Við tökum við rafmynt — svo myntirnar sem þú átt geta keypt alvöru vélbúnað.Greiddu í rafmynt, eyddu henni í alvöru vélbúnað. Frí, tryggð sending um allan heim fyrir pantanir yfir 399 $. Hlutlausir pappakassar án merkinga, sendir frá Þýskalandi.Frí, tryggð sending yfir 399 $. Gengið þitt er læst í 30 mínútur um leið og greiðsluferlið hefst.Gengið læst í 30 mínútur. Hver vara er innsigluð, raðnúmer hennar yfirfarið, og hún nýtur 24 mánaða ábyrgðar.Innsiglað, með 24 mánaða ábyrgð.

Vaktaðar vörur Skrá inn
Black CoolerMaster HAF 500 desktop with tempered glass side panel and ARGB fans

Adamant Custom

Adamant Custom Core Ultra 7 265K 192GB 2TB RTX 5070 gaming PC

SKU CH-0GTS-K47W2

A 20-core Intel Core Ultra 7 265K desktop with 192GB DDR5, 2TB NVMe storage and an RTX 5070 GPU, liquid-cooled in a tempered-glass case

  • Release Year2026
  • TypeBusiness Desktops & Workstations
  • Form FactorDesktop
  • AI FeaturesAI Ready
  • UsageWorkstation
  • ColorBlack
  • 20-core Intel Core Ultra 7 265K reaches 5.50 GHz for heavy workloads
  • 192GB DDR5 memory handles large datasets and multitasking
  • 2TB Samsung 990 PRO NVMe SSD delivers 7450 MB/s read speed
  • RTX 5070 12GB GPU accelerates rendering and AI tasks
  • 240mm liquid cooler keeps CPU temperatures low under load

20-Core Workstation Desktop

This Adamant Custom desktop pairs a 20-core Intel Core Ultra 7 265K processor with 192GB DDR5 memory and a 2TB NVMe SSD. The GeForce RTX 5070 12GB GPU handles rendering, modelling and AI learning workloads. Windows 11 Pro is pre-installed on the Z890 TUF motherboard platform.

Pre-Built Liquid Cooling

The system arrives fully assembled with a 240mm AIO liquid cooler mounted on the CPU. The CoolerMaster HAF 500 case includes front 200mm ARGB fans, a rear 120mm ARGB fan and a dedicated 120mm GPU fan. Internal bays provide two 3.5-inch and two 2.5-inch drive slots for future expansion.

Suits Heavy Multitasking Builders

Buyers needing 192GB memory and 20 threads for parallel workloads will benefit from this configuration. The single RTX 5070 limits multi-GPU compute scaling. Users requiring more than two M.2 drives or a higher-wattage power supply should consider a different chassis.

Helstu kostir

  • 20-core Intel Core Ultra 7 265K reaches 5.50 GHz for heavy workloads
  • 192GB DDR5 memory handles large datasets and multitasking
  • 2TB Samsung 990 PRO NVMe SSD delivers 7450 MB/s read speed
  • RTX 5070 12GB GPU accelerates rendering and AI tasks
  • 240mm liquid cooler keeps CPU temperatures low under load

Tæknilýsing

Release Year2026
TypeBusiness Desktops & Workstations
Form FactorDesktop
AI FeaturesAI Ready
UsageWorkstation
ColorBlack
Cooling SystemLiquid Cooling
Operating SystemWindows 11 Pro
Windows 11Ready
CPU TypeIntel Core Ultra 7
CPU NameIntel Core Ultra 7 265K
CPU Socket TypeLGA 1700
GPU/VGA TypeGeForce RTX 5070
Memory Capacity192GB DDR5
Memory Slots (Available/Total)4 total
Maximum Memory Supported192 GB
SSD2TB NVMe
HDDNo
Optical Drive TypeNo
Screen SizeNo Screen
Audio ChipsetIntegrated
Audio Channels7.1 channels
LAN ChipsetIntegrated
LAN Speed2.5Gbps
WLAN802.11be Wireless LAN
WiFi GenerationWi-Fi 7
BluetoothBluetooth 5.4
RJ451 port

Spurningar um þessa vöru

How can I confirm the liquid CPU cooler is running after I power on the system?

The 240mm AIO Quiet Edition water cooler uses a 140mm fan; you should hear low fan noise and see coolant flow through the transparent tubing, while BIOS or Windows reports CPU temperatures near the 3.7GHz base clock under idle load.

Which components are included in the box and which must be purchased separately?

The box contains the CoolerMaster HAF 500 case with pre-installed 200mm and 120mm ARGB fans, Z890 TUF motherboard, Core Ultra 7 265K CPU, 192GB DDR5 memory, 2TB Samsung 990 PRO NVMe SSD, RTX 5070 12GB graphics card, 750W 80 PLUS Bronze PSU, and Windows 11 Pro; a monitor, keyboard and mouse are not included.

Should I choose the 192GB DDR5 configuration or a lower capacity for AI modelling workloads?

The motherboard supports a maximum of 192GB across four slots and this build already populates that limit, giving full bandwidth for large-model training; a smaller kit would leave slots empty and reduce total memory available for AI learning tasks.

Einnig í þessum vöruflokki

Fólk bar þetta saman

Sama hilla, sama greiðsluferli — átta myntir og 30 mínútna gengislæsing.