Terray Therapeutics needed to search an almost infinite chemical space with generative models, but multi-node training, terabyte-scale datasets, and iterative hyperparameter experiments were too cumbersome on a mix of on-premises and generic cloud systems.
Reported impact
Terray increased resource utilization by 4x compared with alternative clouds, cut model training from a week to a day, and later used cuEquivariance to accelerate TerraBind by 3x to 4x with about 70% cost savings, giving researchers far more room to search for rare high-value molecules.
Terray built the COATI and TerraBind model-development workflow on NVIDIA DGX Cloud, NVIDIA Base Command Platform, and later NVIDIA cuEquivariance. The company starts from a massive experimental engine that measures hundreds of millions of molecule-target interactions daily and stores more than 50 billion biophysical measurements. Those data are used to train COATI, a multimodal encoder-decoder that translates molecules into numerical representations and back again so the company can generate new compounds with desired properties. DGX Cloud simplified multi-node training and Base Command Platform standardized job launch and orchestration, which let Terray move from manual remote-cluster management to push-button 32-GPU experiments. Terray then accelerated TerraBind models further with cuEquivariance, improving potency-prediction performance while cutting the time and cost required to iterate on new drug candidates.
Terray's experimental measurements feed COATI training on DGX Cloud, Base Command orchestrates large distributed runs, generated molecule embeddings flow into screening and optimization loops, and cuEquivariance-accelerated TerraBind models return potency and structure predictions for new candidates.
DGX Cloud offered us a solution that worked seamlessly with the ease and simplicity of cloud.
It used to take us a week to train a model, and we were getting it done in a day.
Terray Therapeutics has further advanced its computational capabilities by integrating NVIDIA's cuEquivariance technology into its proprietary TerraBind models.
The Terray platform measures hundreds of millions of interactions between small molecules and biological targets daily, with a growing database of 50 billion experimental biophysical measurements.
Review how this reference was scored and trace each normalized claim back to its documented source.
Scoring methodology
Customer Reference Value measures usefulness as a customer reference. Evidence Confidence measures the quality and traceability of available evidence, not independent verification.
Traceability
Field coverage shows which facts each source supports. Dates are shown only when documented by the source.
numberOfEmployees":{"value":142,"@type":"QuantitativeValue"}
Related
NVIDIA Sample Sep 22, 2026 DELFI Diagnostics DELFI Diagnostics needed to speed genomic analysis for cancer screening and therapeutic monitoring, but CPU-bound alignment and somatic variant-calling steps created turnaround-time and cost bottlenecks that limited how… 10–20 business days DeepSomatic · MutectCaller · NVIDIA Parabricks NVIDIA Sample Jul 10, 2026 Astellas Pharma Inc. Astellas needed to shorten the slow handoff between experimental scientists and computational chemists in early drug discovery so more researchers could use structure prediction and affinity screening without waiting on… More than 50 regular users and over 2,000 total users within four months NVIDIA BioNeMo · NVIDIA Boltz-2 NIM · NVIDIA DGX H100/H200 NVIDIA Sample Apr 23, 2026 Biofy Hospitals needed pathogen identification and antibiotic-resistance analysis much faster than five-day culture and antibiogram workflows, but CPU-bound genomic processing and limited AI expertise made it hard to turn… 100,000+ synthetic DNA sequences generated by the biology model Abby · H100 Tensor Core GPUs · NVIDIA CUDA-accelerated libraries 