NVIDIA B300 SXM 288GB
Official
Blackwell Ultra SXM 2025 4nm
FP32
75.0
TFLOPS
VRAM
288 GB
HBM3e
TDP
1.4 kW
53.6 TFLOPS/kW
Bandwidth
8.0 TB/s
memory
Performance Metrics
Peak theoretical throughput by precision type
| Precision | Peak TFLOPS | Efficiency |
|---|---|---|
| FP6 | 4500.0 | 3.214 TFLOPS/W |
| INT8 | 4500.0 | 3.214 TFLOPS/W |
| FP8 | 4500.0 | 3.214 TFLOPS/W |
| FP4 | 13500.0 | 9.643 TFLOPS/W |
| FP16 | 2250.0 | 1.607 TFLOPS/W |
| BF16 | 2250.0 | 1.607 TFLOPS/W |
| TF32 | 1125.0 | 0.804 TFLOPS/W |
| FP32 | 75.0 | 0.054 TFLOPS/W |
| FP64 | 1.3 | 0.001 TFLOPS/W |
FP8 Efficiency
3.214 TFLOPS/W
4500.0 TFLOPS / 1400W
FP16 Efficiency
1.607 TFLOPS/W
2250.0 TFLOPS / 1400W
FP32 Efficiency
0.054 TFLOPS/W
75.0 TFLOPS / 1400W
Power Specifications
TDP
1.4 kW
Max Power
1.6 kW
Power Connector
PCIe 16-pin
Cooling
Liquid
Memory Specifications
Capacity
288 GB
Type
HBM3e
Bandwidth
8000 GB/s
Interface
--
Hardware & Design
Form Factor
SXM
Architecture
Blackwell Ultra
Process Node
4nm
Launch Year
2025
Variant
SXM 288GB
Market Segment
Data Center
Common Use Cases
General Compute AI/ML Workloads Data Processing
The NVIDIA B300 is optimized for high-performance computing tasks with Blackwell Ultra architecture delivering 75 TFLOPS of compute power.
Where to Rent
Prices shown are approximate. Click through for current pricing.
Similar GPUs
NVIDIA GB300 Grace Blackwell Ultra Superchip 252GB
Superchip · 2025
FP32: 80.0 TFLOPS
Compare vs GB300Systems Using This GPU
Pre-configured systems featuring the NVIDIA B300
| System | Peak Performance |
|---|---|
NVIDIA DGX B300 DGX · rack · 2025 | 120.00 PFLOPS |
NVIDIA DGX GB300 NVL72 NVL · rack · 2025 | 1080.00 PFLOPS |
Stay Updated on GPU Releases
Get notified when new GPUs are added or specifications are updated.
Loading verification...
No spam, unsubscribe anytime.