GPU makers are redirecting manufacturing capacity toward AI chips, triggering a memory shortage that has driven up prices for graphics cards and related components. JP Morgan projects DRAM prices will rise 400 percent between 2024 and the end of 2026, while Tom's Hardware reports Nvidia's RTX 5090 now sells for nearly $5,000 above its original $1,999 price. The price surge has also hit consoles, laptops and resale markets, with PC shipments falling as a result.
engadget.com
· 2026-11-07
Tom's Hardware reviewed TeamGroup's GA Pro 2TB SSD, a PCIe 5.0 drive built on an InnoGrit IG5666 controller with DRAM and YMTC 232-layer TLC memory, priced at $331.99. The outlet found the drive ran hot and delivered inefficient, generally low performance despite some decent results at QD8 and acceptable steady-state write speeds, prompting it to retest and contact TeamGroup about the results.
tomshardware.com
· 2026-11-07
TSMC publicly rejected a report from The Information claiming it was exploring 3D DRAM production at Nvidia's request amid global memory shortages. The company stated it remains a pure-play logic foundry, though the original report cited two sources saying TSMC had begun preparing plans for the initiative under VP Hsu Kuo-chin, a former Micron Taiwan executive.
tomshardware.com
· 2026-11-07
Raspberry Pi CEO Eben Upton announced another price increase for the 16GB Raspberry Pi 5, now priced at $385, up $80 from before. The 16GB Compute Module 5 also rises by $40, while the Raspberry Pi 500+ keeps its current price despite having 16GB of RAM. Upton attributed the hike to ongoing shortages of 16GB DRAM modules and a lack of alternative suppliers.
techspot.com
· 2026-11-06
TrendForce reports conventional DRAM contract prices jumped roughly 93-98% quarter over quarter in early 2026, driven partly by manufacturers shifting production toward AI data center memory. Micron exited its Crucial consumer RAM business this year citing AI demand, while HP says memory and storage will make up about 35% of its PC component costs in 2026, up from 15-18% previously, and has warned of PC price hikes. Used DDR4 kits offer some savings over new ones but still cost far more than in 2020.
engadget.com
· 2026-11-02
Sam Altman said on the YouTube channel Cheeky Pint that it's inefficient for consumers to keep DRAM sitting mostly idle in personal computers, arguing that memory would be better utilized in AI data centers. Elon Musk responded on X with a mocking meme referencing a WWII film about hunting for hidden items, framing Altman's comment as an attack on consumer RAM ownership.
tomshardware.com
· 2026-11-02
Raspberry Pi CEO Eben Upton announced another price hike for the 16GB Raspberry Pi 5, blaming a lack of alternative DRAM supply combined with shortages of other components. Upton advised customers to avoid paying for more memory than they need, suggesting lower-memory or older Pi models as alternatives. The 2GB Raspberry Pi 5 configuration had already seen a price increase the previous month.
theverge.com
· 2026-11-02
Biwin has launched the M560, a PCIe 5.0 DRAM-less SSD built around the SMI SM2504XT controller paired with 218-layer BiCS8 TLC memory from Kioxia. The drive is available in 1TB ($189.99) and 2TB ($364.99) capacities, rated up to 11,000 MB/s sequential read and 10,000 MB/s sequential write, with a 5-year warranty and endurance ratings of 750TB and 1,500TB respectively.
tomshardware.com
· 2026-11-01
OpenAI CEO Sam Altman met with roughly 50 senior Samsung executives at the company's Resources Development Center in Yongin, with discussions focused on AI technology direction and future business opportunities rather than a new memory supply deal. Samsung confirmed the talks built on an October 2025 letter of intent covering AI infrastructure cooperation, including memory chips for OpenAI's Stargate project. The meeting follows Altman's recent remarks suggesting underused DRAM in personal computers should instead support data centers.
tomshardware.com
· 2026-11-01
A technical blog post explains that a CPU showing 100% utilization may not actually be fully working, because much of that time can be spent waiting on memory accesses rather than executing instructions. It breaks down the latency gap between CPU cores and memory tiers, noting RAM access can take 50-100ns versus roughly 1ns for L1 cache, meaning a CPU could execute 400-800 instructions in the time it takes to fetch data from RAM once.
pivotlake.io
· 2026-10-31
OpenAI CEO Sam Altman said in an interview with YouTuber Cheeky Pint that it is 'quite terrible' for people to keep large amounts of DRAM sitting idle in personal computers, arguing memory should be maximized in the cloud given the current supply shortage. He made the comment while discussing local AI inference, where models run directly on a user's device rather than in data centers.
engadget.com
· 2026-10-31
In a Cheeky Pint interview, OpenAI CEO Sam Altman said it's "quite terrible" that DRAM sits mostly unused inside consumer devices, since he estimates it's active only about 5% of the time. He argued memory would be better utilized in data centers, which run hardware near-continuously, citing PCs with local AI features and open-source LLMs as examples of inefficient use.
techspot.com
· 2026-10-31
GoKawiil