KONST
AI Training CalculatorContact us
EN
AIDC site | North America · United States

NVIDIA B300・United States California

US-California-B300ClusterLive by mid-December 2026

Location
United States California
Availability
Live by mid-December 2026
Cooling
Air Cooling
Reference price
$6.10 USD / hr / GPU, starting price · Monthly
64Nodes
512GPU
288GB HBM3e per GPUMemory per GPUabout 147.5TB total
800G RoCEGPU Fabric
Specifications

Full specifications

GPU, host, network, and site conditions in one place.

COMPUTECompute

GPU model
NVIDIA B300
GPU configuration
512 units64 nodes × 8x B300
Deployment scale
64 nodes512 GPU
Memory per GPU
288GB HBM3e per GPU
Total cluster memory
about 147.5TB total

HOSTHost

Processor
2x Intel Xeon 6776P
System memory
2,048GB DDR5-6400 per node
Storage
OS: 2x 1.92TB NVMeData: 4x 7.68TB NVMe (30.72TB per node)

NETWORKNetwork

GPU interconnect
RoCE6,400Gbps per node (8x 800Gbps)NVIDIA ConnectX-8 SuperNIC

FACILITYSite and delivery

Location
United States California
Cooling
Air Cooling
Availability
Live by mid-December 2026

WORKLOADSuitable workloads

Purpose
AI TrainingInferenceHPC

Remaining specifications are provided when you request a quote. Source: GPU cluster specification sheet · 2026-09-30

Delivery schedule

From signing to go-live

  1. 01Signing3 year term
  2. 02HandoverConfirmed with quote
  3. 03Go-liveLive by mid-December 2026

Prices, specifications, and availability are confirmed in the signed contract and do not reflect live inventory.

Location

United States California

California
Region
North America
Country / city
United States California
Site code
US-California-B300
Sites in the same location
US-California-B300-32

Location is indicative, at country or city level.

Compare | Other sites

Sites with NVIDIA B300

Compare scale, network, cooling, and delivery time side by side.

SiteScaleGPU memoryGPU FabricCoolingAvailabilityStarting price, USD / hr / GPU
US-California-B300United States CaliforniaThis page64 nodes
512 GPU
288GB per GPU
about 147.5TB total
800G RoCEAir CoolingLive by mid-December 2026$6.10 from
AU-Sydney-B300Australia Sydney64 nodes
512 GPU
288GB per GPU
about 147.5TB total
CX8 ×8 embeddedTo be confirmedLive in November 2026$6.80 from
US-California-B300-32United States California32 nodes
256 GPU
288GB per GPU
about 73.7TB total
800G RoCEAir CoolingLive by mid-December 2026$6.10 from
CA-Toronto-B300Canada Toronto64 nodes
512 GPU
288GB per GPU
about 147.5TB total
800G InfiniBandLiquid coolingDelivery 8 to 12 weeks after order$6.20 from
CA-Toronto-B300-XLCanada Toronto140 nodes
1,120 GPU
288GB per GPU
about 322.6TB total
800G InfiniBandLiquid coolingDelivery 8 to 12 weeks after order$6.20 from
PL-B300Poland128 nodes
1,024 GPU
288GB per GPU
about 294.9TB total
800G InfiniBandAir CoolingPhased go-live from January 2027
All units in place in February
$6.00 from
MY-Cyberjaya-B300Malaysia Cyberjaya128 production nodes
1,024 GPU
288GB per GPU
about 294.9TB total
800G InfiniBandDirect-to-chip liquid coolingLive in 2027 Q1$6.00 from
ES-B300Spain128 nodes
1,024 GPU
288GB per GPU
about 295TB total
800G InfiniBandAir Cooling16 to 18 weeks after signing$6.10 from
EU-B300Europe144 nodes
1,152 GPU
288GB per GPU
about 332TB total
800G InfiniBandAir Cooling16 to 18 weeks after signing$6.10 from
FR-Paris-B300France Paris50 nodes
400 GPU
288GB per GPU
about 115TB total
800G InfiniBandAir Cooling14 to 18 weeks after signing$6.30 from
US-Houston-B300United States Houston64 nodes
512 GPU
288GB per GPU
147.5TB total
800G RoCEAir CoolingDelivery in February 2027$6.30 from
CZ-B300Czechia64 nodes
512 GPU
288GB per GPU
about 147.5TB total
800GbpsDLC liquid coolingLive March to June 2027$6.50 from
TW-B300Taiwan New Taipei4 nodes
32 GPU
288GB per GPU
about 9.2TB total
800G InfiniBandAir CoolingPre-sale
Schedule to be determined
$6.00 from
FAQ | Common questions

Common questions

How are the price and delivery schedule confirmed?
Prices and delivery schedules are for evaluation reference. Actual specifications, availability, and commercial terms are set by the signed contract. Prices, specifications, and availability are confirmed in the signed contract and do not reflect live inventory.
How does bare metal differ from GPU cloud?
Bare metal is a physical machine with no virtualization layer, so all performance is available, but you manage the operating system and cluster yourself. GPU cloud comes with the environment prepared and activates quickly, but resources are shared with other tenants. Training at long-term full load usually goes to bare metal.
Besides overseas GPU clouds, what should be considered when renting locally?
Three things: data residency, network latency, and service hours. When data must not leave the country, equipment location is a hard requirement. Local latency is lower than transoceanic links. The local team keeps the same working hours as the lessee, so fault resolution does not wait on time zones.
Can the GPU count be increased or reduced mid-term?
Adding GPUs usually means adding nodes, with pricing folded into the original contract and subject to capacity at the time. Removing GPUs is restricted by take-or-pay terms, so in practice the scale is usually adjusted at renewal.
Who owns the equipment, and what does the rental fee cover?
KONST keeps ownership of the equipment. The renter pays for the right to use it, which covers data center space, power, network, and operations. KONST bears the risks tied to purchase funding, residual equipment value, and generation changes.
From installation to ready for use, where does KONST's delivery scope end?
From the initial operating system installation, KONST sets up the machines, drivers, and base environment, and configures a separate network segment and a dedicated VPN. After that, the renter manages frameworks, scheduling, and data storage on its own.

Want to get B300 compute quickly on the US West Coast?

Tell us how many GPUs you need and the lease term, and we will reply with the delivery time and a quote.