Larger local models
Evaluate model sizes that cannot fit within the current 24 GB GPU-memory limit.
SUPPORT FARYNTH RESEARCH
Help expand what Farynth can study.
Farynth currently conducts research on locally operated hardware. Support toward our present compute goal would help us build a more capable dedicated research station for larger models and more demanding experiments.
CURRENT GOAL01
RESEARCH COMPUTE
From 24 GB to 96 GB.
Our current station uses a single 24 GB GPU. The present goal is an NVIDIA RTX PRO 6000 Blackwell Workstation Edition with 96 GB of on-card memory—four times the capacity available to our current research environment.
Current24 GBWorkstation GPU memory
→ Target96 GBWorkstation GPU memory
Current hardware goal $16,000
Support through Stripe Secure one-time checkout. Choose the amount you wish to contribute.
WHAT IT ENABLES02
More capacity, used with purpose.
Broader studies
Support more demanding experimental conditions, longer contexts, and deeper local analysis.
Independent operation
Keep research execution on Farynth-operated hardware with direct control over the environment and evidence.
TRANSPARENCY03
What support means.
Farynth LLC accepts voluntary support payments toward its independent research work. They are not tax-deductible charitable gifts, purchases, preorders, investments, or exchanges for equity.
Secure checkoutPayments are completed on Stripe’s hosted checkout. Farynth does not receive or store full card details.
Payment termsBecause no product, service, or other benefit is provided in return, completed support payments are generally final. Duplicate, unauthorized, or processor-error payments will be reviewed, and applicable law always controls.
After the goalFarynth will publish an update showing what was acquired and the research capacity it added.
Ongoing recordProgress will reflect completed contributions and will be updated from the payment record.