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.

RackNerd’s BLACK FRIDAY TAKEOVER! NEW DEALS, NEW TECH, AdminBolt HOSTING + HUNDREDS OF GIVEAWAYS!

1…173174175176178

Comments

  • @TrK said:

    @DaDeveloper said:

    @TrK said:

    @op23 said:

    @TrK said:

    @Neat321 said:

    @DaDeveloper said:

    @Neat321 said:

    @DaDeveloper said:

    @TrK said:
    Who uses a dgx spark?

    I use two lol

    Deepseek flash 4.1 ?

    1017 cuz bad internet lol

    Yeah

    I like that model. I also recommend trying hy 4 flash.

    Are they on dgx spark?

    He has 2 sparks, they can easily accommodate GGUF Q4_K_M quantized hy4 flash.

    Hmm sounds good enough but Q4 won't be slow?

    No idea

    Hmm maybe I don't do ai so can't say much

    I am getting my feet wet after writing it off round the gpt 4o era

  • TrKTrK Veteran

    Looks like I need to be on my pc gotta finish some migrations

  • op23op23 Member

    @DaDeveloper said:

    @op23 said:

    @TrK said:

    @Neat321 said:

    @DaDeveloper said:

    @Neat321 said:

    @DaDeveloper said:

    @TrK said:
    Who uses a dgx spark?

    I use two lol

    Deepseek flash 4.1 ?

    1017 cuz bad internet lol

    Yeah

    I like that model. I also recommend trying hy 4 flash.

    Are they on dgx spark?

    He has 2 sparks, they can easily accommodate GGUF Q4_K_M quantized hy4 flash.

    Shoot me a recipe and I’ll set it up when I bring my internet back to Phoenix with me after thanksgiving

    https://github.com/ZackSample/hy4-flash-dgx-sparks

    Thanked by 1TrK
  • TrKTrK Veteran

    Damn it's cold I don't wanna get out of the bed, it's cozy.

  • TrKTrK Veteran

    @DaDeveloper said:

    @TrK said:

    @DaDeveloper said:

    @TrK said:

    @op23 said:

    @TrK said:

    @Neat321 said:

    @DaDeveloper said:

    @Neat321 said:

    @DaDeveloper said:

    @TrK said:
    Who uses a dgx spark?

    I use two lol

    Deepseek flash 4.1 ?

    1017 cuz bad internet lol

    Yeah

    I like that model. I also recommend trying hy 4 flash.

    Are they on dgx spark?

    He has 2 sparks, they can easily accommodate GGUF Q4_K_M quantized hy4 flash.

    Hmm sounds good enough but Q4 won't be slow?

    No idea

    Hmm maybe I don't do ai so can't say much

    I am getting my feet wet after writing it off round the gpt 4o era

    Lol 😂

  • dustincdustinc Patron Provider, Top Host, Veteran

    @ToneJon said:
    Order #7681436058 — please double my bandwidth to 6TB, thank you!

    Hi @ToneJon -- We've doubled your bandwidth 👊

    We sincerely appreciate you trusting us with your business and look forward to growing with you.

  • dustincdustinc Patron Provider, Top Host, Veteran

    @klss36 said:
    Hello, I would like to double the bandwidth.
    Invoice #27000669
    Order #6774254224
    thanks.

    Hi @klss36 -- We've just finished doubling your bandwidth!

    We look forward to continuing to exceed your expectations, we greatly appreciate the trust you've placed with us. :)

  • @TrK said:
    Looks like I need to be on my pc gotta finish some migrations

    Thanked by 1op23
  • op23op23 Member

    @TrK said:

    @op23 said:

    @TrK said:

    @Neat321 said:

    @DaDeveloper said:

    @Neat321 said:

    @DaDeveloper said:

    @TrK said:
    Who uses a dgx spark?

    I use two lol

    Deepseek flash 4.1 ?

    1017 cuz bad internet lol

    Yeah

    I like that model. I also recommend trying hy 4 flash.

    Are they on dgx spark?

    He has 2 sparks, they can easily accommodate GGUF Q4_K_M quantized hy4 flash.

    Hmm sounds good enough but Q4 won't be slow?

    Avg 20 token/sec for a highly capable model and decent quant, it will be enough for a ton of use cases.

    I am planning to use my KS-LE-6 to run qwen flash next at 3 tok/second lol. Not doing interactive coding, but I can leave it overnight to reverse engineer code bases and setup recipes. 20 tok/second would be like rocketing to moon.

    Thanked by 1TrK
  • dustincdustinc Patron Provider, Top Host, Veteran

    @txzhu said:
    Hello, please double the bandwidth for my Black Friday order!
    Order Number: 8402566568
    Invoice ID: 27003399
    Thanks!

    Hi @txzhu -- Thank You for placing an order with RackNerd! We've doubled your bandwidth 🎈

    We sincerely appreciate your business.

  • dustincdustinc Patron Provider, Top Host, Veteran

    @letacc said:
    Hello, I would like to double the bandwidth.
    Order Number: 1472611441
    Thank you very much!

    Hi @letacc -- Your bandwidth has been successfully upgraded. 👊

    Thank You for relying on RackNerd. We're here to ensure your experience is seamless!

  • dustincdustinc Patron Provider, Top Host, Veteran

    @blueroar said:
    Double Bandwidth request for #9807931973

    Hi @blueroar -- Good stuff! We sincerely appreciate you choosing RackNerd as your provider.

    If you ever need anything, feel free to reach out. By the way, we've gone ahead and doubled your bandwidth :)

  • dustincdustinc Patron Provider, Top Host, Veteran

    @dxfong said:
    Please double my bandwidth. Thank you!
    Order Number: 6016288913

    Hi @dxfong -- We've doubled your bandwidth resources 🕺

    Your business is valuable to us at RackNerd. We look forward to continuing to serve you!

  • @op23 said:

    @TrK said:

    @op23 said:

    @TrK said:

    @Neat321 said:

    @DaDeveloper said:

    @Neat321 said:

    @DaDeveloper said:

    @TrK said:
    Who uses a dgx spark?

    I use two lol

    Deepseek flash 4.1 ?

    1017 cuz bad internet lol

    Yeah

    I like that model. I also recommend trying hy 4 flash.

    Are they on dgx spark?

    He has 2 sparks, they can easily accommodate GGUF Q4_K_M quantized hy4 flash.

    Hmm sounds good enough but Q4 won't be slow?

    Avg 20 token/sec for a highly capable model and decent quant, it will be enough for a ton of use cases.

    I am planning to use my KS-LE-6 to run qwen flash next at 3 tok/second lol. Not doing interactive coding, but I can leave it overnight to reverse engineer code bases and setup recipes. 20 tok/second would be like rocketing to moon.

    It would feel slower compared to the 1017 I’m running tho lol

  • TrKTrK Veteran

    @op23 said:

    @TrK said:

    @op23 said:

    @TrK said:

    @Neat321 said:

    @DaDeveloper said:

    @Neat321 said:

    @DaDeveloper said:

    @TrK said:
    Who uses a dgx spark?

    I use two lol

    Deepseek flash 4.1 ?

    1017 cuz bad internet lol

    Yeah

    I like that model. I also recommend trying hy 4 flash.

    Are they on dgx spark?

    He has 2 sparks, they can easily accommodate GGUF Q4_K_M quantized hy4 flash.

    Hmm sounds good enough but Q4 won't be slow?

    Avg 20 token/sec for a highly capable model and decent quant, it will be enough for a ton of use cases.

    I am planning to use my KS-LE-6 to run qwen flash next at 3 tok/second lol. Not doing interactive coding, but I can leave it overnight to reverse engineer code bases and setup recipes. 20 tok/second would be like rocketing to moon.

    Fair enough, you gonna use REA or Craft anything tool for those reverse engineering tasks?

  • dustincdustinc Patron Provider, Top Host, Veteran

    @joerui said:
    Hello, I would like to double my bandwidth.
    Order Number: 3532837684
    Invoice ID: 26931576
    Thanks!

    Hi @joerui -- We've gone ahead and doubled your bandwidth.

    Thank You for doing business with RackNerd :)

  • dustincdustinc Patron Provider, Top Host, Veteran

    @Kawaii said:
    Hi Dustin,
    Order Number: 4687826274
    Please double my bandwidth, thank you!

    Hi @Kawaii -- Your bandwidth allocation has been doubled :)

    We value your trust in us, thanks for being a RackNerd customer!

  • TrKTrK Veteran

    We can't forget the context window now can we?

  • dustincdustinc Patron Provider, Top Host, Veteran

    @lukelzlz said:
    Invoice #26919669 - Double bandwidth please! Thank you!

    Hi @lukelzlz -- awesome stuff! :)

    We were able to locate your order and doubled your bandwidth resources.

  • dustincdustinc Patron Provider, Top Host, Veteran

    @tantide said:
    Hello, I would like to double the bandwidth.
    Order Number: 8082842844
    Invoice ID: #26915132
    Thanks!

    Hi @tantide -- Great news: Your bandwidth has been increased! 🥳

    We are excited to support your growth. Thank You for partnering with us!

  • dustincdustinc Patron Provider, Top Host, Veteran

    @lupeng said:
    Hello Dustin, please double my bandwidth!
    Order Number: 5180046638
    Thank you!

    Hi @lupeng -- Thank You for choosing RackNerd as your provider :)

    We've gone ahead and doubled your bandwidth resources. We sincerely appreciate you trusting us with your business.

  • TrKTrK Veteran

    Brb need to get out of bed

  • dustincdustinc Patron Provider, Top Host, Veteran

    @hello0wrld said:
    Order #2760314855 - please double my bandwidth. Thanks!!

    Hi @hello0wrld -- Thank You for placing an order with RackNerd! We've doubled your bandwidth :)

    We sincerely appreciate your business.

  • dustincdustinc Patron Provider, Top Host, Veteran

    @spikevyxel said:
    Order number: 1709869026

    Please double transfer - ty!

    Hi @spikevyxel -- Boom! Your bandwidth's been doubled 👊

    Thanks for choosing RackNerd - we're here if you need anything!

  • op23op23 Member

    @DaDeveloper said:

    @op23 said:

    @TrK said:

    @op23 said:

    @TrK said:

    @Neat321 said:

    @DaDeveloper said:

    @Neat321 said:

    @DaDeveloper said:

    @TrK said:
    Who uses a dgx spark?

    I use two lol

    Deepseek flash 4.1 ?

    1017 cuz bad internet lol

    Yeah

    I like that model. I also recommend trying hy 4 flash.

    Are they on dgx spark?

    He has 2 sparks, they can easily accommodate GGUF Q4_K_M quantized hy4 flash.

    Hmm sounds good enough but Q4 won't be slow?

    Avg 20 token/sec for a highly capable model and decent quant, it will be enough for a ton of use cases.

    I am planning to use my KS-LE-6 to run qwen flash next at 3 tok/second lol. Not doing interactive coding, but I can leave it overnight to reverse engineer code bases and setup recipes. 20 tok/second would be like rocketing to moon.

    It would feel slower compared to the 1017 I’m running tho lol

    Guess it depends if it’s any better at a specific task, just because a new model comes out it’s not necessarily worth running for sure.

    1017 is hard to beat right now it seems.

  • dustincdustinc Patron Provider, Top Host, Veteran

    @gillg said:
    Order #8263024271,please double my bandwidth. Thanks!!

    Hi @gillg -- good stuff!

    We just completed the upgrade — if you ever need anything, we'll be here to help :)

  • @dustinc never heard anything back from you about the Joe holder/ tea vessel

  • dustincdustinc Patron Provider, Top Host, Veteran

    @what99 said:
    Hi @dustinc,

    I previously requested the free double bandwidth upgrade on November 16 in the older RackNerd Black Friday thread, but the upgrade has not yet been reflected in my VPS control panel.

    Invoice #21331723

    Previous request:
    https://lowendtalk.com/discussion/comment/4920481/#Comment_4920481

    Could you please check whether my VPS is eligible and help process the upgrade?

    Thank you!

    Hi @what99 -- Your bandwidth has been doubled, just as you requested.

    Really appreciate you giving RackNerd a shot. :)

  • @op23 said:

    @DaDeveloper said:

    @op23 said:

    @TrK said:

    @op23 said:

    @TrK said:

    @Neat321 said:

    @DaDeveloper said:

    @Neat321 said:

    @DaDeveloper said:

    @TrK said:
    Who uses a dgx spark?

    I use two lol

    Deepseek flash 4.1 ?

    1017 cuz bad internet lol

    Yeah

    I like that model. I also recommend trying hy 4 flash.

    Are they on dgx spark?

    He has 2 sparks, they can easily accommodate GGUF Q4_K_M quantized hy4 flash.

    Hmm sounds good enough but Q4 won't be slow?

    Avg 20 token/sec for a highly capable model and decent quant, it will be enough for a ton of use cases.

    I am planning to use my KS-LE-6 to run qwen flash next at 3 tok/second lol. Not doing interactive coding, but I can leave it overnight to reverse engineer code bases and setup recipes. 20 tok/second would be like rocketing to moon.

    It would feel slower compared to the 1017 I’m running tho lol

    Guess it depends if it’s any better at a specific task, just because a new model comes out it’s not necessarily worth running for sure.

    1017 is hard to beat right now it seems.

    For the speed and competitiveness yea it’s something I could live with for a good while longer

    Thanked by 1op23
  • dustincdustinc Patron Provider, Top Host, Veteran

    @guangm777 said:
    Hello, I would like to double the bandwidth.
    Order #3467159651
    Invoice #27072154

    thanks.

    Hi @guangm777 -- Your bandwidth has been successfully upgraded. 👊

    Your business is valuable to us at RackNerd. We look forward to continuing to serve you!

Sign In or Register to comment.