Montreal, August 11, 2026 — This week the technology world delivered three stories that are far more interesting under the hood than their headlines suggest. Samsung told us exactly how it engineered most of the crease out of a folding OLED screen — with titanium, of all things. At the Future of Memory and Storage summit, the same company sketched a radical answer to the single biggest bottleneck in AI computing: stacking memory directly on top of the processor. And in the security world, CISA quietly added a batch of actively exploited software flaws to its must-patch list, a reminder that the boring discipline of updating your devices is still the cheapest insurance in tech. In today's edition we explain the science behind all three — metallurgy and neutral planes, through-silicon vias and wafer bonding, remote code execution and hardware-backed authentication — and translate each into practical buying advice you can act on right now.

Today's Tech Radar

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

Deep Dive #1 — Flex Titanium: The Metallurgy That Flattens the Fold

A Samsung foldable smartphone. The folding display is a laminated stack of glass, polymer, and now titanium layers. Photo: 2H Media / Unsplash.

When Samsung Display revealed the details of the Galaxy Z Fold 8's screen this month, the headline was a marketing name — "Flex Titanium" — but the substance is a real change in materials engineering. According to reporting from 9to5Google, Android Authority, and Tom's Guide, Samsung has for the first time overhauled the layered structure of its foldable display, swapping a conventional plastic (polymer) support film for a titanium-alloy film beneath the OLED panel, paired with a titanium support plate. Samsung says the alloy film is roughly twenty times stiffer than the plastic film it replaces, at about one third the thickness of a human hair, and that the new stack cuts overall display thickness by around 12%.

Why folding a screen is hard: the neutral plane problem

To appreciate why titanium helps, it helps to understand what bending does to a laminated panel. A foldable display is a sandwich: an ultra-thin glass (UTG) cover, adhesive layers, the OLED emitting layer itself, and supporting films underneath. When you bend any sandwich of materials, the outer surface of the curve is stretched (tension) while the inner surface is squeezed (compression). Somewhere in the middle lies the neutral plane — a layer that is neither stretched nor compressed. Display engineers fight to position the fragile OLED layer as close to the neutral plane as possible, because organic light-emitting materials and their microscopic wiring tolerate very little strain before they crack or delaminate.

The crease you see on most foldables is plastic deformation: after thousands of folds, the soft polymer layers under the screen take a permanent "set" along the hinge line, like a paperback that has been read too many times. The screen above follows that deformed foundation, and light reflecting off the resulting valley makes the crease visible.

What titanium changes

Titanium alloys have an unusually attractive combination of properties for this job: high stiffness-to-weight ratio, excellent elastic recovery (they spring back rather than taking a set), and good fatigue resistance — the ability to endure hundreds of thousands of stress cycles without cracking. By replacing the compliant polymer film with a far stiffer titanium-alloy film, Samsung spreads the bending stress of the folding zone across a wider area instead of letting it concentrate along a single line. Android Authority's report adds that the new structure has fewer internal air gaps, which means the layers support each other more uniformly during every fold and unfold cycle. The result, per hands-on accounts from Tom's Guide, is a crease that is dramatically less visible to the eye — though reviewers note you can still feel a slight ripple under your finger. Physics hasn't been repealed; it has been managed.

There's a second-order benefit, too. A stiffer, thinner backing plate lets the whole phone get thinner without making the screen feel spongy, and it disperses impact energy from drops and presses across the panel. That's the same principle behind why a steel ruler resists a poke better than a strip of cardboard of equal thickness.

The market is responding

This engineering matters commercially. Multiple outlets, including GSMArena and SamMobile, report that Samsung is raising its Galaxy Z Fold 8 production target by roughly one million units — from about 2.8 million toward 4 million — after record-breaking pre-orders, with the Fold accounting for an unusually high share of demand in several markets. When a manufacturer makes a mid-cycle production increase of that size, it signals that a long-promised technology transition (foldables as everyday flagships) is finally happening.

What this means for your next screen purchase

You don't need to buy a foldable to benefit from this science. The same principles — stiffer substrates, better lamination, thinner optical stacks — flow into conventional displays every year. If you are due for a screen upgrade today, our advice is to prioritize panel quality and resolution for your actual use case rather than chasing the newest form factor. For a home office or creative work, the Samsung ViewFinity S8 32" 4K UHD monitor (in stock) gives you the pixel density where you actually spend your day, while the budget-friendly Samsung Essential 32" Full HD monitor (in stock) is a solid choice for everyday productivity. For meeting rooms, storefronts, or lobbies, large-format professional panels such as the Samsung QBC Series 50" UHD display (in stock) and the Samsung QM85C 85" UHD display (in stock) bring the same Samsung panel engineering to commercial scale, rated for long daily duty cycles. On the smartphone side, our current foldable and flagship stock rotates quickly — models like the Samsung Galaxy S23 FE 5G are listed as available to order — so if you want a specific phone, request a free quote from our team and we'll confirm availability and lead times for you.

Deep Dive #2 — zHBM and the 400-Layer Era: Building Memory Upward

Modern memory chips are already three-dimensional; zHBM proposes stacking them directly on top of the processor. Photo: Vishnu Mohanan / Unsplash.

At the Future of Memory and Storage (FMS) 2026 summit, Samsung presented a trio of memory technologies that, taken together, describe where all of computing is headed: zHBM, a concept that stacks high-bandwidth memory directly on top of an AI processor; zNAND-O, Samsung's take on the emerging "high bandwidth flash" (HBF) idea; and BV-NAND, flash memory that breaks the 400-layer barrier. TrendForce, Tom's Hardware, HotHardware, and the Korea JoongAng Daily all covered the announcements. It's important to say up front: zHBM is a concept architecture with a horizon around 2029, not a product you can buy — Samsung cited no confirmed customers or production schedule. But the physics it addresses affects every computer you own today.

The memory wall, explained

Processors have gotten faster much more quickly than memory has gotten closer. A modern AI accelerator can perform trillions of operations per second, but every operation needs data, and that data lives in memory chips physically separated from the processor. Moving a bit of data across a circuit board costs far more energy — and time — than performing an arithmetic operation on it. Engineers call this the memory wall, and in AI workloads, where models are measured in tens or hundreds of gigabytes, it is usually the binding constraint. This is why "high-bandwidth memory" (HBM) exists: instead of placing memory chips beside the processor on a motherboard, HBM stacks DRAM dies vertically and connects them with thousands of microscopic vertical wires called through-silicon vias (TSVs) — think of an apartment tower with thousands of elevators, instead of suburban houses connected by a single road. Today's HBM sits next to the GPU on a shared silicon base called an interposer.

zHBM: eliminating the last few millimetres

Samsung's zHBM concept takes the logical next step: put the memory tower directly on top of the processor, using next-generation wafer bonding — fusing entire silicon wafers together with copper-to-copper connections at microscopic pitch, rather than soldering individual bumps. Because signals now travel micrometres instead of millimetres, Samsung projects roughly eight times the performance of HBM5 (itself a future standard), more than ten times the memory density, about three times better energy efficiency, and less than half the thermal resistance, according to the TrendForce and HotHardware reports. The thermal point is counterintuitive and crucial: stacking a heater (memory) on top of a furnace (GPU) sounds like a cooling nightmare, which is precisely why the wafer-bonded interface — which conducts heat far better than today's bumped connections — is the enabling trick. If it works at scale, it would mean AI systems that hold far larger models per chip and burn far less power per calculation.

400+ layers of flash, and "high bandwidth flash"

On the storage side, Samsung's BV-NAND pushes past 400 stacked layers of flash memory cells. To picture what that means: each layer is a full grid of storage cells, and manufacturers must etch perfectly vertical holes through all 400+ layers — an aspect ratio comparable to drilling a smooth, straight elevator shaft through a 400-storey building. More layers mean more capacity per chip at lower cost per gigabyte, which is why SSD prices per terabyte keep drifting downward over time. Meanwhile, zNAND-O — Samsung's version of high bandwidth flash — borrows HBM's stacking-and-wide-bus philosophy and applies it to NAND flash, aiming to give AI systems a huge, cheaper tier of memory between DRAM and conventional SSDs, with a 2029-era target according to industry reporting.

Practical takeaways for buyers today

You can't buy zHBM, but the trickle-down of these same ideas is already on our shelves. Today's NVMe SSDs are the direct beneficiaries of 3D NAND stacking: the Samsung 990 PRO 1TB PCIe Gen4 NVMe SSD (in stock) delivers sequential reads around 7 GB/s — bandwidth that would have required exotic server hardware a decade ago — and is the single most cost-effective upgrade for a sluggish PC. If your work moves with you, the Samsung T7 Shield 2TB portable SSD (in stock) puts rugged, 3D-NAND-based storage in your pocket, and the Samsung PRO Ultimate 128GB microSD (in stock) applies the same cell technology to cameras, drones, and tablets. When you're comparing SSDs, the science gives you a simple checklist: prefer a drive with a DRAM cache (it keeps the map of your data in fast memory), match the interface to your machine (PCIe Gen4 x4 for modern desktops and laptops), and check endurance ratings (TBW — terabytes written) if you shuttle large video or dataset files daily.

Deep Dive #3 — Actively Exploited: What This Week's CISA Alerts Teach Us About Digital Hygiene

Patching known flaws quickly is the cheapest, most effective security measure available. Photo: FlyD / Unsplash.

On August 5, the U.S. Cybersecurity and Infrastructure Security Agency (CISA) added three vulnerabilities to its Known Exploited Vulnerabilities (KEV) catalog — flaws in Langflow (a popular tool for building AI applications), Apache Tomcat (one of the internet's most widely deployed web servers), and N-able N-central (a platform IT providers use to manage client computers), as reported by The Hacker News. The N-central case is particularly instructive: N-able's own security team disclosed that a threat actor was caught exploiting a zero-day — a flaw unknown to the vendor at the time — in a customer environment in late July, before an emergency fix shipped on August 6. CISA's weekly vulnerability summary for the same period runs to hundreds of entries.

The science of an exploit

Terms like "remote code execution" get thrown around loosely, so let's define the mechanics. Software is a set of instructions operating on data. A vulnerability is a place where the boundary between the two breaks down — where carefully crafted data (a web request, a file, a message) can trick a program into treating attacker input as instructions. In a command injection flaw, for instance, text an attacker submits into a form field ends up concatenated into a command the server executes, so typing the right hostile string gives the attacker a shell on the machine. Remote code execution (RCE) is the general term for any flaw that lets an attacker run their own code from afar; it is the most severe class because it converts a stranger on the internet into a user on your machine. A zero-day is a flaw being exploited before the vendor knows about it — meaning no patch exists and only layered defenses can help. Once a vendor ships a fix, the race inverts: attackers reverse-engineer the patch to find the flaw, then hunt for machines that haven't updated yet. That window — between patch release and patch installation — is where the overwhelming majority of real-world compromises happen. The KEV catalog exists precisely to shorten it: it is a public, curated list that says "criminals are using this one right now; fix it first."

Why this matters even if you don't run a server

The N-central incident illustrates a modern reality: attackers increasingly target the tools that manage many computers at once, because one breach cascades into hundreds. The same logic applies at household and small-business scale. Your laptop, phone, tablet, router, and even your smart TV all run update-able software with occasional critical flaws. The single highest-value habit is enabling automatic updates on every device you own and rebooting when prompted. The second is reducing what attackers can do if a password leaks — which is where hardware comes in.

Hardware-backed security: the physics of a phishing-proof login

Passwords fail because they are secrets that can be copied. FIDO2 hardware keys fix this with public-key cryptography: the private key is generated inside a tamper-resistant chip and never leaves it. When you log in, the site sends a mathematical challenge; the key signs it — including the website's real domain name in the signature — and a phishing site with a look-alike address simply produces an invalid signature. There is no shared secret to steal, and biometric variants add a fingerprint check so a stolen key alone is useless. We stock the Kensington VeriMark Guard USB-C fingerprint key (in stock), which supports FIDO2, WebAuthn, and U2F — it turns your strongest login from "something you know" into "something you have plus something you are." Physical security layers matter too: a Kensington MicroSaver 2.0 keyed laptop lock (in stock) protects hardware in shared spaces, and a MagPro 27" magnetic privacy filter (in stock) narrows your monitor's viewing angle so shoulder-surfers see a black screen. If you run a small business and aren't sure where your soft spots are — unpatched machines, shared passwords, no backup strategy — request a free quote from our team and we'll help you assess your setup and pick the right protective hardware.

Glossary of the Week

Setup at a Glance

That's today's tour through the science behind the headlines — from titanium films a third the width of a hair to memory towers bonded onto processors, to the quiet arms race between patchers and attackers. If anything here sparked a question about your own setup — whether that's choosing a display, speeding up an aging PC, or hardening your logins — request a free quote from our team and we'll point you to the right gear for your budget. See you tomorrow.

Sources & Further Reading

Foldable display: 9to5Google — Galaxy Z Fold 8 'Flex Titanium' display; Android Authority — How the Z Fold 8 display achieved a barely visible crease; Tom's Guide — Flex Titanium hands-on; GSMArena — Z Fold 8 production increase; SamMobile — 1 million more units on high demand. Memory: TrendForce — Samsung showcases zHBM at FMS 2026; HotHardware — zHBM 3D memory concept; Korea JoongAng Daily — Samsung unveils zHBM, zNAND-O and 400-layer V-NAND; Tom's Hardware — Samsung debuts next-gen memory tech. Security: The Hacker News — CISA flags Langflow, Tomcat and N-central flaws as actively exploited; N-able — N-central security update, August 6, 2026; CISA — Vulnerability summary for the week of August 3, 2026. Radar items: Tech Startups — Top tech news, August 10, 2026; Tech Startups — Top tech news, August 3, 2026. Photos: Unsplash (free commercial license).