Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!


New on LowEndTalk? Please Register and read our Community Rules.

All new Registrations are manually reviewed and approved, so a short delay after registration may occur before your account becomes active.

How are you getting RAM

2»

Comments

  • I'm buying it on eBay before Sam Altman gets it.

  • SWAP

  • For my personal use:

    • Spare 192GB of DDR5 memory I purchased Oct last year has increased 3x to 8x in price since!
    • Spare 2TB/4TB NVMe drives purchased last year have increased in price by 2x to 5x
    • Spare 256GB/512GB microsd cards (lots of security cameras) have increased 2x to 3x in price

    Seriously don't know how web hosting providers going to handle this for next 4-6yrs as it ain't going to get cheaper until at least 2032 I think.

  • rpqurpqu Member

    @eva2000 said:
    Seriously don't know how web hosting providers going to handle this for next 4-6yrs as it ain't going to get cheaper until at least 2032 I think.

    I think of following reasons:
    1. DDR6 soon™
    2. Big 3 + CXMT capacity increase by 2028-2029
    3. Market tightening credit line and regulatory pressure, which may led to early decomm. and lowered demand

    It may or may not happen.

  • SplitIceSplitIce Member, Host Rep

    @rpqu said: Big 3 + CXMT capacity increase by 2028-2029

    Thats still a long way away.

  • @rpqu said:

    @eva2000 said:
    Seriously don't know how web hosting providers going to handle this for next 4-6yrs as it ain't going to get cheaper until at least 2032 I think.

    I think of following reasons:
    1. DDR6 soon™
    2. Big 3 + CXMT capacity increase by 2028-2029
    3. Market tightening credit line and regulatory pressure, which may led to early decomm. and lowered demand

    It may or may not happen.

    Honestly there is also the fact a lot of these hyperscalers have RAM+GPUs sitting in cold storage because they do not have the power to set them up anywhere, eventually they have to be sold I'm assuming.

  • rpqurpqu Member

    @SplitIce said:

    @rpqu said: Big 3 + CXMT capacity increase by 2028-2029

    Thats still a long way away.

    We're 7 weeks from 2027.

    @neighhouse said:

    @rpqu said:

    @eva2000 said:
    Seriously don't know how web hosting providers going to handle this for next 4-6yrs as it ain't going to get cheaper until at least 2032 I think.

    I think of following reasons:
    1. DDR6 soon™
    2. Big 3 + CXMT capacity increase by 2028-2029
    3. Market tightening credit line and regulatory pressure, which may led to early decomm. and lowered demand

    It may or may not happen.

    Honestly there is also the fact a lot of these hyperscalers have RAM+GPUs sitting in cold storage because they do not have the power to set them up anywhere, eventually they have to be sold I'm assuming.

    Yes. If they prioritize new gen, the previous one won't get racked. If they prioritize previous gen, their efficiency will take a hit and it will reduce the demand of new gen.

  • @rpqu said:

    @SplitIce said:

    @rpqu said: Big 3 + CXMT capacity increase by 2028-2029

    Thats still a long way away.

    We're 7 weeks from 2027.

    @neighhouse said:

    @rpqu said:

    @eva2000 said:
    Seriously don't know how web hosting providers going to handle this for next 4-6yrs as it ain't going to get cheaper until at least 2032 I think.

    I think of following reasons:
    1. DDR6 soon™
    2. Big 3 + CXMT capacity increase by 2028-2029
    3. Market tightening credit line and regulatory pressure, which may led to early decomm. and lowered demand

    It may or may not happen.

    Honestly there is also the fact a lot of these hyperscalers have RAM+GPUs sitting in cold storage because they do not have the power to set them up anywhere, eventually they have to be sold I'm assuming.

    Yes. If they prioritize new gen, the previous one won't get racked. If they prioritize previous gen, their efficiency will take a hit and it will reduce the demand of new gen.

    >

    Power being the real bottleneck is an interesting angle. Even when hyperscalers eventually offload un-racked or early-decommissioned inventory, that secondary wave tends to lag behind and skew toward high-density RDIMMs rather than the 16GB/32GB sticks budget hosts rely on. In the meantime, smaller providers are left balancing older DDR4 platforms — where used stick pricing still swings wildly by region — against newer DDR5 builds with multi-month wholesaler lead times.

  • @eva2000 said:
    For my personal use:

    • Spare 192GB of DDR5 memory I purchased Oct last year has increased 3x to 8x in price since!
    • Spare 2TB/4TB NVMe drives purchased last year have increased in price by 2x to 5x
    • Spare 256GB/512GB microsd cards (lots of security cameras) have increased 2x to 3x in price

    Seriously don't know how web hosting providers going to handle this for next 4-6yrs as it ain't going to get cheaper until at least 2032 I think.

    Crash because all these datacenters ain't getting built and AI is becoming a race to the bottom.

    Thanked by 1rpqu
  • The nvidia scam was discovered...

  • I used to stay awake all night downloading it. Now have my AI agent spin up more agents as it downloads it exponentially

  • Price is ridiculous

  • @TimboJones said:
    The nvidia scam was discovered...

    Yup, TSMC demanding cash upfront : https://www.reddit.com/r/pcmasterrace/comments/1xzjodl/with_jensen_huangs_circular_financing_engine_gone/

    SSD spot prices drop 20% : https://www.reddit.com/r/LocalLLaMA/comments/1xzixp0/ssds_laptops_and_phones_may_get_cheaper_as/

    I dont think NVIDIA will be able to take delivery for the silicon they contracted from TSMC.

  • @DrNutella said:
    I used to stay awake all night downloading it. Now have my AI agent spin up more agents as it downloads it exponentially

    I had to check the topic again, I had assumed you were talking about porn.

  • rpqurpqu Member

    @rincewind said:

    @TimboJones said:
    The nvidia scam was discovered...

    Yup, TSMC demanding cash upfront : https://www.reddit.com/r/pcmasterrace/comments/1xzjodl/with_jensen_huangs_circular_financing_engine_gone/

    SSD spot prices drop 20% : https://www.reddit.com/r/LocalLLaMA/comments/1xzixp0/ssds_laptops_and_phones_may_get_cheaper_as/

    I dont think NVIDIA will be able to take delivery for the silicon they contracted from TSMC.

    Great, hopefully my prediction materialized, I was wrong about the mechanism though (nvidia fallout)

    @rpqu said:
    What this means for hardware pricing at Q4 2027 / Q1 2028

    Market rationalization here means a specific trigger: hyperscalers stop or significantly reduce new orders, either from token supply glut making additional capacity economically indefensible, or from liquidity pressure as medium-term corporate bond markets tighten against sustained negative ROI. Hyperscalers raised $121B in new debt in 2025 alone [9], with Morgan Stanley and JP Morgan projecting $1.5T in total debt issuance required over the next few years [10]. Oracle already faces a financing gap from FY2027, with Barclays warning it could run out of cash by November 2026 at current trajectory [11]. CDS spreads — the bond market's forward-looking default insurance — have been rising across the sector [12].

    This aligns roughly with the end of NVIDIA's currently committed production pipeline. If either condition materializes earlier — and the oversupply math suggests it could — the price corrections described below arrive ahead of this timeline, potentially as early as mid-2027.

    Compute (GPU)

    The 2026 installed Blackwell base faces a competitive token economics gap of 350x against Vera Rubin alone, 12,250x against VR+Groq.Hourly GPU rental rates on H100/H200 class hardware will reprice downward as operators compete for utilization against a market where newer hardware produces orders of magnitude more output at the same power cost. The stranded asset isn't theoretical — it's hardware being delivered today against a depreciation schedule that assumed 5 years of competitive relevance.

    RAM (DDR5)

    Prices will be lower than the February 2026 peak. The 400% ramflation was driven by AI factory demand absorbing total DRAM production capacity.As HBM4 and SOCAMM displace DDR5 for the highest-demand AI workloads, and new fab capacity from Micron's Hiroshima expansion comes online in 2027, the DDR5 market should see meaningful relief. The exact magnitude depends on whether consumer and enterprise non-AI demand recovers the slack, or whether the market overshoots into a glut. Directionally: lower, timeline and depth uncertain.

    NVMe

    SOCAMM LPDRAM handling warm KV cache in-flight reduces the NVMe use case to cold KV archival only — sessions idle for hours or days.This is a significant demand reduction for the high-performance NVMe tier. The workload that justified $50-80/TB NVMe pricing — fast random-access KV staging — is being absorbed by the SOCAMM tier on-board. What remains for NVMe is large-block sequential cold storage, a workload that does not require NVMe's random-access performance premium. Expect pricing pressure on datacenter NVMe as the AI workload profile shifts.

    HDD

    Inconclusive near-term, but NVIDIA has drawn the boundary for us.

    The Vera Rubin POD architecture defines four explicit memory tiers [13]:

    Tier Hardware NVIDIA product Latency
    Hot KV HBM4 on Rubin GPU ✅ Nanoseconds
    Warm KV SOCAMM 2TB on Vera CPU ✅ Microseconds
    Warm-cold KV STX rack NVMe via BlueField-4 ✅ Milliseconds
    Cold archive — ❌ Seconds+

    NVIDIA's BlueField-4 STX rack extends GPU memory into NVMe for active KV reuse — sessions resuming within hours. It is not designed for day-scale or month-scale retention. The cold archive tier is explicitly outside the POD boundary, and NVIDIA has no product there.

    At Vera Rubin's 700M tokens/sec throughput, even a 1% session persistence rate generates petabytes of cold KV and artifact data per day per POD. The I/O profile — large sequential writes, infrequent full-block reads, latency tolerance in seconds — is exactly where HDD is cost-optimal versus NVMe.

    S3-compatible object storage on CMR nearline HDD, positioned as the cold archive tier that the NVIDIA POD architecture intentionally leaves unaddressed.

    Thanked by 1TimboJones
  • @TimboJones said:

    @DrNutella said:
    I used to stay awake all night downloading it. Now have my AI agent spin up more agents as it downloads it exponentially

    I had to check the topic again, I had assumed you were talking about porn.

    Too busy to ask AI

Sign In or Register to comment.