–4,928 · low confidence |
TinyLlama-1.1B (1T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
3,080
tok/s
1,848–4,928 · low confidence |
TinyLlama-1.1B (3T token checkpoint) ≈ | 1.1B | Oct 2023 | 1.9 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,824
tok/s
1,694–4,518 · low confidence |
EXAONE 4.0 (1.2B) ≈ | 1.2B | Jul 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,824
tok/s
1,694–4,518 · low confidence |
LFM2-1.2B ≈ | 1.2B | Jul 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,824
tok/s
1,694–4,518 · low confidence |
MinerU2.5 ≈ | 1.2B | Sep 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,824
tok/s
1,694–4,518 · low confidence |
Pleias 1.0 1.2B ≈ | 1.2B | Dec 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,824
tok/s
1,694–4,518 · low confidence |
Pleias-RAG-1B ≈ | 1.2B | Apr 2025 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,755
tok/s
2,341–3,306 |
Llama 3.2 1B | 1.2B | Sep 2024 | 2.2 GB | 131k tokens | Q8_0 | Comfortable |
|
2,716
tok/s
1,630–4,346 · low confidence |
MiniCPM-1.2B ≈ | 1.2B | Jun 2024 | 2.0 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,606
tok/s
1,564–4,170 · low confidence |
DeepSeek Coder 1.3B ≈ | 1.3B | Jan 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,606
tok/s
1,564–4,170 · low confidence |
DeepSeek-VL-1.3B ≈ | 1.3B | Mar 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,606
tok/s
1,564–4,170 · low confidence |
DigiRL ≈ | 1.3B | Jun 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,606
tok/s
1,564–4,170 · low confidence |
GLA Transformer 1.3B ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,606
tok/s
1,564–4,170 · low confidence |
Janus 1.3B ≈ | 1.3B | Oct 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,606
tok/s
1,564–4,170 · low confidence |
Kosmos-2.5 ≈ | 1.3B | Aug 2024 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
|
2,606
tok/s
1,564–4,170 · low confidence |
Otter ≈ | 1.3B | May 2023 | 2.1 GB | 131k tokens ? | Q8_0 | Comfortable |
Speeds are estimates for a single request — one conversation at a time — calculated from memory bandwidth, model size and quantisation. Real throughput varies with the inference runtime and its version. Figures published by hardware vendors measure many simultaneous requests and are much higher.
On record
B300 full specification
Everything on record for this board, ordered by how much it bears on running a language model rather than by how a spec sheet would list it. Memory comes first because it decides the outcome; the rest is context.
Memory
The two specifications that decide what this card can run and how quickly. Capacity sets which models fit; bandwidth sets how many tokens per second they produce once they do.
- Memory size
- 288 GB
- Memory bandwidth
- 8,000 GB/s
- Memory type
- HBM3e
- Memory bus width
- 8,192 bit
- Memory clock
- 1.95 GHz
The chip
Which processor is on the board and how it was manufactured. A smaller process size generally means more performance for the same power.
- Graphics processor
- GB110
- Architecture
- Blackwell Ultra
- Generation
- Server Blackwell(Bxx)
- Foundry
- TSMC
- Process size
- 5 nm
- Transistors
- 104 billion
- Released
- 11 September 2025
Clock speeds
How fast the processor runs. Worth far less here than on a gaming benchmark: generating text is limited by memory bandwidth, so a higher clock barely moves the result.
- Base clock
- 1.67 GHz
- Boost clock
- 2.03 GHz
Processing units
What the chip contains. These drive graphics performance and matter mainly for processing a long prompt rather than for producing the answer.
- Shading units
- 18,944
- Texture mapping units
- 592
- Render output units
- 24
- Streaming multiprocessors
- 148
- Tensor cores
- 592
- L1 cache
- 250 KB
- L2 cache
- 50 MB
Theoretical performance
Peak arithmetic rates published for the board. These are ceilings that no real workload reaches, and generating text reaches a small fraction of them because it is limited by memory rather than arithmetic.
- Half precision (FP16)
- 1,231.8 TFLOPS
- Single precision (FP32)
- 77 TFLOPS
- Double precision (FP64)
- 1.2 TFLOPS
- Pixel rate
- 49 GPixel/s
- Texture rate
- 1,203 GTexel/s
The board
What it takes to physically install and power the card — the practical constraints that decide whether it fits the machine you already own.
- Power draw (TDP)
- 1,400 W
- Suggested power supply
- 1,800 W
- Bus interface
- PCIe 5.0 x16
- Slot width
- SXM Module
Software support
Which graphics and compute interfaces the card supports. CUDA compute capability is the one that bears on inference: below 7.0 there are no tensor cores, and modern inference software falls back to slower code paths.
- CUDA compute capability
- 10.3
- OpenCL
- 3.0
Listings
Where to buy a B300
No vendor is currently listing this card. Listings come from vendors who publish them here directly — browse the vendor directory to see who is selling what.
What the numbers mean
What the memory subsystem means for AI
Memory
288 GB
Bandwidth
8,000 GB/s
Largest model
Nemotron 3 Ultra
B300 holds 288 GB of HBM3e. That puts it in the class of hardware that holds the largest open-weight models without splitting them across machines. An inference runtime can reach roughly 259.2 GB.
Memory bandwidth reaches 8,000 GB/s across a bus of 8,192 bits. That is at the top of what exists. Since each token means reading the whole model out of memory once, it translates almost directly into generation speed — this card is bandwidth-rich enough that model size stops being the limiting factor long before the bus does.
The figure is the bus width multiplied by a memory clock of 1.95 GHz. Widening the bus and raising the clock are the two levers a manufacturer has, which is why a card with unremarkable cores can still generate quickly.
The practical ceiling is Nemotron 3 Ultra, 550B, compressed to Q3_K_M and generating around 92.4 tokens per second.
The chip and how it was built
B300 is built on the graphics processor GB110, using the architecture Blackwell Ultra from NVIDIA, as part of the generation Server Blackwell(Bxx).
The chip is manufactured by TSMC, on a process of 5 nm, holding 104 billion transistors. A smaller process generally means more performance for the same power, though for language models it matters far less than the memory subsystem.
It was released in September 2025, roughly 1.0058533472133 years ago. Inference software support tends to follow hardware by a year or two, so a card of this age generally has mature, well-optimised code paths available to it.
Compute throughput, and why it matters less than it looks
FP16
1,231.8 TFLOPS
FP64
1.2 TFLOPS
Tensor cores
592
On paper B300 reaches 1,231.8 TFLOPS at half precision, and 77 TFLOPS at single precision. These are peak figures no real workload sustains, and generating text reaches only a small fraction of them — decoding is limited by memory rather than arithmetic, which is why a card can look enormously powerful here and still produce tokens at an ordinary rate.
Double-precision throughput reaches 1.2 TFLOPS. It has no bearing on running a language model — no inference runtime uses it — but it separates datacentre parts from consumer ones, since the latter deliberately restrict it.
The card carries 592 tensor cores across 148 streaming multiprocessors. These accelerate the matrix arithmetic at the heart of a transformer, and they are what make prompt processing — reading a long document before answering — dramatically faster than it would otherwise be.
Clocks run from a base of 1.67 GHz to a boost of 2.03 GHz. Worth far less here than on a gaming benchmark: raising the clock speeds up the arithmetic, and the arithmetic is not what generation is waiting on.
Cache and processing units
B300 has an L1 cache of 250 KB, backed by an L2 cache of 50 MB. Cache absorbs a share of the memory traffic that would otherwise hit the main bus, which is the one place on this page where a number other than bandwidth quietly affects generation speed — a large L2 lets more of the working set stay close to the cores.
There are 18,944 shading units, 592 texture mapping units, and 24 render output units. These drive graphics workloads and contribute to prompt processing, but they sit idle for much of the time a model spends generating a reply.
Power, size and installation
Power draw
1,400 W
B300 is rated at 1,400 W, and the suggested system power supply is 1,800 W. Running a language model keeps a card busy in bursts rather than continuously — it draws hard while generating and idles between requests — so sustained draw over a working day is usually well below the rated figure.
The board occupies sxm module. Worth checking against the case and power supply already in the machine, since the largest cards need considerably more of both than a typical desktop provides.
It connects over PCIe 5.0 x16. The interface governs how quickly a model is loaded from disk into the card, not how fast it runs once there, so a narrower link costs a few seconds at startup and nothing thereafter.
The extremes
The largest AI models that run on a B300
The biggest open-weight models that fit on this card, newest first. Each is shown at the best compression the card can hold.
The fastest AI models on a B300
Where this card produces tokens quickest. Smaller models dominate here, because generating each token means reading the whole model out of memory once.
Step by step
How to work out the tokens per second of a B300
You do not have to calculate anything by hand — the gputps.com calculator on this page has already worked it out for every model this card can hold. Reading off the answer takes six steps.
-
01
Start with the model, not the specification
The table lists 689 models this card runs. Search narrows the list by name or by size.
-
02
Match the context to your work
Drag the slider to the conversation length you plan to work at. The cache grows with the conversation, and against a card holding 288 GB so the setting is worth getting right.
-
03
Pin the comparison to one quality level
Each model is shown at the best compression this card can hold. A minimum quality hides the ones that only fit by being squeezed further than you would accept.
-
04
Look at the range, not just the number
Speeds come with error bars for a reason. The best case here is 3,388 tok/s on Gemma 3 QAT 1B. The same card and model vary by thirty to fifty per cent between inference runtimes.
-
05
Check the headroom before you decide
Tight means it works today; comfortable means it still works when the conversation grows. Compare what each model needs against an available 288 GB.
-
06
Cross-check against other hardware
Following a model through to its own page lists all the hardware that can run it, so you can see how it compares against B300.
Answers
B300 — common questions
B300— what is its memory bandwidth?
Memory bandwidth reaches 8,000 GB/s across a bus of 8,192 bits. This is the single best predictor of how fast it generates text, because producing each token means reading the entire model out of memory once.
B300— what type of memory does it use?
It uses HBM3e clocked at 1.95 GHz. HBM types are found on datacentre accelerators and carry far more bandwidth than the GDDR used on desktop cards, which is why they generate tokens considerably faster at the same capacity.
B300— who makes it?
This is a product of NVIDIA, with the chip manufactured by TSMC, on a process of 5 nm.
B300— when was it released?
It was released in September 2025.
B300— how much power does it use?
Rated board power is 1,400 W, and the suggested system power supply is 1,800 W. Generating text draws hard in bursts and idles between requests, so average consumption over a working session is normally well below the rated figure.
B300— how much cache does it have?
The L1 cache is 250 KB, and the L2 cache is 50 MB. Cache absorbs part of the memory traffic that would otherwise reach the main bus, so a larger L2 gives a modest lift to generation speed beyond what bandwidth alone predicts.
B300— what are its TFLOPS?
It is rated at 1,231.8 TFLOPS at half precision and 77 TFLOPS at single precision. These are peak arithmetic ceilings rather than achievable rates, and text generation reaches only a small fraction of them because it is limited by memory bandwidth instead.
B300— how many tensor cores does it have?
It has 592 tensor cores across 148 streaming multiprocessors. They accelerate the matrix arithmetic a transformer is built from, which mainly speeds up processing a long prompt rather than producing the reply.
B300— does it support CUDA?
Yes. It reports CUDA compute capability 10.3. Capability 7.0 and above has tensor cores, which modern inference software uses; below that it falls back to slower code paths for quantised models.
B300— what bus interface does it use?
It uses PCIe 5.0 x16. This governs how fast a model is loaded onto the card rather than how fast it runs once loaded, so it costs a few seconds at startup and nothing during generation.
B300— is it good for running local AI models?
Its memory is large enough for models most desktop hardware cannot touch and its bandwidth puts it among the fastest hardware available for generation. In total it runs 689 of the models we track. Whether that is enough depends entirely on which model you want — the table above answers that directly.
B300— can it run a model that does not fit in its memory?
Offloading past the card's 288 GB sit in system memory and run at a fraction of the speed, so a mostly-offloaded model is rarely worth using. Every figure here assumes it is fully resident on the card.
Would two B300 cards be twice as fast?
No. A second card doubles the memory to 576 GB of combined memory, at roughly the same generation speed as one.
B300— which AI models can it run?
689 of the 721 open-weight language models we track fit on this card and can be run locally. The table on this page lists every one, with the memory it needs, the quantisation it runs at and an estimated generation speed.
B300— what is the largest AI model it can run?
The largest model in our catalogue that fits is Nemotron 3 Ultra at 550B parameters, compressed to Q3_K_M. It generates roughly 92.4 tokens per second and needs about 239.6 GB of the card's memory.
B300— how many tokens per second does it produce?
It depends on the model. The fastest model we track here is Gemma 3 QAT 1B at about 3,388 tokens per second, while larger models run proportionally slower because each token requires reading the whole model out of memory once. Speeds are estimates for a single conversation at a time.
B300— can it run 7B models?
Yes. For example it runs Gemma 4 E4B at Q8_0, using about 9.8 GB of memory and generating around 753 tokens per second.
B300— can it run 13B models?
Yes. For example it runs DeepSeekMoE-16B at Q8_0, using about 17.5 GB of memory and generating around 1,176 tokens per second.
B300— can it run 30B models?
Yes. For example it runs ERNIE-4.5-VL-28B-A3B at Q8_0, using about 29.2 GB of memory and generating around 672 tokens per second.
B300— can it run 70B models?
Yes. For example it runs Qwen3-Coder-Next at Q8_0, using about 80.7 GB of memory and generating around 235 tokens per second.
B300— how much memory does it have?
This card has 288 GB of HBM3e. Around a tenth is reserved by the inference runtime and the driver, leaving roughly 259.2 GB available for a model and its conversation.
The other direction
Looking at it from the other side?
This page starts from the hardware. If you already know which model you want and need to know what it takes to run it, start from the model instead.