PantheonSim / GPUs / NVIDIA A100-SXM4-40GB

Simulated NVIDIA A100-SXM4-40GB: a GPU profile for CUDA and ROCm testing

PantheonSim, the GPU simulator for CUDA and ROCm, simulates the NVIDIA A100-SXM4-40GB (Ampere, 8.0) as the profile nvidia/a100-sxm4-40gb. The A100 as an SXM4 module with 40 GB of HBM2, built for multi-GPU baseboards joined by NVLink. Its profile is verified, which means it was read from a physical card.

The profile at a glance

These are the numbers a program can ask this simulated GPU, and the values nvidia-smi show for it. They come from the profile file in the simulator's repository, pinned to one commit of it.

Property	Value
Profile id	`nvidia/a100-sxm4-40gb`
Vendor and architecture	NVIDIA, Ampere
Compute capability	`8.0`
Device memory	39.5 GiB, HBM
Streaming multiprocessors	108
Warp size	32
Max threads per multiprocessor	2048
Shared memory per block (opt-in maximum)	163 KiB
L2 cache	40 MiB
Memory bus width	5120-bit
Power limit (profile)	400 W
Maximum clocks (profile)	1410 MHz shader, 1215 MHz memory
ECC memory	yes
PCIe link reported	Gen4 x16
NVLink reported	NVLink3 x12
Tensor or matrix cores	yes
Verified (measured on a physical card)	yes

What this GPU is for

The A100 as an SXM4 module with 40 GB of HBM2, built for multi-GPU baseboards joined by NVLink.

Ampere covers two compute capabilities in these profiles: 8.0 for the GA100 data-center parts and 8.6 for the GA10x parts. The simulator runs the SASS a binary carries for the architecture and falls back to PTX when it carries none.

Among the 37 NVIDIA profiles it ranks 23rd by device memory (39.5 GiB) and 19th by SMs (108), counting the largest as first.

Is this profile verified?

Yes. This profile is marked verified, which means its values were read from a physical NVIDIA A100-SXM4-40GB by the simulator's characterization tools, not copied from a data sheet. Verified describes the profile's numbers: it says nothing about speed, because the simulator has no timing model.

Simulate this GPU

Build the simulator (see the docs) or open the playground, then:

nvcc -cudart shared -arch=compute_80 my_app.cu -o my_app
vgpu run --gpu nvidia/a100-sxm4-40gb ./my_app
vgpu shell -y --gpu nvidia/a100-sxm4-40gb --count 4
vgpu info --gpu nvidia/a100-sxm4-40gb

This profile is compute capability 8.0. A kernel built for compute_75 runs on it and on every other NVIDIA profile; a kernel built for a newer architecture is refused by a profile that is older than it. Link the CUDA runtime shared (-cudart shared) so the simulator can stand in for it.

To use it every time, put gpu: nvidia/a100-sxm4-40gb in a pantheonsim.yaml settings file, or in a GitHub Actions job give the PantheonSim action gpu: nvidia/a100-sxm4-40gb (see CUDA and HIP tests in GitHub Actions). The simulator checks behaviour and has no timing model, so a run on a physical GPU before a release still counts. The CUDA guide explains the rest.

Where the numbers come from

Each profile file begins with a note on its sources. This is the note for NVIDIA A100-SXM4-40GB, as written in the file:

Source note for nvidia/a100-sxm4-40gb

Generated by nvidia/tools/characterize.cu on a physical A100 SXM4 40GB (Lambda gpu_1x_a100_sxm4). The 80GB SXM4 variant is a separate profile. Every value below was read from the hardware, not a datasheet.

The GPU index lists every profile, and this one is measured on a physical card.

Other Ampere profiles: