GPU
A GPU is a specialized chip used for fast parallel computation.
Category
Compute Economics
These terms cover hardware, reuse, bundling requests, and keeping systems busy.
How to recognize this theme
Terms that show up when you optimize model serving cost and speed.
In a daily board, this category groups terms by their shared role. Look for four cards that describe the same mechanism, risk area, or workflow rather than four words that merely sound similar.
Educational context
These entries are vocabulary notes for learning. They are not project endorsements, token recommendations, exchange rankings, or trading signals.
Cache
A cache stores data or output so it can be reused without recomputing it.
Batching
Batching groups requests together so a system can process them more efficiently.
Parallelism
Parallelism is the practice of running multiple operations at the same time.
GPU Hour
A GPU hour measures the use of one graphics processor for one hour of computing work.
Power Usage Effectiveness
Power usage effectiveness compares total facility energy with the energy delivered to computing equipment.
Capacity Reservation
A capacity reservation secures computing resources in advance for a workload that needs predictable availability.
Spot Instance
A spot instance is discounted cloud capacity that the provider may reclaim when demand changes.
Latency
Latency is the time between a request and a response.
Throughput
Throughput is how much output a system can produce over a given period.
Batching
Batching groups multiple inputs into one processing pass.
Quantization
Quantization compresses model numbers into lower precision to save memory and speed up inference.
GPU
A GPU is specialized hardware built to run many calculations in parallel.
Throughput
Throughput is the amount of work a system completes over a given time period.
Utilization
Utilization is the share of available compute actively being used.
Autoscaling
Autoscaling adjusts system resources automatically as demand rises or falls.