Tenstorrent Software Engineer Interview Guide

The 60-second brief

A Software Engineer at Tenstorrent operates at the highly critical intersection of cutting-edge artificial intelligence models and custom, high-performance silicon. Unlike traditional software roles that exist purely in the application layer, engineering at Tenstorrent requires a deep understanding of hardware-software co-design. The software team is responsible for building the compilers, runtime systems, toolchains, and programming models that translate complex neural networks into highly optimized instructions for our custom RISC-V CPUs and AI accelerators. The impact of this role cannot be overstated. As AI models continue to scale exponentially, traditional general-purpose processors are hitting physical and architectural limits. Tenstorrent is redefining the computing paradigm by building hardware from the ground up to handle massive tensor operations.

This guide is scoped to a Software Engineer candidate at Tenstorrent.

Focus areas

• Compiler Design
• Code Generation
• RISC-V Architecture

The Tenstorrent interview process

Tenstorrent candidates report 6 rounds over 4-6 weeks. The stages below are what candidates describe, not a published process.

Recruiter Conversation

Initial conversation to align on your background, interests, and the team's current needs.

What to demonstrate

• Initial conversation to align on your background, interests, and the team's current needs
• Depth in Compiler Design

How to prepare

• Be able to walk your CV end to end in two minutes, and say why this company specifically.
• Have your salary expectations, notice period and location constraints ready, and ask for the rest of the loop in writing.

Technical Rounds

Series of rigorous technical interviews conducted by senior engineers and technical leads.

What to demonstrate

• Series of rigorous technical interviews conducted by senior engineers and technical leads
• Depth in Compiler Design

How to prepare

• Answer aloud and timed: How do branch predictors work, and how can software design minimize branch misprediction penalties?
• Answer aloud and timed: Describe the memory hierarchy and explain how cache coherence is maintained in multi-core systems.

Live Coding Challenges

Expect to solve coding challenges in real-time during technical interviews.

What to demonstrate

• Expect to solve coding challenges in real-time during technical interviews
• Depth in Compiler Design

How to prepare

• Answer aloud and timed: How would you design an API to send encoded data efficiently between multiple RISC-V processors?
• Answer aloud and timed: Implement a basic hash map or associative array from scratch in C++ without using standard library containers.

System Design Discussions

Engage in discussions about system design and architecture.

What to demonstrate

• Engage in discussions about system design and architecture
• Depth in Compiler Design

How to prepare

• Answer aloud and timed: Write an algorithm to perform efficient matrix multiplication, and explain how you would optimize it for cache locality.
• Answer aloud and timed: How do you balance and traverse a binary search tree? Implement a node deletion algorithm.

Behavioral Evaluations

Participate in evaluations focusing on behavioral aspects and team fit.

What to demonstrate

• Participate in evaluations focusing on behavioral aspects and team fit
• Depth in Compiler Design

How to prepare

• Prepare three examples from your own work, each with a decision you made and an outcome you can quantify.
• Re-read the description of the behavioral evaluations above and write down what you would ask to confirm before it.

Final Decision

Receive the final decision regarding your application after the interview rounds.

What to demonstrate

• Receive the final decision regarding your application after the interview rounds
• Depth in Compiler Design

How to prepare

• Answer aloud and timed: Describe your proudest engineering project. What was the most difficult technical challenge, and how did you overcome it?
• Answer aloud and timed: Tell me about a time when you had a significant technical disagreement with a teammate or manager. How did you resolve it?

Where candidates lose points

Think from first principles

When presented with a problem, do not just jump to a high-level library or framework. Explain how you would solve it at the system, memory, or hardware level.

Trace your code's hardware footprint

During coding interviews, explain how your data structures impact cache locality, memory bandwidth, and CPU instruction pipelines.

Communicate your thought process clearly

Our interviewers value how you approach a problem just as much as the final solution. Talk through your assumptions, trade-offs, and design choices in real time.

Questions & practice

Describe the memory hierarchy and explain how cache coherence is maintained in multi-core systems.

Medium · Computer Architecture & Low-Level System

Approach

1. Say what the runtime actually does before reasoning about the code.
2. Name what is shared across threads and what owns each piece of state.
3. Identify the window where an invariant is briefly untrue.
4. Distinguish a value from a reference to it, and say which one you handed out.

Follow-up

• What happens if two callers reach this at the same time?
• Where could this allocate more than you expect?

Implement a basic hash map or associative array from scratch in C++ without using standard library containers.

Medium · Data Structures & Algorithms

Approach

1. Restate the input: its shape, its size, and what is guaranteed about it.
2. Name the brute-force solution and its complexity before improving on it.
3. Choose the data structure from the access pattern, not from familiarity.
4. State the target complexity and say which constraint rules the naive version out.

Follow-up

• How does this change if the input no longer fits in memory?
• What is the worst case, and how likely is it on real data?

Write an algorithm to perform efficient matrix multiplication, and explain how you would optimize it for cache

Medium · Data Structures & Algorithms

Write an algorithm to perform efficient matrix multiplication, and explain how you would optimize it for cache locality.

Approach

1. Restate the input: its shape, its size, and what is guaranteed about it.
2. Name the brute-force solution and its complexity before improving on it.
3. Choose the data structure from the access pattern, not from familiarity.
4. State the target complexity and say which constraint rules the naive version out.

Follow-up

• How does this change if the input no longer fits in memory?
• What is the worst case, and how likely is it on real data?

How do you balance and traverse a binary search tree? Implement a node deletion algorithm.

Medium · Data Structures & Algorithms

Approach

1. Restate the input: its shape, its size, and what is guaranteed about it.
2. Name the brute-force solution and its complexity before improving on it.
3. Choose the data structure from the access pattern, not from familiarity.
4. State the target complexity and say which constraint rules the naive version out.

Follow-up

• How does this change if the input no longer fits in memory?
• What is the worst case, and how likely is it on real data?

Describe how you would implement a custom generic class in C++ or Rust to manage low-level memory buffers.

Medium · Data Structures & Algorithms

Approach

1. Say what the runtime actually does before reasoning about the code.
2. Name what is shared across threads and what owns each piece of state.
3. Identify the window where an invariant is briefly untrue.
4. Distinguish a value from a reference to it, and say which one you handed out.

Follow-up

• What happens if two callers reach this at the same time?
• Where could this allocate more than you expect?

Solve a medium-level algorithmic problem involving graph traversal or dynamic programming under tight time com

Medium · Data Structures & Algorithms

Solve a medium-level algorithmic problem involving graph traversal or dynamic programming under tight time complexity constraints.

Approach

1. Restate the input: its shape, its size, and what is guaranteed about it.
2. Name the brute-force solution and its complexity before improving on it.
3. Choose the data structure from the access pattern, not from familiarity.
4. State the target complexity and say which constraint rules the naive version out.

Follow-up

• How does this change if the input no longer fits in memory?
• What is the worst case, and how likely is it on real data?

Explain why the owner filter ignores the listing index

Medium · composite index · query plans · sort elimination

The only index on resource is (tenant_id, status, updated_at DESC, resource_id DESC). A new endpoint returns one user's resources across all statuses, newest created first: WHERE tenant_id = $1 AND owner_user_id = $2 ORDER BY created_at DESC LIMIT 20. On a tenant with 2M rows it takes 900 ms and EXPLAIN shows a sort above a large scan. Explain precisely why the existing index cannot serve it, give the index that can, and state which of these the new index still will not help: owner_user_id alone across tenants; the same query ordered by updated_at. PostgreSQL 16.

Approach

1. Separate the two jobs an index does. For filtering, a composite btree is seekable only on a left prefix, so with no predicate on status the scan can at best range over tenant_id and test owner_user_id per row; PostgreSQL 16 has no btree skip scan to jump the unconstrained column.
2. For ordering, the index is sorted by (status, updated_at) within a tenant and not by created_at, so the LIMIT cannot stop early: every matching row is read and then sorted. That is the 'Sort Method: top-N heapsort' line, and it is why the plan reads 2M rows to answer with 20.
3. Derive the replacement from the access path — equality, equality, then the ordering column: CREATE INDEX CONCURRENTLY ON resource (tenant_id, owner_user_id, created_at DESC). The scan seeks to the (tenant, owner) range and walks 20 entries in order, so the Sort node disappears along with the row-read.
4. Treat INCLUDE (title, status) as conditional, not free. An index-only scan still visits the heap for any row whose page is not marked all-visible, so on a table taking 1.2k writes/second the win depends on autovacuum keeping the visibility map current, and the wider index costs more on every insert.

Follow-up

• 90% of rows are status='active'. Would a partial index WHERE status = 'active' change your answer, and for which of the three queries?
• A dashboard runs this for 40 owners in one page load. What changes about the design?

Denormalise tenant onto revisions and backfill it live

Hard · denormalisation · online migration · backfill

resource_revision (revision_id, resource_id, version, actor_user_id, change_kind, patch, request_id, created_at) has 400M rows and no tenant column; tenant_id lives only on resource. Two reads need it: a tenant-scoped audit feed ordered by created_at DESC, and an offboarding purge. Both join back to resource today. Justify adding tenant_id to resource_revision against those two reads, name the anomaly the copy introduces and the constraint that prevents it, then give the ordered migration for a live table taking 1.2k writes/second — the lock each step takes, how the backfill is batched, and where each step stops being reversible. PostgreSQL 16.

Approach

1. Justify from the access path rather than from taste. Without the column, the audit feed either scans resource_revision by created_at and discards other tenants' rows, or resolves the tenant's resource_ids first and probes with them — both proportional to the tenant's whole history rather than to one page. With (tenant_id, created_at DESC, revision_id DESC) it is a seek that stops at 50 rows, and the purge becomes a ranged delete instead of a join.
2. Name the cost exactly: a second copy of a fact can disagree with the first. Make the disagreement unwritable rather than documented — add UNIQUE (resource_id, tenant_id) on resource so it can serve as a foreign-key target, then FOREIGN KEY (resource_id, tenant_id) REFERENCES resource (resource_id, tenant_id) on the revision table. A revision can then only ever carry its parent's tenant.
3. Step one, expand: ALTER TABLE resource_revision ADD COLUMN tenant_id BIGINT NULL, with no default, so it is a catalogue change and no rewrite. It still needs ACCESS EXCLUSIVE for an instant, and that instant queues behind the longest open transaction on the table while every later query queues behind it — set lock_timeout to 2s and retry rather than wait.
4. Step two, dual-write: deploy the writer that populates tenant_id on every new revision while reads still use the join. Reversible by redeploying the previous build, because nothing reads the column yet.

Follow-up

• The backfill is half finished and a rollback is required. What state is the table in, and what does the previous build do with a half-populated column?
• How do you verify the backfill actually finished, given rows are still being inserted while it runs?

Explain the difference between in-order and out-of-order (OoO) CPU execution pipelines.

Medium · Computer Architecture & Low-Level System

Approach

1. Fix the scope first: who calls this, how often, and what they do when it fails.
2. Name the read and write paths separately; they rarely have the same bottleneck.
3. Choose a partition key and say what query it makes expensive.
4. State the consistency you need, and where you are willing to be stale.

Follow-up

• What breaks first when traffic grows ten times?
• How does this behave when that dependency is down for an hour?

What is register renaming, and what hardware problem does it solve in modern processors?

Medium · Computer Architecture & Low-Level System

Approach

1. Fix the scope first: who calls this, how often, and what they do when it fails.
2. Name the read and write paths separately; they rarely have the same bottleneck.
3. Choose a partition key and say what query it makes expensive.
4. State the consistency you need, and where you are willing to be stale.

Follow-up

• What breaks first when traffic grows ten times?
• How does this behave when that dependency is down for an hour?

How do branch predictors work, and how can software design minimize branch misprediction penalties?

Medium · Computer Architecture & Low-Level System

Approach

1. Fix the scope first: who calls this, how often, and what they do when it fails.
2. Name the read and write paths separately; they rarely have the same bottleneck.
3. Choose a partition key and say what query it makes expensive.
4. State the consistency you need, and where you are willing to be stale.

Follow-up

• What breaks first when traffic grows ten times?
• How does this behave when that dependency is down for an hour?

How would you design an API to send encoded data efficiently between multiple RISC-V processors?

Medium · Computer Architecture & Low-Level System

Approach

1. Fix the scope first: who calls this, how often, and what they do when it fails.
2. Name the read and write paths separately; they rarely have the same bottleneck.
3. Choose a partition key and say what query it makes expensive.
4. State the consistency you need, and where you are willing to be stale.

Follow-up

• What breaks first when traffic grows ten times?
• How does this behave when that dependency is down for an hour?

Walk me through a situation where you had to quickly learn a completely new hardware architecture or software

Medium · Behavioral & Project Management

Walk me through a situation where you had to quickly learn a completely new hardware architecture or software framework to deliver a project.

Approach

1. Fix the scope first: who calls this, how often, and what they do when it fails.
2. Name the read and write paths separately; they rarely have the same bottleneck.
3. Choose a partition key and say what query it makes expensive.
4. State the consistency you need, and where you are willing to be stale.

Follow-up

• What breaks first when traffic grows ten times?
• How does this behave when that dependency is down for an hour?

Every query on one table stalls for forty seconds mid-deploy

Medium · lock contention · ddl · migrations

During a release on PostgreSQL, every query touching resource times out for about 40 seconds and then recovers with no intervention. The release ran one migration, ALTER TABLE resource ADD COLUMN archived_reason TEXT, and the migration log shows it completing in 6 ms. Unrelated tables showed no change in error rate. Explain how a 6 ms statement caused a 40-second stall, give the ordered checks you would run on a live system to confirm it, and give the migration procedure that prevents a repeat.

Approach

1. Separate the statement's duration from the lock's duration. ADD COLUMN with no default is a catalogue-only change and genuinely runs in milliseconds, but it requires ACCESS EXCLUSIVE, and it cannot acquire that until every transaction already touching the table has finished.
2. Account for the queueing, which is the part that surprises people. A lock request that is waiting blocks later requests for conflicting modes behind it rather than letting them overtake, so one long-open transaction holds the DDL and the DDL holds all the traffic. The stall length is set by the longest open transaction, not by the size of the change.
3. Confirm on a live system in this order: pg_stat_activity for that table ordered by xact_start, looking for the oldest transaction and specifically for state = idle in transaction; then pg_locks where granted = false to find the waiter; then join them on pid to name blocker and blocked. pg_blocking_pids() does that join for you and is the fastest single call.
4. Prevent rather than merely time it better. Set lock_timeout to a second or two on the migration session so the DDL abandons the queue after a bounded wait and is retried, instead of holding it for as long as the oldest transaction lives. Be exact about what that buys: queries arriving during the wait still queue behind the pending ACCESS EXCLUSIVE request, so each attempt costs them up to one lock_timeout of added latency. The outage goes from 40 seconds to about one second per attempt, not to zero. Also run migrations away from deploy-time peaks, and put a statement timeout and an idle-in-transaction timeout on the analytics role that opens the long transactions.

Follow-up

• The same release also wants NOT NULL on that column. What is the sequence that gets there without a long lock?
• Your lock_timeout retry fails ten times in a row because the analytics transaction is always open. What do you change?

Suggested practice checklist

Built from the rounds and topics Tenstorrent candidates report.

Day 1 — Map the Tenstorrent loop

• Write out the reported sequence: Recruiter Conversation, Technical Rounds, Live Coding Challenges, System Design Discussions, Behavioral Evaluations, Final Decision.
• For each round, write one sentence on what it is judging, from the description above, and mark the one you are least ready for.

Deliverable: A one-page map of the 6 reported rounds, with the weakest marked.

Day 2 — Work Compiler Design

• Spend the session on Compiler Design, which Tenstorrent candidates report being tested on.
• Write one worked example in Compiler Design and time yourself on it.

Deliverable: One timed worked example in Compiler Design.

Day 3 — Work Code Generation

• Spend the session on Code Generation, which Tenstorrent candidates report being tested on.
• Write one worked example in Code Generation and time yourself on it.

Deliverable: One timed worked example in Code Generation.

Day 4 — Work RISC-V Architecture

• Spend the session on RISC-V Architecture, which Tenstorrent candidates report being tested on.
• Write one worked example in RISC-V Architecture and time yourself on it.

Deliverable: One timed worked example in RISC-V Architecture.

Day 5 — Answer out loud: Computer Architecture & Low-Level Systems

• Answer aloud, timed: Explain the difference between in-order and out-of-order (OoO) CPU execution pipelines.
• Answer aloud, timed: What is register renaming, and what hardware problem does it solve in modern processors?

Deliverable: Spoken answers to 2 reported Computer Architecture & Low-Level Systems question(s), under time.

Day 6 — Answer out loud: Data Structures & Algorithms

• Answer aloud, timed: Implement a basic hash map or associative array from scratch in C++ without using standard library containers.
• Answer aloud, timed: Write an algorithm to perform efficient matrix multiplication, and explain how you would optimize it for cache locality.

Deliverable: Spoken answers to 2 reported Data Structures & Algorithms question(s), under time.

Day 7 — Answer out loud: Behavioral & Project Management

• Answer aloud, timed: Describe your proudest engineering project. What was the most difficult technical challenge, and how did you overcome it?
• Answer aloud, timed: Tell me about a time when you had a significant technical disagreement with a teammate or manager. How did you resolve it?

Deliverable: Spoken answers to 2 reported Behavioral & Project Management question(s), under time.

Behavioural preparation

Behavioural rounds judge the decision you made and what it cost.

Describe your proudest engineering project. What was the most difficult technical challenge, and how did you o

Medium · Behavioral & Project Management

Describe your proudest engineering project. What was the most difficult technical challenge, and how did you overcome it?

Approach

1. Pick a story where you made the decision, not one where you watched it.
2. State the situation in two sentences and spend the rest on the reasoning.
3. Give the blast radius: what could have broken, and what you measured.
4. Name the disagreement and how you resolved it with evidence.

Follow-up

• What would you do differently if you ran that again?
• How did you know your change caused the improvement?

Tell me about a time when you had a significant technical disagreement with a teammate or manager. How did you

Medium · Behavioral & Project Management

Tell me about a time when you had a significant technical disagreement with a teammate or manager. How did you resolve it?

Approach

1. Pick a story where you made the decision, not one where you watched it.
2. State the situation in two sentences and spend the rest on the reasoning.
3. Give the blast radius: what could have broken, and what you measured.
4. Name the disagreement and how you resolved it with evidence.

Follow-up

• What would you do differently if you ran that again?
• How did you know your change caused the improvement?

How do you handle shifting priorities or ambiguous requirements when working on a critical product release?

Medium · Behavioral & Project Management

Approach

1. Pick a story where you made the decision, not one where you watched it.
2. State the situation in two sentences and spend the rest on the reasoning.
3. Give the blast radius: what could have broken, and what you measured.
4. Name the disagreement and how you resolved it with evidence.

Follow-up

• What would you do differently if you ran that again?
• How did you know your change caused the improvement?