Free tools Windows power users keep installed
One-click scans. No signup required.
The Picture-Level Benchmark (PLB) was a historical graphics-performance benchmark that measured how quickly a computer drew specified complete pictures on CRT-based systems. Its standardized tests made performance comparisons possible across hardware platforms, but PLB did not measure image or display quality. The project retired in 2000.
What did the Picture-Level Benchmark measure?
PLB timed the drawing of complete test pictures, including workloads used with engineering workstations and personal computers. It was designed to compare graphics display performance, not to judge how accurate, sharp, or pleasing the resulting image looked.
As an Amazon Associate I earn from qualifying purchases.
The benchmark’s central idea was to use defined picture workloads and consistent timing and reporting rules. A PLB result therefore described performance on a particular test and system configuration—not a general rating of every graphics task a computer could perform.
How did PLB work?
PLB used three formats to define tests, measure performance, and present results. Its benchmark program processed test files, but the program was platform-dependent and had to be adapted to a particular hardware configuration.
#1 Best Overall
- Benchmark Interchange Format (BIF): specified the picture geometry and actions to be performed.
- Benchmark Timing Methodology (BTM): standardized how performance was measured.
- Benchmark Reporting Format (BRF): provided a consistent way to report results and system details.
The most representative approach was to convert a user’s own application into BIF and run it using a vendor’s PLB port. Because that took expertise and effort, PLB also provided standard files based on common applications. Implementation tests and standard benchmark files helped developers test and compare platform ports.
How should you read PLB scores?
PLB results used distinct labels for unoptimized and tuned runs, as well as composite figures for two workload categories. The distinction matters: an optimized result reflects tuning for a platform and does not show what an application would achieve without similar work.
Rank #2
| Result | What it represents | How to interpret it |
|---|---|---|
| PLB_lit | A literal benchmark run without application- or platform-specific optimizations. | Useful for understanding the run without those optimizations. |
| PLB_opt | A result tuned for the platform’s hardware and graphics API; tuning could change drawing order, primitives, or other implementation details. | Represents optimized performance under those conditions, not necessarily performance from an untuned application. |
| PLBwire93 | A geometric-mean composite of literal and optimized results for the standard wireframe tests. | Summarizes the wireframe category; it is not a score for every workload. |
| PLBsurf93 | A geometric-mean composite of literal and optimized results for the standard surface tests. | Summarizes the surface category; it is not a score for every workload. |
For PLB scores, higher means better performance: less time to display the picture under test. For a meaningful comparison, match the benchmark file and category, and check hardware, software configuration, graphics API, and optimization status. The standardized report format included configuration and test details to help readers assess whether results were comparable. SPEC’s archived PLB overview describes the result types and reporting approach.
What workloads did PLB include?
The standard workload set changed over the benchmark’s history. Digital Equipment Corporation’s April 1992 DECstation report listed five then-approved files; SPEC’s August 1996 overview described an eight-file suite divided into three categories. These are snapshots from different dates, not necessarily contradictory accounts.
Rank #3
| Historical source and date | Workloads or suite description |
|---|---|
| Digital Equipment Corporation, April 1992 | pc_board (2-D electrical CAD); sys_chassis (3-D wireframe chassis); cyl_head (3-D engine-cylinder-head model); head (laser-scanned human-head model); and shuttle (low-end 3-D simulation). |
| SPEC, August 1996 | An eight-file standard suite divided into 3-D wireframe, 3-D surface, and “other” categories. |
The 1992 file descriptions come from Digital’s DECstation report; the later suite description is in SPEC’s archived overview.
What PLB scores did not tell you
A PLB score was evidence about the speed of the reported test on the reported configuration. It was not evidence that a monitor or display produced a better-looking image. The PLB project group stated in SPEC’s archived overview: “The PLB allows buyers to compare performance, it does not address the issue of display quality.”
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.When did PLB end?
SPEC’s retirement notice says the PLB project group voted to disband and retire the benchmark in 2000. Interest had declined as vendors shifted toward API-level benchmarks such as Viewperf and GLperf, as well as application-level benchmarks. The notice says PLB was last updated in 1993 and that its information, standard files, and archived results would remain available through its page at that time. That historical notice does not establish whether downloads remain available today. Read SPEC’s PLB retirement notice.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteQuick Recap
Best Value
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
