3 ms·
80 GB per GPU and the consumer GPU is purposefully lower memory to induce demand for server grade.
by spullara 3y ago
80 GB per GPU and the consumer GPU is purposefully lower memory to induce demand for server grade.
- airgapstopgap 3y agoOn top of contracts strictly penalizing utilization of consumer GPUs in data centers, at that! Even with the memory, bandwidth etc. handicaps, servers with 4090/3090s would have been competitive for many ML tasks.