• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Search results

  1. L

    AMD Radeon RX 9070 GRE Tested, Fills Gap Between RX 9060 XT and RX 9070

    the 9070 at $620.... seem to offer quite a bit more, maybe they do not think they will be possible to find at those price for long.
  2. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    need here will tend to be how much more people are ready to pay for sockets, Nvidia can do quite a lot with sockets on their AI system with LPCAMM type of connexion, people are ready to pay a lot for the ability to change one of those gpu if they fail to keep the 8 GPU rack going, the more...
  3. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    They will move it more and more on customer machine to save money, the version they made for Nvida machine will use local power quite a bit would be my guess, the era of cheap subsidized AI is almost over...
  4. L

    More DLSS...

    https://www.youtube.com/watch?v=5CWIcKaDCSU FPS in pragmata seem quite similar (virtually the same to a tiny bit lower it seem) on the blackwell in the video
  5. L

    [SemiAccurate] upcoming Nvidia chip has a problem that doesn't have a fix as of now

    Models are getting more and more capable at 32GB/64GB has well, the N1X 650 and lowers, but will still be a fortune in the current memory world https://videocardz.com/newz/nvidia-n1x-650-and-n1x-675-appear-in-lenovo-yoga-pro-7-listings If you want to talk to your computers that use blender-a...
  6. L

    [SemiAccurate] upcoming Nvidia chip has a problem that doesn't have a fix as of now

    For those top of the line, over $4400..... probably, it is the same SOC than the DGX Spark and the same a lot of memory with a nice monitor. They can remove the multi-node NVLink, 400gbps connectX Mellanox), those laptop will have regular cheap wifi-7/10 gbps or less ethernet port in...
  7. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    Maybe but now all modern arm phone cpu (mandatory for those) and AWS Graviton, google server cpu, AmpereOne, Snapdragon X2 and Nvidia N1X has both (SVE and SVE2), Apple going for SME-SME2 (or arm going into Apple AMX and calling it SME...) only do throw a bit of a wrench here but they have such...
  8. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    There is arm device tree and other standard outthere now: https://www.arm.com/architecture/system-architectures/systemready-compliance-program That why standard Linux for arm work for the new Nvidia Vera cpu system, Nvidia usually use UEFI like x86 cpus...
  9. L

    [SemiAccurate] upcoming Nvidia chip has a problem that doesn't have a fix as of now

    https://youtu.be/suhxvEb5l2s Curious about lower end sku if they make them, the 128GB full version is such a big amount of computer, the market for such workstation in a laptop format must be a bit small, like for full StrixHalo affair. Local agents change that math for who can use 128gb of...
  10. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    That seem quite ISA independent for the most part, that could well be true but it seem there often some missing because of ARM fixed instruction side or rules that you must put data in register before manipulating it which mean that in this case X,y,z over x86 performance will always be there...
  11. L

    GN: Collapse of Personal Computing - Investigation Into the Destruction of Ownership (video)

    Table 1: Preliminary Worldwide PC Vendor Unit Shipment Estimates for 1Q26 (Thousands of Units) Company 1Q26 Shipments 1Q26 Market Share (%) 1Q25 Shipments 1Q25 Market Share (%) 1Q26-1Q25 Growth (%) Lenovo 16,645 26.5 15,199 25.2 9.5 HP Inc. 12,142 19.3 12,766 21.1 -4.9 Dell 10,337...
  12. L

    GN: Collapse of Personal Computing - Investigation Into the Destruction of Ownership (video)

    In a world where you can buy a Neo for $599, it is still not 1994 level of expensive to own a personal computer/smarthphone that can do a lot. https://www.tweaktown.com/news/110956/global-pc-shipments-actually-grew-in-q1-2026-despite-memory-crisis-and-price-crunch/index.html Global PC shipments...
  13. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    not sure looking at apple or nvidia core count if that what going on here. 88 cores was not an high amount in phoronix benchmark (that went from 64 to 256 x86 cores in its competition), same for Apple M5 cpu, they often have similar or smaller core count then their intel-amd competition. why ...
  14. L

    GPU prices — is the worst behind us ?

    who is they ? at the current margin, I am not sure they mind selling stuff at all.... Intel consumer business is still more profitable, memory maker has well, Apple I do not imagine has plan to kill the expensive iPhone to turn it into a dumb terminal until sales ever go down, Nvidia just...
  15. L

    More DLSS...

    If you have modern (lovelace and up) Tensor core it could be very small performance impact, at least that Nvidia claim: Efficient Denoiser: The new model delivers 35% more compute capability, and processes 20% more parameters, while maintaining similar performance to the previous model. Which...
  16. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    they are not with the current power usage as they are not power tuned up yet (that the reason giving to phoronix about why they would not be allowed to run power metric on them) You cannot trust that fast fourier transform (a good for gpu workload), ran by phoronix that compiled the open source...
  17. L

    AMD’s new pitch: our old tech is so good you should just keep using it

    could see the scenario of using old stock of binned down but still quite good 8 core zen4 and putting cache on them to add value making a lot of sense, as tsmc 5/zen 4 3d v-cache must be quite mature now and does not mess up zen5 line up as much as a 9700x3d would
  18. L

    [SemiAccurate] upcoming Nvidia chip has a problem that doesn't have a fix as of now

    commitment to do 3 generation/5 years of them over is the announcement here,, for both the hardware partners but also the software world that important. Could be a bit of a fast pace for the competition here, if they use the yearly refresh (a la apple) on the whole stack and not just datacenter.
  19. L

    GPU prices — is the worst behind us ?

    I think you are the first one to suspect Nvidia of using secret gaming revenues to boost their datacenter number. But no data center is its own revenues segment and they separate it in 2 bucket, hyperscaler and the neo clouds, industrial-enterprise (i.e. datacenter hardware but not bought by...
  20. L

    GPU prices — is the worst behind us ?

    isn't H2 2026 cap-ex a new record high and by a lot ? Not sure anything has slowed yet.
  21. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    phoronix was super clear about this, in fact you took this from there first page, no one missed this. And datacenter CPU being tested for datacenter type of workload is not that special, and here they were able to go a little bit outside of that compiling godot game engine is not necessarily the...
  22. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    are the cpu themselve not doing this a bit ? with the way decoder cut big operation in uOps anyway, not so sure how much gain there would be here to be had. Vera big advantage here is support for much faster memory, NVLInk, being monolithic instead of cutting cost with chiplet and ISA wise...
  23. L

    NVIDIA "Vera" CPU Benchmarked: Beating Intel Xeon and AMD EPYC

    Yes Jensen is well aware and he paid 20 billions to buy Groq for a reason (they are low latency LPU... an ASIC only for LLM type inference), that do only this at an extreme level of simplicity and efficacy, build from the ground up to do this by the guy that made the google TPU a decade before...
  24. L

    Steam Deck Price Hike. top model now $949

    They sell a little amount of them, maybe 2 millions in a big year and they do make billions in net profit/free cash flow say 2.5-3B. They could easily loose $150 per unit sold yes... specially if they think it can increase sales by $90 a year in average among its buyer and pay for itself by 5...
  25. L

    Steam Deck Price Hike. top model now $949

    only reasonable way to do this, is if you can rent capacity or rent your excess if you have your own server. They probably rent more third party capacity during peak period (for the content delivery part of the cloud), akamai for example is one of those, it is not a pure steam infracstructure...
  26. L

    Steam Deck Price Hike. top model now $949

    but they could be making so little of them that the price does not really matter either way, small price tag and if you do not have volume end up bought by scalper and resold at the askable price.
  27. L

    Majority of CEOs Report Zero Payoff From AI Splurge

    It is incredibly ambitious, zero pay outside really high result that climb up to a giant payday (they are of course saying risk of getting $0 for a lot of work are high here): would have well earned those bonus if he reach them, that said everyone was saying the same for the previous deal he...
  28. L

    AI's Water and Electricity Use Soars In 2026

    They tend to do, the H20 did not sell well there when it was remade possible to be sold and I am not sure if the H100 would, you and others bring that up as a big issues, that I am not sure it is. They would still do in some ways of course but if the H100-200 sell well that would mean local...
  29. L

    Steam Deck Price Hike. top model now $949

    that would ahve been true until a PS6 becamse popular enough for game to not care and optimise for a RX 6700/3700x CPU/regular pci 4.0 nvme speed/16-20GB of available system ram + vram type of hardware. Crimson desert, resident evil requiem, pragmata, forza horizon 6, 007 first light all run...
  30. L

    Welcome to Archality - AI Inference by Enthusiasts

    Has an example of this, Nvidia "equivalent" they are making for this tech (not at all the same in some ways, but weight resident on silicon and made only for inference chips, very direct competitor) https://www.sammobile.com/news/samsung-makes-groq-3-lpu-chips-nvidia/...
  31. L

    AI's Water and Electricity Use Soars In 2026

    That one can be a bit of a lagging take, let alone the serious money Google-Meta-Amazon has been doing with their Ai datacenter all along, Anthropic is currently paying 1.25 billion a month for Collosus 1 alone. 15 billion a year for a single memphis location that did cost between 4-7 billions...
  32. L

    AI's Water and Electricity Use Soars In 2026

    Who said there was anything wrong with China developping some AI models and capacity or anything wrong with the US doing it.... That back and forth can be oversold here a bit, otherwise selling 2022's H100/AMD MI325X would not be a talking point, at no point China was ahead in anyway yet, as...
  33. L

    Microsoft steeply raises Xbox Series X/S, game prices in response to tariffs

    Seem to have been the case, could have simply been old stock unaffected by nand-ram-tariff
  34. L

    Welcome to Archality - AI Inference by Enthusiasts

    I think you are saying the same..... the cheaper the model the more the agents will use it, the better the model the more the agents will use and the cheaper/better it get new things will be available. If cars went at horse speed moving similar weight than them they would have been quite better...
  35. L

    Welcome to Archality - AI Inference by Enthusiasts

    Not necessarily exactly the same, cerebras was TSMC 14 for a long time and latest on TSMC 5, Groc on samsung node for the future chips, they were on old TSMC and global foundry until recently. SRAM does not scale well anyway and the super predeterministic dataflow goes fast on older gen affair...
  36. L

    Welcome to Archality - AI Inference by Enthusiasts

    Nvidia is going toward that direction has well, with groq, so much money and so much compute for inference that from the ground up built just for inference solution will appear a bit everywhere, AMD will have (and obviously the google-openAI-amazon has well), starting with upcoming Rubin in...
  37. L

    So why cards?

    well you had one signal that apu does not have for the board to handle, between cpu and gpu and now the gpu need its own memory controller to talk to its own memory and its own memory slot on the board, way more costly if not harder (which maybe are the same word here, it is all about cost...
  38. L

    So why cards?

    It would not be if you do it under those limitation of relatively slow gpu bandwith and relatively low power, that were all the challenge of discrete level GPU in socket would come from. i.e. we already have gpu in socket that use motherboard ram outhere, those AMD APU, we already have high...
  39. L

    AI's Water and Electricity Use Soars In 2026

    Because it tend to be illegal to build nuclear anywhere in the US in a reasonable timeline has well (would they hve China regulatory regime and building capacity, they would gladly pay a bit more if it is even actually more for nuclear), Solar and its vast intermission often need great and...
  40. L

    So why cards?

    Not soldered in vs soldered in is a big difference regardless of the socket format I think. If an Nvidia arm cpu (or mediatek or Ampere arm cpu) use a 450w-600w 5090 gpu and if you add more power-heat by not having the ram soldered and farther away from the gpu, it will need even more cooling...
Back
Top