GPU model · price + stock · hourly snapshot

Nvidia A800 cloud rental prices

Name: Nvidia A800 cloud GPU rental
Brand: Nvidia
Availability: OutOfStock

1 provider lists the Nvidia A800, 0 have it in stock. It rents for $1.08 per GPU hour.

Ampere 80 GB VRAM Q4 2022 8 configurations

LIVE Prices and stock checked 1 minute ago November 4, 2026 UTC

Cheapest listed

$1.08

Compshare · no stock reported

Median

$1.08

across 1 provider

Most expensive

$1.08

the only price listed

Providers

Every provider renting the Nvidia A800

All providers →

Billing:

On-demandfrom $1.08/hrno stock reportedPay as you go, stop any time.Spotfrom $0.45/hrno stock reportedCheaper, but the provider can take the machine back.Reservednone listedPrepay a term for the lowest hourly rate.

Listed, but nobody reports stock at this price type right now.

Providers renting the Nvidia A800, in stock first, then cheapest first
Provider	Country	Configs	Scale	From	Stock	Days in stock	Grab
Compshare	China	8	1x-8x	$1.08	Not reported	0/4	Grab

In stock first, then cheapest. Stock is what the provider reported at our last check. One GPU for a month (730h) at $1.08: $786.

Price history

Price trend

Cheapest Nvidia A800 rate has held at $1.08/GPU/hr since Nov 1.

$1.08 /GPU/hr · Nov 4

One point per day: the lowest listed $/GPU/hr across every provider we track, captured daily. 76 offers in the latest snapshot.

Configurations

All 8 configurations

In stock only8 of 8 configurations

Nvidia A800 configurations from every provider
Configuration	Provider	Region	VRAM	vCPU	RAM	Per GPU/hr	Total/hr	Stock
8x A800PreemptibleCompshare	Compshare	—	640 GB	124	1800 GB	$0.45	$3.59	Not reported
4x A800PreemptibleCompshare	Compshare	—	320 GB	64	940 GB	$0.45	$1.81	Not reported
1x A800PreemptibleCompshare	Compshare	—	80 GB	16	240 GB	$0.45	$0.45	Not reported
2x A800PreemptibleCompshare	Compshare	—	160 GB	32	480 GB	$0.45	$0.91	Not reported
8x A800Compshare	Compshare	—	640 GB	124	1800 GB	$1.08	$8.61	Not reported
4x A800Compshare	Compshare	—	320 GB	64	940 GB	$1.09	$4.37	Not reported
1x A800Compshare	Compshare	—	80 GB	16	240 GB	$1.10	$1.10	Not reported
2x A800Compshare	Compshare	—	160 GB	32	480 GB	$1.10	$2.20	Not reported

Model fit

What fits in 80 GB

The question behind most rentals: will the model load. Sized against one Nvidia A800.

VRAM required to serve open-weight LLMs on a Nvidia A800, by quantization precision
Model	16-bit	8-bit	4-bit
DeepSeek-V3 DeepSeek · 671B params · 37B active	1,610 GB 21× · 70 GB free	805 GB 11× · 75 GB free	403 GB 6× · 77 GB free
Llama 3.1 405B Meta · 405B params	972 GB 13× · 68 GB free	486 GB 7× · 74 GB free	243 GB 4× · 77 GB free
Qwen3 235B-A22B Alibaba · 235B params · 22B active	564 GB 8× · 76 GB free	282 GB 4× · 38 GB free	141 GB 2× · 19 GB free
Mixtral 8x22B Mistral · 141B params · 39B active	338 GB 5× · 62 GB free	169 GB 3× · 71 GB free	85 GB 2× · 75 GB free
Mistral Large 2 Mistral · 123B params	295 GB 4× · 25 GB free	148 GB 2× · 12 GB free	74 GB 1× · 6 GB free
gpt-oss-120b OpenAI · 117B params · 5.1B active	281 GB 4× · 39 GB free	140 GB 2× · 20 GB free	70 GB 1× · 10 GB free
Qwen2.5 72B Alibaba · 72.7B params	174 GB 3× · 66 GB free	87 GB 2× · 73 GB free	44 GB 1× · 36 GB free
Llama 3.3 70B Meta · 70.6B params	169 GB 3× · 71 GB free	85 GB 2× · 75 GB free	42 GB 1× · 38 GB free
Qwen3 32B Alibaba · 32.8B params	79 GB 1× · 1 GB free	39 GB 1× · 41 GB free	20 GB 1× · 60 GB free
Gemma 2 27B Google · 27.2B params	65 GB 1× · 15 GB free	33 GB 1× · 47 GB free	16 GB 1× · 64 GB free
Mistral Small 3 Mistral · 24B params	58 GB 1× · 22 GB free	29 GB 1× · 51 GB free	14 GB 1× · 66 GB free
Llama 3.1 8B Meta · 8B params	19 GB 1× · 61 GB free	10 GB 1× · 70 GB free	5 GB 1× · 75 GB free

Required memory is weights plus 20% for KV cache and runtime at short context; long contexts and large batches need more. Cells tinted green fit on a single Nvidia A800.

Specifications

Nvidia A800 specifications

Nvidia A800 hardware specifications
Architecture	Ampere
Memory per GPU	80 GB HBM2e (40 GB PCIe board also exists)
Memory bandwidth	1,935 GB/s
Release date	Q4 2022
GPU interconnect	NVLink 400 GB/s
FP16 / BF16 compute rate	312 TFLOPS (dense), 624 with sparsity
INT8 compute rate	623.5 TOPS (dense), 1,247 with sparsity
Process node	TSMC N7
Board power	240 W
Partitioning	MIG, up to 7 instances

Alternatives

Every provider renting the Nvidia A800

Price trend

All 8 configurations

What fits in 80 GB

Nvidia A800 specifications

Alternatives to the Nvidia A800