What is Throughput?
The amount of work, requests or data processed per unit time, a different metric from latency, which describes how fast one task completes.
Field: Performance Analysis
Throughput in the explanations
Little's law states that average WIP equals throughput (output per unit time) times cycle time.
Which correctly states Little's law, relating work in progr…Throughput measures how much work can be done; latency measures how quickly a single piece of work completes.
Which correctly describes throughput and latency as server …DDR transfers data on both the rising and falling clock edges, doubling throughput at the same clock frequency.
What does 'DDR (double data rate)' in DDR SDRAM mean?
Questions that mention Throughput (15)
- Which model plots whether a program's performance is limited by compute throughput or memory bandwidth, using arithmetic intensity (operations per byte) on the horizontal axis?
- Which correctly states Little's law, relating work in progress (WIP), throughput and cycle time in a factory?
- Which correctly describes throughput and latency as server performance measures?
- What does 'DDR (double data rate)' in DDR SDRAM mean?
- What is the main purpose of dividing the inside of a DRAM into multiple banks?
- What is the main purpose of pipelining in a CPU?
- What is MOCVD (metal-organic chemical vapor deposition) mainly used for?
- What is the main cause of the stochastic fluctuations that are a problem in EUV lithography?
- What is an advantage of single-wafer cleaning tools, which process one wafer at a time?
- What is a point of caution for batch cleaning, in which many wafers are immersed in the same tank of cleaning solution?
- Which method bonds chips to substrates one at a time while heating and applying pressure, and is used to connect thin chips with large warpage and fine-pitch bumps?
- Which is a challenge of wafer-on-wafer (W2W) bonding, which bonds entire wafers together?
- How do the design philosophies of CPUs and GPUs differ?
- What is a drawback of making the inference batch size too large?
- Which is closest to the efficiency of the 128b/130b encoding adopted from PCI Express 3.0 onward?
Related terms
Branch prediction Arithmetic intensity Roofline model Little's law DDR SDRAM Bank Pipelining MOCVD (metal-organic chemical vapor deposition) GaN (gallium nitride) EUV lithography