Skip to content
InfratoApp
InfratoApp / Compare / H200-NVL-141GB vs H200-SXM-141GB

H200-NVL-141GB vs H200-SXM-141GB

H200-SXM-141GB is the cheaper of the two: $2.09/hr at Beam, against $3.62/hr for H200-NVL-141GB at Massed Compute — 1.7× the price.

Cheapest on each

H200-NVL-141GB
4 SELLERSView H200-NVL-141GB →
$3.62/hr
cheapest, at Massed Compute
H200-SXM-141GB
26 SELLERSView H200-SXM-141GB →
$2.09/hr
cheapest, at Beam

Aligned facts

Same sheet, same order, both cards — including the gaps.

AttributeH200-NVL-141GBH200-SXM-141GB
VRAM141 GB141 GB
Bandwidth4.8 TB/s4.8 TB/s
FP16835 TFLOPS989 TFLOPS▲
FP81670 TFLOPS1979 TFLOPS▲
TDP600 W700 W▲
NVLink2- or 4-way NVIDIA NVLink bridge: 900GB/s per GPU PCIe Gen5: 128GB/sNVLink 4▲
ArchitectureHopperHopper
Released20242024

A blank is not a zero. Where one card states a figure and the other states none, neither wins the row — the second number is unknown, and ranking an unknown as nothing is how a comparison starts to lie.

Every seller, both cards

One row per seller, cheapest tier, on-demand list prices only. Sellers of both cards lead, cheapest combined rate first.

SellerH200-NVL-141GBH200-SXM-141GB
Lium
Vast.ai
RunPod
Massed Compute○ no answer
Beam○ no answer
Shadeform○ no answer
Theta EdgeCloud○ no answer
Sesterce○ no answer
Koyeb○ no answer
Civo○ no answer
DeepInfra○ no answer
Hyperbolic○ no answer
Oblivus○ no answer
Hyperstack○ no answer
Canopy Wave○ no answer
Cerebrium○ no answer
DigitalOcean○ no answer
Modal○ no answer
Runcrate○ no answer
Spheron○ no answer
Replicate○ no answer
Enverge○ no answer
Crusoe○ no answer
Fal.ai○ no answer
Verda (DataCrunch)○ no answer
Lyceum○ no answer
CoreWeave○ no answer

What this page will not tell you

On-demand list rates only · spot and committed terms priced separately and not mixed in here · no performance-per-dollar figure appears on this page, because it needs a throughput benchmark we do not hold and dividing a price by a vendor’s own TFLOPS claim produces a number that looks authoritative and means nothing