Learners engaging with GPU architecture material

LEARNER EXPERIENCES

What Learners Say About Veyl

Feedback from individuals and teams who have worked through our GPU architecture programmes.

← Back to Home

Programme Reviews

FZ

Farhan Zulkifli

ML Engineer · Kuala Lumpur

"I had been working with PyTorch for about two years before this but kept hitting questions I couldn't answer about why certain batch sizes behaved the way they did. The Reading Track laid out the memory hierarchy in a way that actually connected to what I was observing. Module 5 on shared memory was where things clicked for me."

GPU Fundamentals Reading Track · May 2025

NK

Nurul Khairina

Data Scientist · Petaling Jaya

"The Lab Companion is genuinely practical in a way that most 'hands-on' courses aren't. Working through actual spec sheets rather than toy examples made the exercises feel relevant. I would say the workload sizing sessions (8 and 9) are worth the price of the whole companion on their own."

Hands-On Lab Companion · April 2025

CW

Chen Wei Liang

Software Architect · Cyberjaya

"We put six people from our infrastructure team through the Team Knowledge Programme in March. The pre-session context call was useful — the facilitator clearly read what we sent and adjusted the workshop examples to match our deployment environment. The reference handbook is still sitting on the team's shared drive and gets opened regularly."

Team Knowledge Programme · March 2025

SA

Siti Amirah

Final-year CS student · Shah Alam

"As someone finishing my degree, I was looking for something that would help me talk credibly about hardware in interviews for ML roles. The Reading Track covered exactly the kind of foundational questions that come up. The glossary at the end of each module is genuinely useful — not just a list of definitions but small explanations of why each term matters."

GPU Fundamentals Reading Track · May 2025

RN

Rajesh Nair

MLOps Engineer · Kuala Lumpur

"I work primarily on the operations side of ML deployments and found the Lab Companion covered the dashboard interpretation exercises extremely well. The section on SM occupancy and what it means for throughput was more actionable than anything I found in the official documentation. One suggestion: more exercises around multi-GPU setups would be a welcome addition."

Hands-On Lab Companion · April 2025

LH

Lim Hui Shan

Engineering Manager · Selangor

"We are a team of twelve and the hardware literacy gap was creating friction in planning conversations — some people had strong intuitions about GPU behaviour while others were working from very general assumptions. The Team Knowledge Programme brought everyone to a shared vocabulary over four weeks. The fact that it was HRDC-claimable made the budget conversation simple."

Team Knowledge Programme · February 2025


Learning Journeys

How three learners moved from initial gaps to applied hardware understanding.

FZ

ML Engineer — From unexplained slowdowns to confident workload reasoning

STARTING POINT

Two years of ML experience with PyTorch. Understood model architecture well but kept encountering GPU-related slowdowns without knowing where to look for the cause. Had read documentation but found it hard to connect spec numbers to observed behaviour.

PROGRAMME PATH

Completed the GPU Fundamentals Reading Track over five weeks, focusing particularly on memory hierarchy and the throughput modules. Followed up with two Lab Companion exercises on his own after the track to check his understanding against practical scenarios.

OUTCOME

Reported being able to diagnose a memory bandwidth bottleneck in a production training job within two weeks of finishing the track — an issue that had been attributed to a dataset problem for several months. Now reads GPU profiler outputs with confidence.

LH

Engineering Team — Building shared vocabulary across twelve engineers

STARTING POINT

A twelve-person engineering team split between those with hardware intuition and those working from software abstractions only. Planning conversations around GPU provisioning often stalled or produced decisions that senior engineers later revised without explanation.

PROGRAMME PATH

Enrolled in the Team Knowledge Programme. Facilitator conducted a pre-programme context session with the engineering manager and two senior engineers. Workshop scenarios were drawn from the team's actual provisioning decisions and common friction points.

OUTCOME

GPU provisioning discussions in sprint planning now conclude in under thirty minutes where they previously ran long without resolution. The reference handbook is actively used. Three team members requested individual Lab Companions for deeper study after the programme ended.

SA

Computer Science Student — Preparing for ML hardware questions in technical interviews

STARTING POINT

Final-year CS student with project experience in ML but no formal hardware study. Aware that ML infrastructure roles often ask hardware-related questions and wanted to build defensible knowledge before internship and graduate applications.

PROGRAMME PATH

Worked through the GPU Fundamentals Reading Track at two modules per week alongside university coursework, taking advantage of the self-paced format. Used the knowledge checks to identify where to reread before progressing.

OUTCOME

Received and accepted an ML infrastructure internship offer the month after completing the track. Credited the programme with giving her concrete, structured answers to hardware questions that had previously made her uncertain in technical conversations.


Get in Touch

Address

Jalan Ampang 128
50450 KL, MY

Hours (MYT)

Mon–Fri 9am–6pm
Sat 10am–2pm


In Numbers

120+

Learners enrolled

14

Team programmes delivered

4.7

Average satisfaction (out of 5)

MY

Based in Malaysia — MYT hours

MDec Digital Education Listing

Recognised provider under the Malaysia Digital Economy Corporation continuing education directory.

HRDC Claimable Programmes

Team Knowledge Programme eligible for Human Resources Development Corporation training claims for Malaysian employers.

Defined Review Schedule

All content reviewed on a documented cycle to stay current with GPU architecture changes.

Add Your Experience to This List

Find out which programme fits your background and enrol to start building hardware literacy that lasts.