Machine learning server requirements
Machine learning server requirements for training and inference: compare CPU, GPU, RAM, storage, provider fit, and operational risks.
All sampled tariffs for this provider, grouped by vertical and normalized for comparison.
| Plan | Vertical | CPU | RAM | Disk type | GPU | Location | SLA | Price | Action |
|---|---|---|---|---|---|---|---|---|---|
vc2-1c-0.5gb-v6 vps · SSD | Servers | 1 | 0.5 GB | 10 GB | No GPU | EWR, Atlanta | 100% | $2.50/mo | Go to provider |
vc2-1c-1gb vps · SSD | Servers | 1 | 1 GB | 25 GB | No GPU | EWR, ORD +6 | 100% | $5/mo | Go to provider |
vc2-2c-4gb vps · SSD | Servers | 2 | 4 GB | 80 GB | No GPU | EWR, ORD +6 | 100% | $20/mo | Go to provider |
vc2-8c-32gb vps · SSD | Servers | 8 | 32 GB | 640 GB | No GPU | EWR, ORD +6 | 100% | $160/mo | Go to provider |
vhf-1c-1gb vps · HIGHFREQUENCY | Servers | 1 | 1 GB | 32 GB | No GPU | EWR, ORD +6 | 100% | $6/mo | Go to provider |
vhf-4c-16gb vps · HIGHFREQUENCY | Servers | 4 | 16 GB | 384 GB | No GPU | EWR, SEA +6 | 100% | $96/mo | Go to provider |
vhp-1c-1gb-amd vps · AMDHIGHPERF | Servers | 1 | 1 GB | 25 GB | No GPU | EWR, ORD +6 | 100% | $6/mo | Go to provider |
vhp-4c-8gb-amd vps · AMDHIGHPERF | Servers | 4 | 8 GB | 180 GB | No GPU | EWR, ORD +6 | 100% | $48/mo | Go to provider |
vhp-8c-16gb-amd vps · AMDHIGHPERF | Servers | 8 | 16 GB | 350 GB | No GPU | EWR, ORD +6 | 100% | $96/mo | Go to provider |
vbm-4c-32gb dedicated · SSD | Servers | 4 | 32 GB | 480 GB | No GPU | EWR, LAX +4 | 100% | $120/mo | Go to provider |
vbm-6c-32gb dedicated · SSD | Servers | 6 | 32 GB | 1,920 GB | No GPU | EWR, ORD +6 | 100% | $185/mo | Go to provider |
vbm-8c-132gb dedicated · NVME | Servers | 8 | 128 GB | 3,840 GB | No GPU | EWR, SEA +6 | 100% | $350/mo | Go to provider |
vbm-24c-384gb-amd dedicated · SSD | Servers | 24 | 384 GB | 960 GB | No GPU | EWR, ORD +6 | 100% | $825/mo | Go to provider |
vbm-6c-32gb-amd dedicated · NVME | Servers | 6 | 32 GB | 1,920 GB | No GPU | EWR, ORD +6 | 100% | $295/mo | Go to provider |
vbm-8c-64gb-amd dedicated · NVME | Servers | 8 | 64 GB | 3,890 GB | No GPU | ORD, DFW +5 | 100% | $365/mo | Go to provider |
vbm-72c-480gb-gh200-gpu training · NVME | GPU Servers | 72 | 480 GB | 960 GB | 1 | Atlanta, Frankfurt | 100% | $2,009.28/mo | Go to provider |
vbm-112c-2048gb-8-a100-gpu training · NVME | GPU Servers | 112 | 2,048 GB | 3,840 GB | 8 | EWR, Frankfurt | 100% | $15,052.80/mo | Go to provider |
vbm-112c-2048gb-8-h100-gpu training · NVME | GPU Servers | 112 | 2,015.35 GB | 1,920 GB | 8 | Amsterdam, Atlanta +31 | 100% | $16,074.24/mo | Go to provider |
vbm-256c-3072gb-8-mi325x-gpu training · NVME | GPU Servers | 128 | 3,072 GB | 28,608 GB | 8 | Amsterdam, Atlanta +31 | 100% | $24,810.24/mo | Go to provider |
vbm-256c-3072gb-8-b200-gpu training · NVME | GPU Servers | 128 | 3,072 GB | 28,608 GB | 8 | Amsterdam, Atlanta +31 | 100% | $45,696/mo | Go to provider |
Object Storage 1 TB included s3 | Cloud | — | — | — | No GPU | Amsterdam, Atlanta +31 | 100% | $5.99/mo | Go to provider |
Block Storage NVMe 10 GB storage · NVME | Cloud | — | — | — | No GPU | Amsterdam, Atlanta +31 | 100% | —/mo | Go to provider |
Managed Database HP 1 vCPU / 1 GB databases | Cloud | 1 | 1 GB | — | No GPU | EWR | 100% | $14.96/mo | Go to provider |
Load Balancer managed | Cloud | — | — | — | No GPU | Amsterdam, Atlanta +31 | 100% | $10.80/mo | Go to provider |
Load Balancer managed | WordPress Hosting | — | — | — | No GPU | Amsterdam, Atlanta +31 | 100% | $10.80/mo | Go to provider |
Serverless Inference input tokens inference | Cloud | — | — | — | No GPU | Amsterdam, Atlanta +31 | 100% | —/mo | Go to provider |
Affiliate link: we may earn a commission if you buy after clicking. Rankings remain editorially independent. Disclosure
Machine learning server requirements for training and inference: compare CPU, GPU, RAM, storage, provider fit, and operational risks.
What is an IP address? Learn how IPv4, IPv6, public, private, static, and shared addresses affect hosting, DNS, security, and server choice.
Slow DNS lookup fixes for hosting buyers: isolate resolver delay, clean bad records, tune TTLs, and use CDN prefetch without hiding risk.
VSCode Remote SSH on a VPS lets you code against a real Linux server; compare starter plans, setup steps, and operational risks.
Cloud storage free plans compared by real limits, object storage prices, egress traps, and lock-in checks before you upload files.
HostScout uses cookies and analytics to measure visits, improve comparisons, and track outbound provider clicks. Privacy details