/
Navigation
C
Chronicles
Browse all articles
C
E
Explore
Semantic exploration
E
R
Research
Entity momentum
R
N
Nexus
Correlations & relationships
N
~
Story Arc
Topic evolution
S
↻
Drift Map
Semantic trajectory animation
D
P
Posts
Analysis & commentary
P
Browse
@
Entities
Companies, people, products, technologies
◇
Domains
Browse by publication source
☉
Handles
Browse by social media handle
Detection
?
Concept Search
Semantic similarity search
!
High Impact Stories
Top coverage by position
+
Sentiment Analysis
Positive/negative coverage
*
Anomaly Detection
Unusual coverage patterns
Analysis
vs
Rivalry Report
Compare two entities head-to-head
/\
Semantic Pivots
Narrative discontinuities
!!
Crisis Response
Event recovery patterns
Connected
Nav: C E R N
Search: /
Command: ⌘K
Embeddings: large
VOICE ARCHIVE

Ronald Mannak

@ronaldmannak
8 posts
2026-03-10
MLX benchmarks are in and I did not expect these results. the M5 Max blows the M3 Ultra out of the water, despite having more GPU cores and higher memory bandwidth. Compute-bound prefill is much faster (up to 2x) thanks to the new M5 Neural Accelerators, but also memory-bound [image]
2026-03-10 View on X
PCMag

MacBook Pro 16" (M5 Max) review: the “Super core” architecture and 40-core GPU mean beastly performance, but it has a five-year-old design and is $100+ pricier

EDITORS' CHOICE  —  4.5  —  Outstanding  —  THE BOTTOM LINE  —  The 2026 MacBook Pro 16-inch with M5 Max …

View original
Tl;DR: Time to first token, where GPU speed matters most, is 2.5× faster on the M5 Max than on the M4 Max! It's also 1.5× faster than the M3 Ultra in llama.cpp testing, though not in MLX. I assume llama.cpp is tapping into the new Neural Accelerators. Now I really want to see
2026-03-10 View on X
PCMag

MacBook Pro 16" (M5 Max) review: the “Super core” architecture and 40-core GPU mean beastly performance, but it has a five-year-old design and is $100+ pricier

EDITORS' CHOICE  —  4.5  —  Outstanding  —  THE BOTTOM LINE  —  The 2026 MacBook Pro 16-inch with M5 Max …

View original
MLX benchmarks are in and I did not expect these results. the M5 Max blows the M3 Ultra out of the water, despite having more GPU cores and higher memory bandwidth. Compute-bound prefill is much faster (up to 2x) thanks to the new M5 Neural Accelerators, but also memory-bound [image]
2026-03-10 View on X
Tom's Hardware

MacBook Neo review: premium chassis, vivid display, mechanical touchpad is clickable anywhere, and affordable, but no backlit keyboard and no USB-C port labels

The MacBook Neo looks and feels like it's expensive.  That's the magic. … Why you can trust Tom's Hardware …

View original
Tl;DR: Time to first token, where GPU speed matters most, is 2.5× faster on the M5 Max than on the M4 Max! It's also 1.5× faster than the M3 Ultra in llama.cpp testing, though not in MLX. I assume llama.cpp is tapping into the new Neural Accelerators. Now I really want to see
2026-03-10 View on X
Tom's Hardware

MacBook Neo review: premium chassis, vivid display, mechanical touchpad is clickable anywhere, and affordable, but no backlit keyboard and no USB-C port labels

The MacBook Neo looks and feels like it's expensive.  That's the magic. … Why you can trust Tom's Hardware …

View original
2026-03-09
MLX benchmarks are in and I did not expect these results. the M5 Max blows the M3 Ultra out of the water, despite having more GPU cores and higher memory bandwidth. Compute-bound prefill is much faster (up to 2x) thanks to the new M5 Neural Accelerators, but also memory-bound [image]
2026-03-09 View on X
PCMag

MacBook Pro 16" (M5 Max) review: “Super core” architecture and 40-core GPU deliver beastly performance, but still retains a five-year-old design

Outstanding  —  THE BOTTOM LINE  —  Pros & Cons  — “Super core” architecture delivers incredible performance

View original
Tl;DR: Time to first token, where GPU speed matters most, is 2.5× faster on the M5 Max than on the M4 Max! It's also 1.5× faster than the M3 Ultra in llama.cpp testing, though not in MLX. I assume llama.cpp is tapping into the new Neural Accelerators. Now I really want to see
2026-03-09 View on X
PCMag

MacBook Pro 16" (M5 Max) review: “Super core” architecture and 40-core GPU deliver beastly performance, but still retains a five-year-old design

Outstanding  —  THE BOTTOM LINE  —  Pros & Cons  — “Super core” architecture delivers incredible performance

View original
2024-10-29
Interesting choice. The power button is on the bottom of the new Mac mini. And if you place it on the right side of your monitor, the stand will block access. Granted, you'll likely only press that button 5 times or less over its lifetime, but still [image]
2024-10-29 View on X
The Verge

Apple updates the Mac mini with a much smaller design, M4 or M4 Pro, two front-facing USB-C ports, up to 64GB of RAM, and more, shipping November 8 for $599+

As part of the company's week of Mac-focused announcements, Apple has just introduced a smaller, yet even more powerful Mac Mini.

View original
2024-02-04
This is Apple's second tool (after MLX) that doesn't have Swift support at launch. What's going on?
2024-02-04 View on X
Pkl Blog

Apple open sources Pkl, a configuration-as-code language with rich validation and tooling, with Swift, Go, Java, and Kotlin integration

We are delighted to announce the open source first release of Pkl (pronounced Pickle), a programming language for producing configuration. Threads: @danielpunkass . Mastodon: @lkan...

View original