Phoenix Global Investments @scuff_peg

China’s CXMT now expects 110–120 billion yuan ($17.6B) in first-half 2026 revenue as AI-driven DRAM demand continues to outpace supply. Even more striking, Q1 revenue surged over 700% YoY to 50.8 billion yuan. AI infrastructure is creating a global memory supercycle where DRAM and HBM are becoming just as critical as GPUs. The interesting part is that this is no longer dominated only by Samsung, SK Hynix and Micron. China is aggressively scaling its own memory ecosystem through players like CXMT and YMTC to secure AI independence. And this is exactly why next-gen memory technologies like ReRAM and in-memory computing are getting attention. …rpriseai.economictimes.indiatimes.com/news/industry/…

Weebit Nano now has a market cap of roughly $1.49B, showing how much attention next-gen memory technologies are starting to attract. The company recently raised A$15 million through a strongly supported share purchase plan, another signal that investors are increasingly betting on alternatives to traditional memory architectures. As AI models scale, the industry is running into a major problem: moving data between processors and memory is becoming too slow, expensive and power hungry. That’s why technologies like ReRAM are getting serious attention. Unlike conventional DRAM or NAND, ReRAM has the potential to support ultra-efficient in-memory computing, where data can be processed closer to where it’s stored. markets.businessinsider.com/news/stocks/we…

New industry forecasts suggest the biggest growth driver is no longer just EVs, it’s battery energy storage systems (BESS), electric trucks, and AI-driven power infrastructure. Some analysts now expect global lithium demand to reach 4.6 million tonnes LCE by 2030, far above consensus estimates. Energy storage alone could consume over 1.1 million tonnes of lithium by the end of the decade, while electric trucks may add another 630,000 tonnes annually because of their massive battery sizes. The lithium narrative is shifting fast: from oversupply fears to structural deficits and long-term infrastructure demand. That’s why lithium stocks are rallying again as investors begin repricing the sector. miningmx.com/trending/65347…

Wood Mackenzie is warning that the lithium market could face supply deficits as early as 2028, and the numbers are massive. Under a net-zero scenario, global lithium demand could surge to 13.2 million tonnes LCE by 2050, while an additional 8.5 million tonnes of new supply would be needed just to meet demand. Analysts estimate the industry may require up to $276 billion in new investment. EVs remain the biggest demand driver, but energy storage systems are now the “sleeper story,” growing 6–7% annually as AI power demand and renewable grids expand. The lithium market is shifting from fears of oversupply to concerns about long-term shortages and strategic resource security. That’s why lithium stocks are back in focus again. woodmac.com/press-releases…

Pioneer Minerals (ASX: $PMM) Secures Drilling Contractor for Maiden Springfield Drill Program in Idaho Key Highlights • Alaska Midnight Sun Drilling appointed for Pioneer’s maiden RC drill campaign at the Springfield Project in Idaho • Maiden drill program targeted for Q3 2026, subject to approvals and completion of ongoing site preparation activities. • Phase 1 drilling to include up to 15 RC holes to 200m depth plus shallow drilling across the historic Springfield Mine footprint. • Drilling designed to test historical tungsten mineralisation, newly identified LiDAR structures, magnetic anomalies and EM targets. • High-grade gallium results include up to 128.7 ppm Ga₂O₃, alongside tungsten assays up to 2.98% WO₃ and gold up to 7.75 g/t Au. • Road upgrades, EM surveys and geological programs progressing ahead of mobilisation. @PioneerMins Link to the Announcement: clients3.weblink.com.au/clients/pionee…

keyboard_arrow_left Previous keyboard_arrow_right Next

dorsaVi (ASX: $DVL) Advances RRAM Commercialisation Pathway for Wearable Robotics & Ultra-Edge AI Key Highlights • Three RRAM material stacks qualified, reducing development risk and increasing flexibility for commercial production. • Technology confirmed compatible with standard CMOS foundry processes, enabling scalable manufacturing on existing commercial lines. • Clear roadmap from 180nm validation to 22nm commercial scaling, with Tier-1 foundry tape-out targeted for Q3 2026. • Compute-in-Memory (CIM) integrated directly onto the RRAM array, improving energy efficiency and lowering latency. • 22nm scaling expected to support applications across exoskeletons, prosthetics, medical implants and industrial robotics. • Platform underpins dorsaVi’s strategy to evolve FDA-cleared wearable sensors into autonomous on-body AI inference systems. • Standard CMOS manufacturing approach supports cost-effective, high-volume commercial production. @dorsavi @dorsaviUSA Link to the Announcement: cdn-api.markitdigital.com/apiman-gateway…

keyboard_arrow_left Previous keyboard_arrow_right Next

Battery storage is now growing even faster than EVs and it’s completely changing the lithium demand story. The market used to view lithium mainly as an EV battery material. But the next major wave is grid-scale energy storage, where massive battery deployments are needed to support renewable power and AI-driven electricity demand. As global battery storage installations accelerate, lithium demand is expanding far beyond the auto sector. That’s why lithium prices and lithium stocks are gaining momentum again in 2026. This is no longer just an EV trend, lithium is becoming core infrastructure for the future energy economy carboncredits.com/battery-storag…

The market is finally waking up to the “AI memory wall.” A memory-focused ETF called DRAM has already surged nearly 100% since launch and pulled in around $6B in assets within weeks, one of the fastest ETF growth stories ever. The reason is simple: AI scaling is now running into memory limits, not just compute limits. Samsung, SK Hynix and Micron alone make up roughly 74% of the ETF, showing how dependent the AI industry has become on a handful of memory suppliers. Researchers are now openly describing memory bandwidth as the primary bottleneck for modern AI systems because GPUs are getting faster far quicker than DRAM and interconnect speeds can keep up. That’s why technologies like ReRAM, neuromorphic chips and in-memory computing are suddenly becoming serious conversations. aol.com/finance/dram-e…

“The AI memory wall” is quickly becoming one of the biggest themes in semiconductors. The new DRAM ETF has surged nearly 90–100% in just weeks, while attracting billions in inflows, all because investors are realizing AI scaling is now constrained by memory, not just compute. What’s driving this? HBM demand is exploding as larger AI models require massive memory bandwidth to keep GPUs fully utilized. Researchers are now openly calling memory bandwidth the primary bottleneck for modern AI systems. And unlike CPUs or GPUs, the memory market is dominated by just three major players: Samsung, SK Hynix and Micron. That concentration is creating a powerful supply squeeze and pricing cycle. This is also why next-gen ideas like ReRAM, neuromorphic chips and in-memory computing are gaining momentum. …esemiconductornewsletter.substack.com/p/dram-etf-the…

Nanoveu (ASX: $NVU) Expands Defence & Drone Strategy Through Proposed Spinoff Robotics Acquisition Key Highlights • Proposed acquisition of Spinoff Robotics adds proprietary drone platforms and aerial robotics expertise in-house. • Strengthens Nanoveu’s full-stack drone ecosystem across ECS-DoT silicon, ALICE drones, METRON sensing and on-device AI. •Targeting defence and commercial UAV markets, with the military drone market forecast to reach ~US$98.24B by 2033. • ALICE tethered drones designed for high-resilience surveillance in GPS-denied environments. • METRON provides sub-millimetre sensing and AI-powered anomaly detection capabilities. • ECS-DoT drone flight trials underway with a specialist US partner. • Spinoff technologies previously validated with Tier-1 Singapore customers. @nanoveuofficial Link to the Announcement: wcsecure.weblink.com.au/clients/nanove…

keyboard_arrow_left Previous keyboard_arrow_right Next

China’s lithium prices just hit their highest level in more than two years, and the market is starting to realize this isn’t a short-term spike. Tight raw material supply, rising EV demand, and accelerating energy storage deployment are pushing battery-grade lithium carbonate above ¥177,000 per ton again. Futures markets are also climbing fast as supply concerns grow. The conversation around lithium used to be about oversupply. Now the focus is shifting back to deficits, strategic resources, and long-term energy infrastructure demand. Lithium stocks are rallying again as investors recognize lithium’s growing role in powering the AI, EV, and energy storage boom. spglobal.com/energy/en/news…

According to S&P Global, automotive LPDDR4 DRAM prices jumped nearly 70% YoY by January 2026 as memory makers shifted more capacity toward high-margin HBM for AI data centers. What’s happening is bigger than a normal chip cycle: AI servers are consuming enormous amounts of memory, and hyperscalers are locking in supply years ahead. Some reports now estimate AI datacenters could consume 70% of global DRAM output in 2026. That leaves automakers, smartphones and consumer electronics competing for the same shrinking pool of DRAM and NAND supply. This is also why ReRAM, neuromorphic computing and in-memory architectures are becoming more important. spglobal.com/automotive-ins…

Lithium is becoming one of the most important commodities of the next decade. The market used to focus only on EV batteries. Now the real demand surge is coming from energy storage, AI data centers, and global electrification. Battery-grade lithium prices have already rebounded to around $24,000 per ton in 2026 as supply tightens and demand accelerates. Analysts are now warning of a potential lithium supply deficit ahead. That’s why lithium stocks are moving sharply higher again, this is no longer just a battery story, it’s a global infrastructure story. carboncredits.com/lithium-prices…

Microsoft revealing that $25B of its AI spend is going toward memory and chips is a huge signal for where AI infrastructure is headed. The conversation used to be all about GPUs. Now the real pressure point is memory bandwidth and data movement. AI systems are consuming massive amounts of HBM, DRAM and high-speed storage just to keep models running efficiently. That’s why technologies like ReRAM and neuromorphic computing are suddenly getting serious attention. The idea is simple: reduce the constant back-and-forth between processor and memory by allowing computation to happen inside the memory layer itself. AI scaling is no longer just a compute problem, it’s becoming a memory architecture problem byteiota.com/memory-chip-cr…

Samsung hitting the $1T valuation mark proves AI’s biggest bottleneck is no longer compute, it’s memory. HBM, DRAM and NAND are now the backbone of the AI race, with demand exploding faster than supply can keep up. But the really interesting part is what comes next. Technologies like ReRAM and neuromorphic in-memory computing are starting to gain attention because moving data between memory and processors is becoming too slow and power-hungry for future AI systems. The next AI breakthrough may not come from bigger GPUs, it could come from memory that can actually think and compute on its own. byteiota.com/samsung-hits-1…

Pioneer Minerals (ASX: $PMM) Identifies New Cross-Cutting Structures at Springfield Tungsten, Gold & Gallium Project Key Highlights • High-resolution LiDAR interpretation completed over the Springfield Project in Idaho, USA. • Newly identified NE-trending fault interpreted to crosscut directly through the historic Springfield Mine area. • Cross-cutting structures considered prospective fluid pathways for tungsten, gold and gallium mineralisation. • Springfield gallium results include up to 128.7 ppm Ga₂O₃, confirming widespread gallium enrichment across the system. • Tungsten assays returned up to 2.98% WO₃ and gold assays up to 7.75 g/t Au. • Structural interpretation to support upcoming EM surveys and maiden drill targeting at Springfield. @PioneerMins Link to the Announcement: clients3.weblink.com.au/clients/pionee…

keyboard_arrow_left Previous keyboard_arrow_right Next

dorsaVi (ASX: $DVL) Advance DoD-Funded Program Targeting Military Training Injury Reduction Key Highlights • Georgia Southern University renewed its DoD-funded research agreement with dorsaVi for a further 12 months. • Study uses wearable sensors to monitor gait and running patterns in real-world military training environments. • Program targets reduction of stress fractures and load-bearing training injuries through real-time biomechanical feedback. • Platform captures key running metrics including impact force, cadence, ground contact time and foot strike pattern. • Technology builds on peer-reviewed research focused on reducing tibial shock and improving running technique in the field. • dorsaVi previously worked with U.S. fighter pilots capturing G-force data, strengthening its defense application credentials. • Research also supports broader exoskeleton and robotics strategy using sensor intelligence and neuromorphic systems. @dorsavi @dorsaviUSA Link to the Announcement: cdn-api.markitdigital.com/apiman-gateway…

keyboard_arrow_left Previous keyboard_arrow_right Next

Phoenix Global is pleased to highlight one of its standout ASX microcap outcomes with Viridis Mining & Minerals (ASX: $VMM). Since Phoenix Global’s initial coverage in Feb 2023, VMM rose from ~$0.28 to a peak of ~$2.86, delivering a ~904% peak return. Today, VMM remains ~800% above initial entry, reflecting the impact of consistent investor communication, targeted market visibility and strategic exposure. @Viridis_VMM

Shares of Sandisk ($SNDK) have already soared 429% in 2026, after the company posted a massive 252% YoY revenue jump to $5.95B, and now some analysts are projecting around a $4,000 price target AI data centres + AI devices are driving a huge storage boom, with NAND prices projected to surge another 234% in 2026 amid ongoing shortages. What this really shows is how desperate the market is becoming for faster and more efficient memory solutions as AI scales globally. The next wave won’t just be about producing more NAND, it’ll be about entirely new memory architectures designed specifically for AI workloads. That’s where technologies like ReRAM are starting to become hard to ignore. finance.yahoo.com/markets/stocks…