expired
HP Omen 45L NVIDIA® GeForce RTX 5090, 64GB RAM, 2TB SSD, $4987.49 (original $6999 lol)
$4,987
HP Small & Medium Business
This Deal Has Expired We can notify you when similar deals are posted. Set up Deal Alert
HP Small & Medium BusinessDeal Details
NVM. lol
HP Omen 45L NVIDIA® GeForce RTX 5090, 64GB RAM, 2TB SSD, $4987.49 (original $6999 lol)
Stacking with EPP -$350
Coupon code: PARTYUSA25 for 25%
START here
https://www.hp.com/us-en/shop/pdp...91wjaa-aba
This is not customizable, it goes straight to cart
Intel® Core™ Ultra 9 285K (up to 5.7 GHz with Intel® Turbo Boost Technology, 36 MB L3 cache, 24 cores, 24 threads)[6,7]. This is far better than the usual AMD 9700X.
64GB of RAM
1TB of storage.
Final price: $4,987.49
HP Omen 45L NVIDIA® GeForce RTX 5090, 64GB RAM, 2TB SSD, $4987.49 (original $6999 lol)
Stacking with EPP -$350
Coupon code: PARTYUSA25 for 25%
START here
https://www.hp.com/us-en/shop/pdp...91wjaa-aba
This is not customizable, it goes straight to cart
Intel® Core™ Ultra 9 285K (up to 5.7 GHz with Intel® Turbo Boost Technology, 36 MB L3 cache, 24 cores, 24 threads)[6,7]. This is far better than the usual AMD 9700X.
64GB of RAM
1TB of storage.
Final price: $4,987.49
Community Notes
About the Poster
Deal Details
Community Notes
About the Poster
NVM. lol
HP Omen 45L NVIDIA® GeForce RTX 5090, 64GB RAM, 2TB SSD, $4987.49 (original $6999 lol)
Stacking with EPP -$350
Coupon code: PARTYUSA25 for 25%
START here
https://www.hp.com/us-en/shop/pdp...91wjaa-aba
This is not customizable, it goes straight to cart
Intel® Core™ Ultra 9 285K (up to 5.7 GHz with Intel® Turbo Boost Technology, 36 MB L3 cache, 24 cores, 24 threads)[6,7]. This is far better than the usual AMD 9700X.
64GB of RAM
1TB of storage.
Final price: $4,987.49
HP Omen 45L NVIDIA® GeForce RTX 5090, 64GB RAM, 2TB SSD, $4987.49 (original $6999 lol)
Stacking with EPP -$350
Coupon code: PARTYUSA25 for 25%
START here
https://www.hp.com/us-en/shop/pdp...91wjaa-aba
This is not customizable, it goes straight to cart
Intel® Core™ Ultra 9 285K (up to 5.7 GHz with Intel® Turbo Boost Technology, 36 MB L3 cache, 24 cores, 24 threads)[6,7]. This is far better than the usual AMD 9700X.
64GB of RAM
1TB of storage.
Final price: $4,987.49
Community Voting
Deal
Score
-22
Good Deal
Bad Deal
Join The Conversation
Share your experience with the Slickdeals community
Share information with the community. Please follow our Community Guidelines and be kind!
13 Comments
Sign up for a Slickdeals account to remove this ad.
"at least its not an HP" is the new way to quickly evaluate noname brand computers ....
Sign up for a Slickdeals account to remove this ad.
Top secret use, don't want online models to see your data. sure.
Batch oriented processing can work. Offline agent can work as long as you are not waiting for the response - ie: you msg it and wait it to come back.
Anytime you want to have more than 1 concurrent request and more than say 50 tgs average, I wouldn't use local. Long think time requests? oh god no, just takes up more time and frustrates me.
And if you don't have large enough VRAM (ie: just 16GB, you are limited to smaller models which are not smart)
This means coding project that needs smart enough LLM, online is still my go to ---- concurrency and subagents working at the same time, online is the only play. Waiting for your LLM while you code is time sink. And you can't loop it enough for me to want to waste time to just set it and come back in a day hoping it did enough.
Image and video processing will always be better with nvidia. This is best ROI if you are doing video generation as that is the highest cost item with online model providers.
I don't run local models on my M5 Max 128GB, super large models are not that useful to me either and 27B was the smartest for a while but it's slow as a dog on Mac (like 25 tgs) and I can't run those 500B quantized at 128GB so all that ram is pretty useless.
I do have smaller local models, mainly for random prompt generation but I also wrote code to just go online since there are enough free subs so I wrote routing code to just go to free online ones for these random use.
With advent of diffusiongemma - it's faster 200-250% for me and but there were some weird bug so I will wait for other models to catch up with better tgs techniques like dflash.
I don't need to train local models or have top secret use to worry about online model look at my API requests or data I am passing it. Some people may and they would need GPU cycles. Not my use case
Did a test the other day, clone a specific website with an online model Minimax-M3-highspeed and it ripped thru 100M+ tokens in about 6 hours. It would have taken much longer on local AI. Wasn't 100% functional but the the UX was done and had some bugs in the actual APIs but in the shell was there.
Couldn't do this with local AI with just 32GB VRAM with this kind of throughput. Now I can tweak the code and fix whatever it didn't do. Qweb 27B would have been too slow even on a 5090 (50 tgs) before think time and only one agent at time. You can get higher with nvfp4.
Join The Conversation
Share your experience with the Slickdeals community
Share information with the community. Please follow our Community Guidelines and be kind!