H100 vs P4

Jump to: Specs Pricing Offers Similar GPUs

Overview

Nvidia H100

The Nvidia H100 is a data center GPU on the Hopper architecture, released Q3 2022, with 80 GB HBM3 and 3,350 GB/s memory bandwidth on the SXM variant.

High Performance Datacenter

Nvidia P4

The Nvidia P4 is a data center GPU on the Pascal architecture, released Q3 2016, with 8 GB GDDR5 and 192 GB/s memory bandwidth.

Budget Datacenter

Memory per GPU

10x more on H100

H100

80 GB HBM3 94 GB on NVL

8 GB GDDR5

Memory bandwidth

17x higher on H100

H100

3,350 GB/s 2,000 GB/s on PCIe 3,938 GB/s on NVL

192 GB/s

GPU interconnect

H100

NVLink 900 GB/s NVLink 600 GB/s on PCIe, NVL

PCIe only

Peak performance (INT8)

90x higher on H100

H100

1,979 TOPS

22 TOPS

Vendor peak figures, dense and per GPU. Not measured throughput.

INT8 per watt

9.6x higher on H100

H100

2.83 TOPS/W

0.29 TOPS/W

Board power

9.3x higher on H100

H100

700 W

75 W

Precision support

H100

FP8 FP16 BF16 TF32 FP32 FP64 INT8

FP16 FP32 INT8

Highlighted formats are supported by only one of the two.

Architecture

H100

Hopper

Pascal

Released

6 years newer on H100

H100

Q3 2022

Q3 2016

Figures are for each model's headline configuration. Where a form factor carries a different number, it is named under the figure.

Source: H100 datasheet and P4 datasheet.

Salad

Access 60,000+ daily active GPUs across 191 countries, from $0.02/hour

Run RTX 50-series and 40-series cards without datacenter pricing

Pick a priority level to balance cost against interruptions

Per-second billing, with no charge for cold starts or image pulls

View Pricing

Price comparison

Cheapest on-demand

2.4x higher on H100

H100

$1.55 /GPU/hr

$0.65 /GPU/hr

Typical rate

4.2x higher on H100

H100

$3.39 /GPU/hr

$0.81 /GPU/hr

Median across providers, one vote each, so a large catalog does not outweigh a small one.

Spot floor

2.9x higher on H100

H100

$0.35 /GPU/hr

$0.12 /GPU/hr

Providers listing it

52 more on H100

H100

Price, last 90 days

H100

+10%

Too few providers pricing it

Change in the market-wide median, which moves when providers reprice and when cheaper or more expensive listings enter the market.

Heads up: A provider's own page may quote a different figure, for example on tax, region or a promotion. A monthly-only plan shows a derived hourly rate. Verify before provisioning. More on how we price.

Offers and availability

How this list works

Order

Each provider appears once per GPU. We pick the one configuration that speaks for that provider, and rank on it in this order:

Availability: in stock, then waitlist, then not reported, then out of stock. A price you can act on outranks a cheaper one you cannot.
Billing type: on-demand first, then reserved, then spot, then quote-only. On-demand leads because it is the only rate you can hold without committing to anything.
Price: the lowest per-GPU hourly rate the provider lists, converted to USD.

The configuration we pick is chosen the same way, availability before billing before price. The rest of that provider's configurations are on the GPU's own page.

Ties sort alphabetically. Reserved rates are not separated by term length, so a three-year commitment can sit above a one-month one: the label beneath each price says which you are looking at. Providers who publish no rate are listed last within their group.

Transparency and funding

Ads and sponsors: Paid placements sit at the top of the list and are always labeled as sponsored content. Sponsorship never influences the organic ranking itself.
Affiliates: Affiliate links are marked. We may earn a commission if you click them, but commissions never affect the order.
Prices: Shown in USD, converted at daily reference rates where a provider publishes in another currency. A month means 720 hours. How we estimate costs.