Community resourceWorksheet

OCR H446 1.1.1 Von Neumann, Harvard and contemporary architectures

Part 8 of 12 · H446 1.1.1 · Structure and function of the processor

OCR H446 1.1.1: Von Neumann, Harvard and contemporary architectures. Students learn to make matched Von Neumann and Harvard comparisons and apply broad contemporary architecture models to suitable workloads.

Students will:

  • compare matching features of Von Neumann and Harvard architectures
  • apply contemporary architecture models to workloads
  • justify an architecture recommendation in context

Inside: 6 explanation cells, 1 multiple-choice question, 2 fill-in-the-blanks cells and 3 written answers. 21 marks, about 30 to 40 minutes.

Series: H446 1.1.1 · Structure and function of the processor, part 8 of 12.

Shared by Coding PathwayVerified teacher

  • 12 cells
  • About 30 minutes
  • CC BY-SA 4.0
  • Shared 31 Aug 2026
  • Updated 3 Sept 2026

Preview

The whole resource, exactly as a class sees it. Answers and marking are held back.

Von Neumann, Harvard and contemporary architectures

Architecture describes how processor components, memories and communication routes are organised. A complete comparison discusses the same feature for both designs.

By the end, you will be able to

  • compare Von Neumann and Harvard storage and access;
  • explain benefits and trade-offs;
  • use information supplied about a contemporary architecture without treating every modern system as identical.

Compare arrangements

Von Neumann, Harvard and varied contemporary approachesVon Neumannprocessorone shared memoryinstructions and datashared route can be contendedHarvardprocessorinstructionmemorydatamemoryseparate routes allow concurrent accessContemporary architecture is a collection, not one designSIMDone instruction, multiple dataMIMDmultiple instructions, multiple dataDistributed computingwork across networked computersThese are broad examples for comparison, not a claim that every modern system is organised identically.

A Von Neumann design stores instructions and data in shared memory and commonly uses shared buses. Instruction and data access can therefore compete for the same route, sometimes called the Von Neumann bottleneck. The shared space is flexible because it is not rigidly divided.

A Harvard design uses separate instruction and data memories and buses. It can fetch an instruction and access data concurrently, and each memory can be designed for its purpose. The extra paths and memories can increase hardware cost and complexity.

Supplied contemporary examples

A question may supply information about a contemporary arrangement and ask you to interpret or compare it. For example:

Supplied exampleMeaning
SIMDone instruction is applied to multiple data items
MIMDmultiple processors or cores perform different instructions on different data
distributed computingwork is divided across networked computers

Use the information supplied in the question. These names are examples, not a compulsory list of modern designs to memorise. A modern system may also combine features, such as separate instruction and data caches with shared main memory.

Worked comparison point

Incomplete: ‘Harvard has separate memory.’ Complete: ‘Harvard stores instructions and data in separate memories, whereas Von Neumann stores both in one shared memory. Harvard can therefore fetch an instruction while accessing data, although the extra memory paths increase hardware complexity.’

The second response compares the same feature on both sides, explains an effect and includes a limitation.

Multiple choice1 mark

Which is a matched comparison?

  • AVon Neumann uses shared instruction/data memory whereas Harvard uses separate memories
  • BHarvard is newer
  • CHarvard is always faster
  • DVon Neumann has no control unit
Fill in the blanks3 marks
Von Neumann uses gap 1 instruction and data memory. Harvard uses gap 2 memories and buses, which can permit gap 3 access.
  • shared
  • separate
  • concurrent
  • random
Written answer4 marks

Compare two differences between Von Neumann and Harvard architectures.

Use two criteria. For each, state both sides of the comparison.

Students type their answer here.

Written answer4 marks

The table supplies three contemporary examples. Choose one and explain how it goes beyond the simple model of one processor handling one instruction stream and one data item at a time.

Name the supplied example, use its definition and make one clear comparison. You are applying the table, not recalling an unnamed proprietary design.

Students type their answer here.

Checkpoint

Complete the architecture comparison from memory. There is no answer bank.

Fill in the blanks4 marks
Von Neumann uses checkpoint gap 1 memory for instructions and data. Harvard uses checkpoint gap 2 instruction and data memories with separate communication paths, so the processor may access both checkpoint gap 3. The extra paths can increase hardware checkpoint gap 4.

Apply the comparison independently

Use the matched architecture model and contemporary examples to make a justified recommendation in a new context.

Written answer5 marks

A signal-processing device needs predictable, simultaneous instruction and data access but has a generous hardware budget. Recommend an architecture.

Apply access arrangement and one trade-off before concluding.

Students type their answer here.

Consolidate your understanding

Check that every architecture comparison uses the same criterion on both sides and every recommendation follows from the stated workload.