Follow Us:
+62 21 6124030
Tesla M40 Passive 24GB

 

TESLA GPU ACCELERATORS FOR SERVERS

Accelerate your most demanding HPC, hyperscale, and enterprise data center workloads with NVIDIA® Tesla® GPU accelerators. Scientists can now crunch through petabytes of data up to 10x faster than with CPUs in applications ranging from energy exploration to deep learning. Plus, Tesla accelerators deliver the horsepower needed to run bigger simulations faster than ever before. For enterprises deploying VDI, Tesla accelerators are perfect for accelerating virtual desktops to any user, anywhere.

 

DEEP LEARNING INFERENCING WITH TESLA M40

Deep learning is redefining what's possible. From early-stage startups to large web service providers, deep learning has become the fundamental building block in delivering amazing solutions for end users. Today's leading deep learning models typically take days to weeks to train, forcing data scientists to make compromises between accuracy and time to deployment. The NVIDIA® Tesla® M40 GPU accelerator is the world's fastest accelerator for deep learning training, purpose-built to dramatically reduce training time.  Training neural networks demands a lot GPU memory, the Tesla M40 is built to handle these workloads for more accurate speech and image recognition and deeper understanding of video and natural language content.

 

TECHNICAL SPECIFICATIONS

Tesla M40 24GB

Tesla M40

Tesla M4

Peak single-precision floating point performance (board)

~7 Tflops

~7 Tflops

~2.2 Tflops

Number of GPUs

1 x GM200

1 x GM200

1 x GM206

Number of CUDA cores

3072

3072

1024

Memory size per board (GDDR5)

24GB

12GB

4GB

Memory Interface

384-bit

384-bit

128-bit

Memory bandwidth 

288 Gbytes/sec

288 Gbytes/sec

88 Gbytes/sec

Thermal Solution

Passive

Passive

Passive

Max Power Consumption

250W

250W

50W-75W

Form Factor

4.376"H x 10.5" L

4.376"H x 10.5" L

Low Profile

System

 NVIDIA Certified Servers NVIDIA Certified Servers

NVIDIA Certified Servers

 * Please check with your server vendor for the CPU power connector usage before turn on the server to avoid potential damage.