home.social

#nand — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #nand, aggregated by home.social.

  1. The Q3 memory numbers are in: DRAM contracts headed up 13–30%, NAND 10–40%, depending on the forecast. Earnings season backs it up — record quarters at Samsung and Micron, with guidance still climbing. SK hynix now expects the shortage to run past 2030, with 2027 the tightest year yet. The full market update, including what it means for DDR4/DDR5 values:
    buysellram.com/blog/market-upd
    #DRAM #MemoryPrices #MemoryShortage #Semiconductors #AIInfrastructure #DataCenter #ServerHardware #tech #NAND

  2. The Q3 memory numbers are in: DRAM contracts headed up 13–30%, NAND 10–40%, depending on the forecast. Earnings season backs it up — record quarters at Samsung and Micron, with guidance still climbing. SK hynix now expects the shortage to run past 2030, with 2027 the tightest year yet. The full market update, including what it means for DDR4/DDR5 values:
    buysellram.com/blog/market-upd
    #DRAM #MemoryPrices #MemoryShortage #Semiconductors #AIInfrastructure #DataCenter #ServerHardware #tech #NAND

  3. IBM股價寫下54年來最慘一日,華爾街讀出更大訊號:AI排擠效應開始反噬
    TNL國際編譯 2026-07-15 15:49:00 CST
    IBM 7月14日崩跌25%,跌幅超越1987年「黑色星期一」、市值蒸發550億美元。因為客戶把買軟體的錢挪去搶購AI硬體,也拖累ServiceNow、Salesforce等軟體股。
    https://www.thenewslens.com/article/269189
    #OpenAI #半導體 #股價 #三星 #基礎設施 #台積電 #伺服器 #IBM #記憶體 #AI晶片 #華爾街 #NAND #軟體 #美光 #Anthropic #HBM #硬體 #科技 #SK海力士 #DRAM #SaaS
  4. Samsung Fast-Tracks Yongin Fab to 2029 as Global Memory Expansion Race Heats Up — BigGo Finance

    The memory chip shortage crisis ignited by surging global artificial intelligence (AI) demand is pushing South Korea’s semiconductor…
    #EuropeSays #Korea #KR #SamsungElectronics #DRAM #HBM #HowardLutnick #KwakNoh-jung #MicronTechnology #NAND #NVIDIA #Samsung #SKhynix #Yonginsemiconductorcluster
    europesays.com/korea/83429/

  5. Samsung estimates its Q2 2026 profit to be around $58.56 billion

    Samsung Electronics has announced its earnings estimates for the second quarter (Q2) of 2026. The company expects revenue…
    #EuropeSays #Korea #KR #Samsung #DRAM #HBM #NAND #SamsungDeviceSolutions #SamsungElectronics #SamsungFoundry #SamsungGroup #SamsungMemory
    europesays.com/korea/77149/

  6. Palantir vs. SanDisk: Two Paths to AI Riches, One Winner

    Quick Read Both PLTR and SNDK posted blowout quarters, with Palantir’s U.S. commercial revenue surging 133% and SanDisk’s…
    #NewsBeep #News #Artificialintelligence #AI #ArtificialIntelligence #CA #Canada #commercialrevenue #NAND #palantir #Sandisk #Technology #top10AIstocks
    newsbeep.com/ca/776303/

  7. AI inference keeps getting pricier, and a lot of that cost is memory. High Bandwidth Flash (HBF) aims to change that: HBM-class speed, up to ~16x the capacity, far less per GB. What it is, how it could cut inference cost, and why it isn't shipping yet. buysellram.com/blog/high-bandw

    #HBF #HighBandwidthFlash #AIinference #HBM #AImemory #GPU #DataCenter #Semiconductors #NAND #SKhynix #SanDisk #MemoryWall

  8. Micron posted record DRAM revenue but can fill only ~two-thirds of demand. Three reasons supply won't catch up: HBM eats the wafers, new fabs are years out, and output is pre-sold. Pulled server memory is now real supply.

    buysellram.com/blog/high-bandw

    #HBF #HighBandwidthFlash #AIinference #HBM #AImemory #GPU #DataCenter #Semiconductors #NAND #SKhynix #SanDisk #MemoryWall

  9. AI inference keeps getting pricier, and a lot of that cost is memory. High Bandwidth Flash (HBF) aims to change that: HBM-class speed, up to ~16x the capacity, far less per GB. What it is, how it could cut inference cost, and why it isn't shipping yet.
    buysellram.com/blog/high-bandw
    #HBF #HighBandwidthFlash #AIinference #HBM #AImemory #GPU #DataCenter #Semiconductors #NAND #SKhynix #SanDisk #MemoryWall

  10. If you run AI models in production, you've probably noticed how much of the cost traces back to memory.

    Every token a model generates is served from fast memory sitting right next to the GPU. That memory — High Bandwidth Memory, or HBM — is scarce, expensive, and effectively sold out for 2026. It's the same crunch that pushed DDR5 prices several times higher this past year. For long-context models and agents, memory is now the single largest cost driver in inference.

    High Bandwidth Flash (HBF) is built to attack that. The idea: take the cheap, dense flash from SSDs and rebuild it to keep pace with a GPU — much of HBM's bandwidth, up to 8–16x the capacity, at a fraction of the cost per GB. In one vendor simulation, an HBM+HBF hybrid hit up to 2.69x higher throughput per watt than HBM alone.

    The catch: HBF isn't shipping. Samples are due in late 2026, first systems in 2027 — the savings so far live in simulations, not invoices.

    New explainer covers what HBF is, how it would cut cost, the alternatives (CXL, processing-in-memory, LPDDR modules), and how seriously to take the claim.

    buysellram.com/blog/high-bandw

  11. Running AI in production? A lot of the bill is memory. Every token a model generates is served from fast memory next to the GPU — and that memory, HBM, is scarce and expensive.
    HBF is built to attack that cost: much of HBM's bandwidth at a fraction of the price per GB. A new explainer covers what HBF is, how it could lower inference costs, and how close it actually is.

    buysellram.com/blog/high-bandw

    #HBF #HighBandwidthFlash #AIinference #HBM #AImemory #GPU #DataCenter #NAND #SanDisk #MemoryWall

  12. Running AI in production? A lot of the bill is memory. Every token a model generates is served from fast memory next to the GPU — and that memory, HBM, is scarce and expensive.
    HBF is built to attack that cost: much of HBM's bandwidth at a fraction of the price per GB. A new explainer covers what HBF is, how it could lower inference costs, and how close it actually is.
    buysellram.com/blog/high-bandw

    #HBF #HighBandwidthFlash #AIinference #HBM #AImemory #GPU #DataCenter #NAND #SanDisk #MemoryWall #tech

  13. Running AI in production? A lot of the bill is memory. Every token a model generates is served from fast memory next to the GPU — and that memory, HBM, is scarce and expensive.
    HBF is built to attack that cost: much of HBM's bandwidth at a fraction of the price per GB. A new explainer covers what HBF is, how it could lower inference costs, and how close it actually is.
    buysellram.com/blog/high-bandw

    #HBF #HighBandwidthFlash #AIinference #HBM #AImemory #GPU #DataCenter #NAND #SanDisk #MemoryWall #tech

  14. Running AI in production? A lot of the bill is memory. Every token a model generates is served from fast memory next to the GPU — and that memory, HBM, is scarce and expensive.
    HBF is built to attack that cost: much of HBM's bandwidth at a fraction of the price per GB. A new explainer covers what HBF is, how it could lower inference costs, and how close it actually is.

    buysellram.com/blog/high-bandw

    #HBF #HighBandwidthFlash #AIinference #HBM #AImemory #GPU #DataCenter #NAND #SanDisk #MemoryWall

  15. Running AI in production? A lot of the bill is memory. Every token a model generates is served from fast memory next to the GPU — and that memory, HBM, is scarce and expensive.
    HBF is built to attack that cost: much of HBM's bandwidth at a fraction of the price per GB. A new explainer covers what HBF is, how it could lower inference costs, and how close it actually is.

    buysellram.com/blog/high-bandw

    #HBF #HighBandwidthFlash #AIinference #HBM #AImemory #GPU #DataCenter #NAND #SanDisk #MemoryWall

  16. “Memory-maker Micron has found a way to keep #prices for its products sky-high for another five years, by signing 16 “strategic customer agreements” #SCA s that include a #FloorPrice the company says comes with “a very robust #GrossMargin for #Micron, well above our peak quarterly margins in any past cycle.”

    #DRAM / #NAND / #economics <theregister.com/systems/2026/0>

  17. “Memory-maker Micron has found a way to keep #prices for its products sky-high for another five years, by signing 16 “strategic customer agreements” #SCA s that include a #FloorPrice the company says comes with “a very robust #GrossMargin for #Micron, well above our peak quarterly margins in any past cycle.”

    #DRAM / #NAND / #economics <theregister.com/systems/2026/0>

  18. Apple has substantially raised prices across multiple product lines in response to shortages of memory and storage. This has driven sharp increases in wholesale costs for DRAM and NAND memory - the so‑called “RAMpocalypse”.

    computing.co.uk/news/2026/chip

    #technews #apple #mac #ipad #dram #nand #rampocalypse

  19. Apple has substantially raised prices across multiple product lines in response to shortages of memory and storage. This has driven sharp increases in wholesale costs for DRAM and NAND memory - the so‑called “RAMpocalypse”.

    computing.co.uk/news/2026/chip

    #technews #apple #mac #ipad #dram #nand #rampocalypse

  20. 記憶體漲272%、iPhone漲200美元,庫克卸任前夕遇「百年水災」
    TNL國際編譯 2026-06-20 19:42:00 CST
    庫克卸任前對《華爾街日報》罕見表態「漲價無法避免」,iPhone 18 Pro起價恐從1099漲至1299美元。AI資料中心吃掉全球記憶體產能,但三星、SK海力士與台廠卻創紀錄獲利。
    https://www.thenewslens.com/article/268694
    #輝達 #南亞科 #台灣 #iPhone 18 Pro #高頻寬記憶體 #手機 #RAMageddon #華邦電 #旺宏 #記憶體 #力積電 #蘋果 #毛利率 #AI晶片 #AI伺服器 #三星電子 #電動車 #消費電子產業 #NAND #努斯 #晶圓廠 #美光 #iPhone #HBM #郭明錤 #科技 #SK海力士 #筆電 #韓國 #庫克 #DRAM #供應鏈 #漲價