Desmond

System Requirements

NOTE: Available on Linux only. See System Requirements for Linux.

	Required	Considerations
Driver	The driver (master job) must run for the complete duration of the job without being interrupted. This means the computing resource on which it runs cannot be a spot or preemptible cloud instance. These nodes can be pre-empted (terminated) and if that happens your whole job will be lost. The -DRIVERHOST argument determines where the driver runs. Select a host entry that is for an on-demand (i.e. not preemptible) node type.	If sufficient licenses and computational resources are available to run multiple Active Learning Glide jobs simultaneously, it is recommended to configure the driver host entry so that it requests an entire node, to avoid multiple drivers potentially using the same node and scratch filesystem, and thereby doubling (or more) the space requirement.
Processor (CPU)	x86_64 compatible processor	For large jobs, computing on a cluster with a queueing system is recommended, with the following hardware components: A highly capable file server for the external network. Shared storage for the intra-cluster network, to reduce traffic to and from the external network. Fast processors, large memory, and high-quality motherboards and network interfaces, especially on the management nodes.
System memory (RAM)	64 GB memory for the entire node	RAM is not related to the input file size, only the disk space is related.
Disk space		The amount of scratch space required on the DRIVERHOST is relative to the size of the input ligand file. See example below.

(General-purpose computing on graphics processing units)

Architecture	Server / HPC	Workstation
Volta	Tesla V100
Turing	Tesla T4	Quadro RTX 5000
Ampere	A100	RTX A4000 RTX A5000
Ada Lovelace	L4	RTX 4000 Ada RTX 2000 Ada
Hopper	H100
Blackwell	B200 (SXM)	RTX PRO 6000 Blackwell RTX PRO 4000 Blackwell

Unless otherwise specified, we only support and test on the PCIe variant of the cards listed above.
All form factors and editions (SFF, Workstation, Server, etc.) of listed GPU models are supported unless explicitly noted.
Please note that published Desmond performance data for supported cards / FEP+ performance data for supported cards performance benchmarks were conducted using a specific GPU edition and that actual performance may vary depending on the hardware variant.
To check the compute capability of NVIDIA cards, see NVIDIA CUDA GPU Compute Capability.

Schrödinger's Desmond GPU-based products—including Desmond, WaterMap, and FEP+—utilize a token-based licensing model scaled by the computational power (CUDA cores) of the underlying hardware. A 2026 update simplifies the previous decimal-based system by rounding the tokens in a more standardized fashion. See License Usage per GPU Calculator.

We support only the NVIDIA recommended / certified / production branch Linux drivers for these cards with minimum CUDA version 12.9. Download from NVIDIA's Drivers webpage.

We support NVIDIA’s Multi-Instance GPU (MIG) feature. Please make sure that MIG-enabled GPUs are being used in conjunction with a queueing system which supports this feature, see How do I make use of Multi-instance GPUs (MIG).

For information on pre-configured Schrödinger compatible GPU boxes see this article.

Pascal GPU Architecture Deprecation in 26-3: Support for all Pascal architecture GPU cards (Tesla P40, Tesla P100, and Quadro P5000) is deprecated starting in the 26-3 release. We still expect our software to run on this architecture until we upgrade to CUDA 13.
Volta GPU Architecture Deprecation & CUDA 13 upgrade: GPU compute nodes and workstations will need a CUDA 13.x compatible NVIDIA driver (>= 580.XX.YY) installed to run our GPU enabled products starting in release 27-1. As a result of this CUDA upgrade, support for all Volta architecture GPU cards (Tesla V100) will be deprecated in the 27-1 release.
Standard support does not cover consumer-level GPU cards such as GeForce GTX cards. Learn more about our rigorous validation process and why we exclusively support professional-grade NVIDIA hardware in this article.
If you already have another NVIDIA GPGPU and would like to know if we have experience with it, please contact our support at help@schrodinger.com.