Skip to content

Nemotron 3.5 Bolt 74B A9B VRAM requirements

How much VRAM Nemotron 3.5 Bolt 74B A9B needs
4-bit weights: 42.3 GBWith 8K context: 43.3 GBWith 32K context: 43.5 GB8-bit, 8K context: 75.2 GB
Parameters
74.9B
Active per token
9B
Longest context
256K
Released
Oct 2026
License
OpenMDW 1.1
Strongest at
Chat, Coding, Agents, Docs

How much VRAM Nemotron 3.5 Bolt 74B A9B needs

Memory Nemotron 3.5 Bolt 74B A9B needs at each precision and context length
PrecisionWeightsWith 8KWith 32KWith 128K
16-bit139 GB141 GB141 GB141 GB
8-bit74.1 GB75.2 GB75.3 GB76.1 GB
6-bit57.2 GB58.2 GB58.4 GB59.2 GB
5-bit49.6 GB50.7 GB50.8 GB51.6 GB
4-bit42.3 GB43.3 GB43.5 GB44.3 GB
3-bit34.1 GB35.1 GB35.3 GB36.1 GB
2-bit26.2 GB27.2 GB27.4 GB28.1 GB
1.6-bit14.4 GB15.4 GB15.6 GB16.4 GB

One conversation, 16-bit context cache, runtime overhead included.

GPUs and Macs that run Nemotron 3.5 Bolt 74B A9B

The best way Nemotron 3.5 Bolt 74B A9B runs on common GPUs, Macs and AI boxes
MachineBest fitContextSpeedFirst replyReport
RTX 4060Too bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 4060 report
RTX 5060Too bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 5060 report
RTX 3070Too bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 3070 report
RTX 3080 10 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 3080 10 GB report
RTX 3060 12 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 3060 12 GB report
Arc B580Too bigPDF: Nemotron 3.5 Bolt 74B A9B on Arc B580 report
RTX 4070Too bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 4070 report
RTX 4070 SuperToo bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 4070 Super report
RTX 5070Too bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 5070 report
RTX 4060 Ti 16 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 4060 Ti 16 GB report
RTX 5060 Ti 16 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 5060 Ti 16 GB report
Arc A770 16 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on Arc A770 16 GB report
Radeon RX 9070 XTToo bigPDF: Nemotron 3.5 Bolt 74B A9B on Radeon RX 9070 XT report
RTX 4070 Ti SuperToo bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 4070 Ti Super report
RTX 4080 SuperToo bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 4080 Super report
RTX 5070 TiToo bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 5070 Ti report
RTX 5080Too bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 5080 report
Radeon RX 7900 XTXToo bigPDF: Nemotron 3.5 Bolt 74B A9B on Radeon RX 7900 XTX report
RTX 3090Too bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 3090 report
RTX 4090Too bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 4090 report
Radeon AI PRO R9700Too bigPDF: Nemotron 3.5 Bolt 74B A9B on Radeon AI PRO R9700 report
RTX 5090Too bigPDF: Nemotron 3.5 Bolt 74B A9B on RTX 5090 report
2× RTX 30904-bit216K72 tok/s1.0 sPDF: Nemotron 3.5 Bolt 74B A9B on 2× RTX 3090 report
2× RTX 40904-bit216K75 tok/s0.4 sPDF: Nemotron 3.5 Bolt 74B A9B on 2× RTX 4090 report
RTX PRO 5000 Blackwell 48 GB4-bit256K89 tok/s0.9 sPDF: Nemotron 3.5 Bolt 74B A9B on RTX PRO 5000 Blackwell 48 GB report
RTX 6000 Ada4-bit256K73 tok/s0.6 sPDF: Nemotron 3.5 Bolt 74B A9B on RTX 6000 Ada report
L40S4-bit256K69 tok/s1.1 sPDF: Nemotron 3.5 Bolt 74B A9B on L40S report
2× RTX 50906-bit248K89 tok/s0.3 sPDF: Nemotron 3.5 Bolt 74B A9B on 2× RTX 5090 report
4× RTX 30908-bit256K50 tok/s0.5 sPDF: Nemotron 3.5 Bolt 74B A9B on 4× RTX 3090 report
RTX PRO 6000 Blackwell8-bit256K77 tok/s0.5 sPDF: Nemotron 3.5 Bolt 74B A9B on RTX PRO 6000 Blackwell report
Mac mini M4 16 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on Mac mini M4 16 GB report
Mac mini M4 24 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on Mac mini M4 24 GB report
Mac mini M4 32 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on Mac mini M4 32 GB report
Mac mini M4 Pro 24 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on Mac mini M4 Pro 24 GB report
Mac mini M4 Pro 48 GBToo bigPDF: Nemotron 3.5 Bolt 74B A9B on Mac mini M4 Pro 48 GB report
Mac mini M4 Pro 64 GB4-bit256K27 tok/s20 sPDF: Nemotron 3.5 Bolt 74B A9B on Mac mini M4 Pro 64 GB report
M4 Max 128 GB8-bit256K28 tok/s9.9 sPDF: Nemotron 3.5 Bolt 74B A9B on M4 Max 128 GB report
M5 Max 128 GB8-bit256K30 tok/s2.9 sPDF: Nemotron 3.5 Bolt 74B A9B on M5 Max 128 GB report
M3 Ultra 96 GB6-bit256K38 tok/s5.5 sPDF: Nemotron 3.5 Bolt 74B A9B on M3 Ultra 96 GB report
M3 Ultra 256 GB8-bit256K33 tok/s5.5 sPDF: Nemotron 3.5 Bolt 74B A9B on M3 Ultra 256 GB report
M3 Ultra 512 GB8-bit256K33 tok/s5.5 sPDF: Nemotron 3.5 Bolt 74B A9B on M3 Ultra 512 GB report
M5 Ultra 512 GB8-bit256K40 tok/s1.7 sPDF: Nemotron 3.5 Bolt 74B A9B on M5 Ultra 512 GB report
Ryzen AI Max+ 395 64 GB4-bit256K22 tok/s5.7 sPDF: Nemotron 3.5 Bolt 74B A9B on Ryzen AI Max+ 395 64 GB report
Ryzen AI Max+ 395 128 GB8-bit256K14 tok/s5.7 sPDF: Nemotron 3.5 Bolt 74B A9B on Ryzen AI Max+ 395 128 GB report
RTX Spark8-bit256K19 tok/s2.0 sPDF: Nemotron 3.5 Bolt 74B A9B on RTX Spark report
DGX Spark8-bit256K18 tok/s2.0 sPDF: Nemotron 3.5 Bolt 74B A9B on DGX Spark report
Dual DGX Spark8-bit256K18 tok/s1.1 sPDF: Nemotron 3.5 Bolt 74B A9B on Dual DGX Spark report
DGX Station GB3008-bit256K138 tok/s0.2 sPDF: Nemotron 3.5 Bolt 74B A9B on DGX Station GB300 report
2× RTX PRO 6000 Blackwell8-bit256K77 tok/s0.3 sPDF: Nemotron 3.5 Bolt 74B A9B on 2× RTX PRO 6000 Blackwell report
4× RTX PRO 6000 Blackwell8-bit256K77 tok/s0.1 sPDF: Nemotron 3.5 Bolt 74B A9B on 4× RTX PRO 6000 Blackwell report
8× RTX PRO 6000 Blackwell8-bit256K77 tok/s0.1 sPDF: Nemotron 3.5 Bolt 74B A9B on 8× RTX PRO 6000 Blackwell report
A100 80 GB SXM8-bit112K83 tok/s1.3 sPDF: Nemotron 3.5 Bolt 74B A9B on A100 80 GB SXM report
H100 SXM8-bit112K106 tok/s0.4 sPDF: Nemotron 3.5 Bolt 74B A9B on H100 SXM report
4× H200 NVL8-bit256K122 tok/s0.1 sPDF: Nemotron 3.5 Bolt 74B A9B on 4× H200 NVL report
8× H200 NVL8-bit256K122 tok/s0.1 sPDF: Nemotron 3.5 Bolt 74B A9B on 8× H200 NVL report

Best fit is the highest quality, up to 8-bit, that leaves 16K of context. First reply is the wait for a 2K prompt. Speeds are rough, plus or minus 25%.

How to run it

Ollama
ollama run nemotron-3.5-bolt:74b-a9b
llama.cpp
llama-server -hf unsloth/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-GGUF
vLLM
vllm serve nvidia/NVIDIA-Nemotron-3.5-Bolt-74B-A9B-BF16

Questions

How much VRAM does Nemotron 3.5 Bolt 74B A9B need?

Nemotron 3.5 Bolt 74B A9B needs 42.3 GB for its weights at 4-bit, 43.3 GB with an 8K context and 43.5 GB with 32K. At 8-bit, 75.2 GB.

What is the minimum VRAM for Nemotron 3.5 Bolt 74B A9B?

About 35.1 GB at 3-bit with an 8K context, the lowest precision we recommend for a model this size. The smallest machine here that runs it without spilling is the 2× RTX 3090.

How big is Nemotron 3.5 Bolt 74B A9B?

74.9B parameters (9B active per token): 139 GB of weights at 16-bit, 42.3 GB at 4-bit.

Can Nemotron 3.5 Bolt 74B A9B run on an RTX 4090?

Not on one card. The smallest machine here that runs it is the 2× RTX 3090.

More models