
TRIPS
Experimental EDGE processor from University of Texas at Austin
TRIPS is a microprocessor architecture based on Explicit Data Graph Execution. Developed by the University of Texas at Austin with IBM, Intel, and Sun Microsystems, it aims to boost performance by grouping instructions into hyper-blocks for parallel processing.
- Uses Explicit Data Graph Execution to group instructions
- Prototype has 16 functional units per core
- Compilers create hyper-blocks of 128 instructions
- Aims for 1 TFLOP performance on a single processor
- General purpose units allow any instruction on any core
What it is
TRIPS is a microprocessor design that uses Explicit Data Graph Execution. It was created by researchers at the University of Texas at Austin alongside IBM, Intel, and Sun Microsystems. The system groups instructions into large blocks called graphs. These graphs run on independent processing elements. This approach keeps data close to the processor in high speed memory. It reduces the need for expensive data reads and writes. The prototype contains sixteen such elements. The goal was to reach one teraflop of performance on a single chip.
How it came about
The TRIPS team began building a prototype chip in 2003. Papers published between 2003 and 2006 outlined the performance targets. The design addresses bottlenecks found in modern systems. It moves away from linear instruction streams. Instead, it treats related instructions as blocks. By 2008, graphics processors had already passed the teraflop mark for specific tasks. Traditional central processing units remained far below this level for single applications.
Inside the chip
The architecture relies on compilers to find code blocks that share data. These become hyper-blocks fed into the CPU. Each block runs in a single functional unit with private registers. This isolation prevents extra memory fetching. The compiler manages data hand-offs between blocks explicitly. Functional units are general purpose. Any instruction can run on any core. This polymorphic design avoids idle units during mixed workloads. It allows scaling by adding more cores without changing the scheduler logic.
How it was made
The prototype implementation features two complete cores. Each core has sixteen functional units arranged four wide and four deep. Compilers create hyper-blocks of 128 instructions each. The system keeps eight blocks active simultaneously. This allows 1,024 instructions per core to be in flight. The basic design supports interconnecting up to 32 chips. Such a configuration approaches 500 gigaflops of total performance.
Where it went
TRIPS targets high performance computing tasks. It aims to outperform traditional CPUs in varied workloads. Traditional chips struggle when specific units like floating point processors are busy. TRIPS distributes all tasks across all units. It may replace lower performance chips used for media processing. It is less likely to replace highly customised graphics processors. The design suits applications requiring high parallelism and data locality.
Versions and family
The prototype serves as the primary reference implementation. It demonstrates the scalability of the EDGE concept. A single core version would run slower but use the same logic. Designs can range from one core to hundreds. The number of cores determines the performance ceiling. No commercial successors or predecessors are listed in the provided material.
Why it matters
TRIPS introduced a method to bypass instruction scheduling limits. It shifts parallelism constraints from hardware to the compiler. This allows wider scaling than traditional four-wide designs. It reduces communication delays by keeping data local. The project highlighted the potential of graph-based execution. It showed how general purpose units could handle diverse tasks efficiently.
Questions people ask about TRIPS
What is the main idea behind TRIPS?
TRIPS uses Explicit Data Graph Execution to group related instructions into hyper-blocks. This allows the processor to run these blocks in parallel on independent units, reducing data movement and improving efficiency.
Who developed the TRIPS architecture?
A team at the University of Texas at Austin designed TRIPS. They worked in conjunction with major technology companies including IBM, Intel, and Sun Microsystems to develop the concept.
How does TRIPS handle different types of tasks?
It uses general purpose functional units. Any instruction can run on any core. This prevents idle units when workloads mix integer and floating point operations, unlike traditional CPUs with specialized units.
Good reading and sources
- TRIPS architecture on WikipediaEncyclopedia
- More photos on Wikimedia CommonsPhotos
- Wikidata recordOpen data
The facts on this page come from the Wikipedia article "TRIPS architecture", written by its contributors and used under the CC BY-SA 4.0 licence, and from its Wikidata record (CC0). EveryChip rewrote the text in its own words, with AI assistance, and compares every page with the article to make sure no passage is copied. It may be reused under the same CC BY-SA 4.0 licence with a link back.
It may contain mistakes, and figures change over time. Nothing here is investment or buying advice. If something is wrong, tell us. Last refreshed 2026-11-14.
Like TRIPS
All CPUs and microprocessors
🇺🇸 United StatesAlpha 21064
First commercial microprocessor implementing the Alpha instruction set architecture.CPU · Other · 1992
More from University of Texas at Austin
All of them
🇺🇸 United StatesRapidCAD
A 486DX in a 386 socket, with a dummy FPU chip to match.CPU · x86
WorldMOS Technology 8502
An 8-bit CPU from MOS Technology that powered the Commodore 128CPU · 8-bit
🇺🇸 United StatesRAD5500
Radiation-hardened PowerPC processor for space systemsCPU · Power