GPU Cloud
GPU buluti: soatbay to'lov bilan NVIDIA GPU serverlar ijarasi
LLM xulosalash (inference), fine-tuning va tasvirlar yaratish uchun NVIDIA GPU serverlarini ijaraga oling. Jonli narxlar, soatbay to'lov, PyTorch, vLLM va ComfyUI shablonlari, SSH orqali kirish.
- Haqiqiy vaqtdagi mavjudlik, buyurtmadan oldin tasdiqlanadigan narx
- PyTorch, vLLM, Ollama va ComfyUI uchun tayyor shablonlar
- To'xtatilgan soatbay serverlarda faqat disk uchun haq olinadi
Cloudj GPU Cloud orqali siz jonli inventardan GPU serverini ijaraga olasiz: GPU modeli, VRAM va disk hajmini tanlang, PyTorch, vLLM, Ollama yoki ComfyUI kabi tayyor shablonlardan birini ishga tushiring va SSH orqali ulaning. Soatbay serverlar uchun toʻlov hisob balansingizdan yechiladi, toʻxtatilgan soatbay server uchun esa faqat disk toʻlovi olinadi.
GPU ijara narxlari va mavjudligi
Jonli inventar: mavjudlik va narxlar bozor holatiga qarab oʻzgaradi. Boshqaruv paneli buyurtma berishdan oldin yakuniy narxni tasdiqlaydi.
| GPU | Har bir GPU uchun VRAM | Arxitektura | FP32 unumdorligi | Quvvat sarfi (TDP) | Boshlang'ich | |
|---|---|---|---|---|---|---|
| NVIDIA RTX PRO 6000 Max-Q | 96 GB | Blackwell | ≈ 119 TFLOPS | 300 W | – | Batafsil → |
| NVIDIA RTX PRO 6000 WS | 96 GB | Blackwell | ≈ 119 TFLOPS | 600 W | – | Batafsil → |
| NVIDIA RTX 5090 | 32 GB | Blackwell | ≈ 107.6 TFLOPS | 575 W | – | Batafsil → |
| NVIDIA RTX 5090D | 32 GB | Blackwell | ≈ 107.6 TFLOPS | 575 W | – | Batafsil → |
| NVIDIA RTX PRO 6000 S | 96 GB | Blackwell | ≈ 93.6 TFLOPS | 600 W | – | Batafsil → |
| NVIDIA RTX 4090 | 24 GB | Ada Lovelace | ≈ 81.4 TFLOPS | 450 W | – | Batafsil → |
| NVIDIA RTX 6000Ada | 48 GB | Ada Lovelace | ≈ 81.4 TFLOPS | 300 W | – | Batafsil → |
| NVIDIA RTX 6000D | 84 GB | Blackwell | ≈ 77.6 TFLOPS | 600 W | – | Batafsil → |
| NVIDIA L40S datacenter | 45 GB | Ada Lovelace | ≈ 73.3 TFLOPS | 350 W | – | Batafsil → |
| NVIDIA RTX PRO 5000 | 48 GB | Blackwell | ≈ 69.6 TFLOPS | 300 W | – | Batafsil → |
| NVIDIA H100 SXM datacenter | 80 GB | Hopper | ≈ 53.5 TFLOPS | 700 W | – | Batafsil → |
| NVIDIA RTX 5080 | 16 GB | Blackwell | ≈ 53.4 TFLOPS | 360 W | – | Batafsil → |
| NVIDIA RTX 4080S | 16 GB | Ada Lovelace | ≈ 50.9 TFLOPS | 320 W | – | Batafsil → |
| NVIDIA RTX 4080 | 16 GB | Ada Lovelace | ≈ 48.6 TFLOPS | 285 W | – | Batafsil → |
| NVIDIA H100 NVL datacenter | 94 GB | Hopper | ≈ 48.3 TFLOPS | 400 W | – | Batafsil → |
| NVIDIA H200 NVL datacenter | 140 GB | Hopper | ≈ 48.3 TFLOPS | 600 W | – | Batafsil → |
| NVIDIA RTX 5070 Ti | 16 GB | Blackwell | ≈ 44.3 TFLOPS | 300 W | – | Batafsil → |
| NVIDIA RTX PRO 4000 | 24 GB | Blackwell | ≈ 44.3 TFLOPS | 145 W | – | Batafsil → |
| NVIDIA RTX 4070S Ti | 16 GB | Ada Lovelace | ≈ 42.6 TFLOPS | 285 W | – | Batafsil → |
| NVIDIA H100 PCIE datacenter | 80 GB | Hopper | ≈ 41 TFLOPS | 350 W | – | Batafsil → |
| NVIDIA RTX 4500Ada | 24 GB | Ada Lovelace | ≈ 38.2 TFLOPS | 210 W | – | Batafsil → |
| NVIDIA RTX 3090 Ti | 24 GB | Ampere | ≈ 36.4 TFLOPS | 450 W | – | Batafsil → |
| NVIDIA RTX A6000 | 48 GB | Ampere | ≈ 36.1 TFLOPS | 300 W | – | Batafsil → |
| NVIDIA RTX 4070S | 12 GB | Ada Lovelace | ≈ 35.8 TFLOPS | 220 W | – | Batafsil → |
| NVIDIA RTX 3090 | 24 GB | Ampere | ≈ 35.4 TFLOPS | 350 W | – | Batafsil → |
| NVIDIA RTX 3080 Ti | 12 GB | Ampere | ≈ 35.1 TFLOPS | 350 W | – | Batafsil → |
| NVIDIA A40 datacenter | 45 GB | Ampere | ≈ 29.9 TFLOPS | 300 W | – | Batafsil → |
| NVIDIA RTX 3080 | 10 GB | Ampere | ≈ 29.5 TFLOPS | 380 W | – | Batafsil → |
| NVIDIA RTX 4070 | 12 GB | Ada Lovelace | ≈ 29.3 TFLOPS | 200 W | – | Batafsil → |
| NVIDIA L4 datacenter | 22 GB | Ada Lovelace | ≈ 24.2 TFLOPS | 72 W | – | Batafsil → |
| NVIDIA RTX 5060 Ti | 16 GB | Blackwell | ≈ 22.8 TFLOPS | 180 W | – | Batafsil → |
| NVIDIA RTX 4060 Ti | 16 GB | Ada Lovelace | ≈ 21.6 TFLOPS | 165 W | – | Batafsil → |
| NVIDIA RTX A4000 | 16 GB | Ampere | ≈ 20.6 TFLOPS | 140 W | – | Batafsil → |
| NVIDIA RTX 3070 | 8 GB | Ampere | ≈ 19.8 TFLOPS | 220 W | – | Batafsil → |
| NVIDIA RTX 3060 Ti | 8 GB | Ampere | ≈ 15.7 TFLOPS | 200 W | – | Batafsil → |
| NVIDIA A100 PCIE datacenter | 40 GB | Ampere | ≈ 15.6 TFLOPS | 250 W | – | Batafsil → |
| NVIDIA A100 SXM4 datacenter | 40 GB | Ampere | ≈ 15.6 TFLOPS | 400 W | – | Batafsil → |
| NVIDIA A800 PCIE datacenter | 80 GB | Ampere | ≈ 15.6 TFLOPS | 300 W | – | Batafsil → |
| NVIDIA RTX 2080 Ti | 11 GB | Turing | ≈ 14.6 TFLOPS | 250 W | – | Batafsil → |
| NVIDIA RTX 3060 laptop | 12 GB | Ampere | ≈ 12.9 TFLOPS | 80 W | – | Batafsil → |
| NVIDIA Tesla V100 datacenter | 16 GB | Volta | ≈ 12.5 TFLOPS | 300 W | – | Batafsil → |
| NVIDIA CMP 170HX datacenter | 64 GB | Ampere | ≈ 12.1 TFLOPS | 250 W | – | Batafsil → |
| NVIDIA RTX 3060 | 12 GB | Ampere | ≈ 12 TFLOPS | 170 W | – | Batafsil → |
| NVIDIA Titan Xp | 12 GB | Pascal | ≈ 11.7 TFLOPS | 160 W | – | Batafsil → |
| NVIDIA RTX A2000 | 6 GB | Ampere | ≈ 11.2 TFLOPS | 70 W | – | Batafsil → |
| NVIDIA GTX 1080 Ti | 11 GB | Pascal | ≈ 11 TFLOPS | 250 W | – | Batafsil → |
| NVIDIA GTX 1080 | 8 GB | Pascal | ≈ 7.8 TFLOPS | 130 W | – | Batafsil → |
| NVIDIA GTX 1070 Ti | 8 GB | Pascal | ≈ 7.4 TFLOPS | 180 W | – | Batafsil → |
| NVIDIA GTX 1650 | 4 GB | ≈ undefined TFLOPS | W | – | Batafsil → | |
| NVIDIA Q RTX 8000 | 45 GB | Turing | ≈ 11.9 TFLOPS | 250 W | – | Batafsil → |
| NVIDIA Tesla P40 datacenter | 24 GB | Pascal | ≈ 9.4 TFLOPS | 250 W | – | Batafsil → |
| NVIDIA RTX 2070S | 8 GB | Turing | ≈ 8.8 TFLOPS | 210 W | – | Batafsil → |
| NVIDIA Q RTX 4000 | 8 GB | Turing | ≈ 7.7 TFLOPS | 125 W | – | Batafsil → |
| NVIDIA RTX 2070 | 8 GB | Turing | ≈ 7.7 TFLOPS | 150 W | – | Batafsil → |
| NVIDIA RTX 2060S | 8 GB | Turing | ≈ 7.5 TFLOPS | 175 W | – | Batafsil → |
| NVIDIA GTX TITAN X | 12 GB | Maxwell | ≈ 6.8 TFLOPS | 150 W | – | Batafsil → |
| NVIDIA RTX 2060 | 12 GB | Turing | ≈ 6.5 TFLOPS | 130 W | – | Batafsil → |
| NVIDIA GTX 1660 Ti | 6 GB | Turing | ≈ 5.3 TFLOPS | 120 W | – | Batafsil → |
| NVIDIA GTX 1660 S | 6 GB | Turing | ≈ 3.6 TFLOPS | 125 W | – | Batafsil → |
AI, LLM va rendering yuklamalari uchun GPU shablonlari
LLM inferensi
Katta til modellarini ishga tushirish va sinovdan oʻtkazish uchun Open WebUI bilan vLLM, SGLang yoki Ollama.
Fine-tuning
Til modellarini oʻz maʼlumotlaringiz asosida fine-tuning qilish uchun Unsloth Studio.
Rasm va video generatsiyasi
Flux, SDXL va video ish jarayonlari uchun ComfyUI.
PyTorch va CUDA da ishlab chiqish
SSH orqali ulanish mumkin bo‘lgan NVIDIA CUDA bilan PyTorch va Ubuntu 24.04 obrazlari.
Ish stoli yoki to‘liq VM
Brauzerda Ubuntu ish stoli yoki kerak bo‘lganda to‘liq KVM virtual mashinasi.
4 qadamda GPU serverni qanday ijaraga olish mumkin
- 01
Shablon va disk hajmini tanlang
- 02
Jonli inventardan mos keluvchi GPU-ni tanlang
- 03
Narxni tasdiqlang va hisob balansingizdan serverni ishga tushiring
- 04
SSH yoki brauzer terminali orqali ulaning
GPU buluti narxlari va hisob-kitob
- Soatbay: server ishlab turgan har bir soat uchun GPU narxi hisoblanadi. Toʻxtatilgan soatbay server uchun faqat disk toʻlovi olinadi.
- Toʻlov davrlari: oylik, choraklik, yarim yillik yoki yillik — har qanday chegirma toʻlovdan oldin koʻrsatiladi.
- Disk xotirasi GPU narxi bilan birga har bir GB uchun hisoblanadi.
- Narxlar USD da koʻrsatilgan; ularni joriy kurs boʻyicha VND da ham koʻrishingiz mumkin.
GPU buluti bo'yicha ko'p beriladigan savollar
GPU xotirasiga (VRAM) mos keladigan har qanday yuklamalarni: LLM xulosalash (inference) va fine-tuning, tasvir va video yaratish, 3D rendering hamda CUDA dasturlash. PyTorch, vLLM, SGLang, Ollama, Unsloth va ComfyUI shablonlari barcha dasturlarni siz uchun avtomatik o'rnatadi.
Yuqoridagi jadvalda ayni paytda mavjud bo'lgan konfiguratsiyalar, jumladan, bazada bor bo'lganda 32 GB VRAM'ga ega NVIDIA RTX 5090 ko'rsatilgan. Mavjudlik kun davomida o'zgarib turadi.
Narxlar real vaqtdagi GPU mavjudligiga qarab belgilanadi va o'zgarishi mumkin. Boshqaruv paneli buyurtma berishingizdan oldin tasdiqlangan narxni ko'rsatadi va serveringizga aynan shu tasdiqlangan narx bo'yicha hisob yoziladi.
Ha. Soatbay to'lovdan tashqari oylik, choraklik, yarim yillik yoki yillik to'lov davrlarini tanlashingiz mumkin, barcha chegirmalar to'lovdan oldin ko'rsatiladi. Disk xotirasi uchun to'lov har GB hisobida GPU narxiga qo'shib hisoblanadi.
Soatbay to'lovda to'xtatilgan server uchun faqat disk xotirasi hisoblanadi. Agar serverdagi ma'lumotlar ortiq kerak bo'lmasa, serverni shunchaki o'chirib tashlang.
Shaxsiy kalitingiz bilan SSH orqali yoki to'g'ridan-to'g'ri boshqaruv panelidagi brauzer terminalidan ulanishingiz mumkin. Desktop va VM shablonlari grafik interfeys (GUI) orqali ulanish imkoniyatini ham beradi.