Montreal, August 6, 2026. This week the technology world offered a rare alignment: a genuinely new display technology went on sale, the first laptops built on a next-generation chip process reached store shelves in volume, and the economics of computer memory — the invisible ingredient in every device we sell — continued their most dramatic upheaval in a decade. Each of these stories rewards a closer look, because each one is rooted in real physics and real engineering rather than marketing language. In today's edition we explain how a television can paint colour with light sources smaller than a human hair, why a transistor that stands upright changes what your next laptop can do, and why the same AI boom that powers chatbots is quietly raising the price of the RAM in your desktop. As always, we close with practical, stock-checked buying advice and a glossary of every technical term we use along the way.

Today's Tech Radar

Our shortlist of the ten most significant technology stories of the past week, before we deep-dive into three of them:

Deep Dive 1 — Micro RGB: How a Television Learns to Paint With Three Colours of Light

Large-format televisions are entering a new era of backlight engineering. Photo: Prydumano Design / Unsplash.

The science: why the backlight is the whole story

Every LCD television you have ever owned is, at heart, a light valve. The liquid-crystal layer does not emit light; it merely twists to let more or less light through, pixel by pixel, while colour filters tint that light red, green or blue. The quality of the picture therefore depends enormously on the light source behind the panel. For two decades that source has evolved: from fluorescent tubes, to white LEDs along the edges, to full arrays of blue-white mini-LEDs grouped into dimming zones that can darken parts of the screen independently. Each step improved contrast, but one compromise remained constant — the backlight was always white-ish, and the colour filters had to carve red, green and blue out of that broadband white light, wasting energy and limiting colour purity in the process.

Samsung's Micro RGB architecture, launched on a 115-inch flagship first in South Korea and now reaching international markets at $29,999 USD, removes that compromise. Instead of white or blue LEDs, the backlight is an ultra-fine grid of discrete red, green and blue LEDs, each less than 100 micrometres across — thinner than a human hair. Because the backlight itself can produce coloured light at any point on the screen, the colour filters no longer fight their light source. The result, according to Samsung and to the first independent hands-on reports from reviewers at Forbes, T3 and HDTVTest, is a much wider colour gamut, dramatically finer local dimming — more than four times the dimming zones of top mini-LED sets — and roughly 20% better energy efficiency, because far less light is thrown away in the filters.

Why this is not quite Micro-LED — and why that matters for pricing

It is important to distinguish Micro RGB from true Micro-LED. In a Micro-LED display, each subpixel is its own LED emitting light directly, with no LCD layer at all — a technology that remains spectacularly expensive to manufacture at living-room sizes. Micro RGB keeps the LCD light-valve layer but upgrades the illumination behind it to per-region coloured light. Think of it as the difference between repainting a stained-glass window (Micro-LED) and replacing the sun behind it with thousands of tiny coloured spotlights (Micro RGB). The clever part is economic as much as optical: Micro RGB reuses mature LCD manufacturing, which is why analysts expect the technology to migrate down to 98-inch and 75-inch classes far faster than Micro-LED ever could.

The AI in the picture

Driving tens of thousands of tri-colour zones in real time is a computational problem, not just an optical one. Samsung's Micro RGB AI Engine analyses every frame and adjusts the colour output of the backlight grid to match the source material, while a 'Color Booster' pass identifies scenes with dull tones and selectively enhances saturation. Whether one wants a television making aesthetic decisions is a matter of taste — reviewers note the processing can be toned down in settings — but the principle is sound: the more independently controllable light sources a display has, the more useful per-frame analysis becomes.

What it means for buyers today

A $29,999, 115-inch television is a technology statement, not a shopping suggestion. The practical lesson for buyers in 2026 is that large-format LCD with advanced backlighting is now the sweet spot of price versus performance, and it is falling in price precisely because these flagship technologies push the ceiling upward. If you are equipping a boardroom, a classroom, a restaurant or a storefront, commercial-grade 4K panels deliver the reliability numbers that matter more than exotic backlights: rated brightness, glare treatment and continuous-operation certification. In our current inventory, the Samsung 55-inch Crystal UHD Signage QBC (in stock) is a slim 4K display rated for 16/7 operation, while the Samsung QM85C 85-inch UHD display (in stock) offers 500-nit non-glare output rated for 24/7 duty — the same engineering lineage, one rung below the bleeding edge. For a desk rather than a wall, the Samsung ViewFinity S8 27-inch 4K monitor (in stock) brings factory-calibrated colour to creators, and the Samsung Odyssey G5 34-inch 165 Hz curved monitor (in stock) covers immersive gaming. If you are unsure which panel class fits your space and budget, request a free quote from our team and we will size it for you.

Deep Dive 2 — Panther Lake and the 18A Process: The Transistor Stands Up

Modern processors pack tens of billions of transistors; the way each one is built determines a laptop's speed and battery life. Photo: BoliviaInteligente / Unsplash.

The science: RibbonFET, or why geometry beats brute force

Intel's Core Ultra Series 3 processors — codenamed Panther Lake, unveiled at CES in January and now shipping in volume in laptops from every major manufacturer — are the first consumer chips built on Intel's 18A process. Two structural innovations define 18A, and both are easier to understand than their names suggest.

The first is RibbonFET, Intel's implementation of the gate-all-around (GAA) transistor. A transistor is a switch: a gate electrode controls whether current flows through a channel. For decades the gate sat on top of a flat channel, touching one side of it. As transistors shrank, current began leaking when the switch was 'off', because the gate could not fully control a channel it only touched on one face. The FinFET era (roughly 2011–2024) raised the channel into a vertical fin so the gate could grip three sides. Gate-all-around completes the logic: the channel becomes a stack of horizontal ribbons, and the gate material wraps entirely around each ribbon — four sides out of four. Full enclosure means near-total control of the current, which means transistors can switch faster, leak less, and be packed closer together. Compared with the previous Intel 3 node, Intel cites roughly 15% better power efficiency and 30% higher transistor density from the node alone.

PowerVia: moving the plumbing to the basement

The second innovation is PowerVia, or backside power delivery. A modern chip is wired like a city: dozens of metal layers above the transistors carry both signals (data) and power (electricity to run the switches). Routing both through the same layers creates congestion — power lines are thick and bully their way through, distorting signal routing and wasting area. PowerVia relocates the entire power network to the back side of the silicon wafer, feeding transistors from below, while the front side is reserved for signals. The effect is like moving a building's plumbing and electrical conduits out of the hallways and into a dedicated basement: everything above flows more freely. Power reaches transistors with less resistance (less voltage droop under load), and signal wiring gets shorter and cleaner.

What the numbers mean in a laptop

Intel's own figures for Panther Lake — which reviewers will continue to verify through 2026 — claim about 40% lower power at the same performance as the prior generation for single-threaded work, or roughly 10% more speed at equal power, and up to 50% higher multi-threaded performance at the same power. Top configurations pair up to 16 CPU cores with 12 Xe3 graphics cores and a 50-TOPS neural processing unit for on-device AI. The honest translation for a buyer: the practical wins are battery life and sustained performance — a thin laptop that does not throttle after ten minutes of load — rather than headline clock speeds. With over 200 laptop designs in the pipeline, these chips will define the mainstream Windows laptop for the next two years.

Buying advice: today's proven silicon at yesterday's structure

First-generation devices on a new node command premium prices, and current-generation Core Ultra machines remain excellent value — especially as memory prices push new-model costs upward (see our next section). From our in-stock inventory: the HP EliteBook X G1i 14-inch Copilot+ PC with Core Ultra 7 266V (in stock) is a premium business ultrabook with 16 GB of memory; the Dell Pro 14 Plus with Core Ultra 7 265U and 32 GB RAM (in stock) is built for managed business fleets with vPro; and for classrooms and light duty the Lenovo 100e Chromebook Gen 4 (in stock) is a rugged, affordable option. Need help matching processor class to your workload? Request a free quote from our team — we spec machines for businesses every day.

Deep Dive 3 — The Great Memory Crunch: How AI Data Centres Ended Up in Your Shopping Cart

DRAM chips like these have become the most contested commodity in computing. Photo: Liam Briese / Unsplash.

The science: DRAM, HBM, and why they compete for the same factories

Dynamic random-access memory (DRAM) is the working memory of every computer: billions of microscopic capacitors, each holding one bit as an electrical charge that must be refreshed thousands of times per second. The DDR5 modules in a desktop and the LPDDR5X soldered into a phone are variations of the same underlying cells. So is HBM — high-bandwidth memory — the type AI accelerators crave. HBM takes ordinary DRAM dies, stacks them vertically (currently 8 to 16 layers), drills thousands of microscopic vertical connections called through-silicon vias through the stack, and mounts the tower right next to the GPU die. The proximity and the massive parallel connections give HBM several times the bandwidth of conventional DRAM — essential, because an AI processor that cannot be fed data fast enough simply idles.

The catch: HBM is made on the same wafer production lines, by the same three companies — Samsung, SK Hynix and Micron, who together control roughly 95% of global DRAM output — as the memory in your laptop. HBM sells at far higher margins, and each HBM stack consumes disproportionately more wafer area per gigabyte (industry estimates run around three times as much, due to stacking overhead and lower yields). Every wafer redirected to HBM is a wafer of consumer memory that is never produced.

The industry context: a structural shift, not a blip

Through 2025 and into 2026, AI data-centre construction accelerated to the point where, as Tom's Hardware reported this week, single hyperscaler orders can absorb a quarter's worth of HBM output. Reporting across Tom's Hardware, Wccftech and industry analysts describes consumer DDR5 prices rising on the order of 90–110% over recent quarters, SSD prices climbing sharply as NAND capacity is likewise diverted toward data-centre storage, and PC makers stockpiling components. There are early signs of cooling at the retail edge — consumers are hitting affordability limits — but contract prices are still forecast to rise through Q3 2026, and both memory makers and PC vendors have warned the imbalance could persist into 2027. This is not a speculative bubble in memory chips; it is a reallocation of a physically limited resource toward the highest bidder.

What it means for your next purchase

Three practical consequences follow. First, buy the memory you will need up front: many modern thin laptops solder their RAM, upgrades are impossible later, and configured-to-order RAM is now priced at a premium — a 32 GB configuration like the in-stock Dell Pro 14 Plus is insurance against both obsolescence and further price rises. Second, storage is caught in the same current, so if an SSD upgrade is on your list, sooner beats later: the Samsung 990 PRO 2TB NVMe SSD (in stock) and the portable Samsung T7 Shield 1TB rugged SSD (in stock) are both on our shelves today. Third, finished devices bought at fixed prices are a hedge: devices like the Samsung Galaxy Tab A9+ 11-inch tablet (in stock) and the flagship Samsung Galaxy S25 Ultra 256 GB in Titanium Black (in stock, limited quantity) carry 2025-negotiated memory costs that new 2026 models will not.

Security Watch: Three Patches That Should Not Wait

This week's security news is a study in management-plane risk. Cisco shipped emergency hotfixes for CVE-2026-20316, an actively exploited zero-day in Secure Firewall Management Center caused by static hard-coded credentials in its web interface. Broadcom's advisory VMSA-2026-0006 addressed CVE-2026-59309, an authentication bypass in VMware vCenter's Directory Service that lets a network-adjacent attacker take over the virtualization management plane. And N-able disclosed that attackers gained administrative access to N-central remote-management servers after an earlier fix proved incomplete (CVE-2026-18577). The common thread: the tools that manage your infrastructure are themselves the most valuable targets, because one compromise cascades to everything they control. On August 5, CISA added three more actively exploited flaws to its Known Exploited Vulnerabilities catalog. If your business runs any of these platforms — or if you simply do not know what is exposed on your network — request a free quote from our team and we will help you assess your patching and hardware refresh priorities.

Glossary of the Week

Setup at a Glance

That closes today's edition. The thread connecting all three deep dives is worth stating plainly: the frontier of computing — coloured micro-light, wrapped transistors, stacked memory — reaches your desk faster than ever, and understanding the science is the best defence against paying for marketing instead of engineering. If anything in today's article raised questions about your own setup, from a single laptop to a whole office refresh, request a free quote from our team — advice is free, and we enjoy the conversation.

Sources & Further Reading

Micro RGB: Samsung Newsroom, HDTVTest, Forbes, T3, Neowin. Panther Lake / 18A: Neowin, PC Perspective, Wccftech, Guru3D. Memory crunch: Tom's Hardware, Wccftech memory crisis roundup, IPC2U analysis. Security: Cybersecurity News weekly, The Hacker News, Check Point Research. Other radar items: TechStartups, Samsung Newsroom (Galaxy S26). Photos: Unsplash (free commercial license).