g4dn.metal by Amazon Web Services

g4dn.metal is a Graphics intensive [Instance store volumes] [Network and EBS optimized] Gen4 metal server offered by Amazon Web Services with 96 vCPUs, 384 GiB of memory and 1.8 TB of storage. The pricing starts at 0.916 USD per hour.
96 vCPU
384 GiB Memory
1.8 TB Storage
8 GPU
Spare SCore
81,985
(All-cores)
1,459
(Single-core)

Specifications

Server Metadata

Vendor ID
aws
Name
g4dn.metal
Description
Graphics intensive [Instance store volumes] [Network and EBS optimized] Gen4 metal
Family
g4dn
Hw Virt
Status
active
Observed At
2026-09-15T14:22:29.878813

Availability

REGION / IDSPOTONDEMAND
Ohio (US) / us-east-2
4.142567 USD/h 7.824 USD/h
Northern Virgina (US) / us-east-1
5.83158 USD/h 7.824 USD/h
Oregon (US) / us-west-2
4.484567 USD/h 7.824 USD/h
Stockholm (SE) / eu-north-1
2.500033 USD/h 8.3 USD/h
Mumbai (IN) / ap-south-1
2.9709 USD/h 8.612 USD/h
Quebec (CA) / ca-central-1
6.7642 USD/h 8.687 USD/h
Dublin (IE) / eu-west-1
4.4961 USD/h 8.724 USD/h
Paris (FR) / eu-west-3
3.524533 USD/h 9.148 USD/h
London (GB) / eu-west-2
2.124433 USD/h 9.154 USD/h
Milan (IT) / eu-south-1
1.3157 USD/h 9.16 USD/h
California (US) / us-west-1
3.65485 USD/h 9.389 USD/h
Seoul (KR) / ap-northeast-2
4.5343 USD/h 9.624 USD/h
Frankfurt (DE) / eu-central-1
4.389333 USD/h 9.78 USD/h
Sydney (AU) / ap-southeast-2
4.012067 USD/h 10.174 USD/h
Cape Town (ZA) / af-south-1
10.381 USD/h 10.381 USD/h
Osaka (JP) / ap-northeast-3
4.5222 USD/h 10.562 USD/h
Tokyo (JP) / ap-northeast-1
5.3947 USD/h 10.562 USD/h
Singapore (SG) / ap-southeast-1
3.8876 USD/h 10.948 USD/h
Hong Kong (HK) / ap-east-1
1.774333 USD/h 12.048 USD/h
Sao Paulo (BR) / sa-east-1
2.127067 USD/h 13.298 USD/h

Processor

vCPUs
96
CPU Allocation
Dedicated
CPU Cores
48
CPU Speed
2.5 GHz
CPU Architecture
x86_64
CPU Manufacturer
Intel
CPU Family
Xeon
CPU Model
8259CL
CPU L1D Cache
32 KiB
CPU L1D Cache Total
2 MiB
CPU L1I Cache
32 KiB
CPU L1I Cache Total
2 MiB
CPU L2 Cache
1 MiB
CPU L2 Cache Total
48 MiB
CPU L3 Cache
36 MiB
CPU L3 Cache Total
72 MiB
CPU Flags
fpu, vme, de, pse, tsc, msr, pae, mce, cx8, apic, sep, mtrr, pge, mca, cmov, pat, pse36, clflush, dts, acpi, mmx, fxsr, sse, sse2, ss, ht, tm, pbe, syscall, nx, pdpe1gb, rdtscp, lm, constant_tsc, art, arch_perfmon, pebs, bts, rep_good, nopl, xtopology, nonstop_tsc, cpuid, aperfmperf, pni, pclmulqdq, dtes64, monitor, ds_cpl, vmx, smx, est, tm2, ssse3, sdbg, fma, cx16, xtpr, pdcm, pcid, dca, sse4_1, sse4_2, x2apic, movbe, popcnt, tsc_deadline_timer, aes, xsave, avx, f16c, rdrand, lahf_lm, abm, 3dnowprefetch, cpuid_fault, epb, cat_l3, cdp_l3, intel_ppin, ssbd, mba, ibrs, ibpb, stibp, ibrs_enhanced, tpr_shadow, flexpriority, ept, vpid, ept_ad, fsgsbase, tsc_adjust, bmi1, avx2, smep, bmi2, erms, invpcid, cqm, mpx, rdt_a, avx512f, avx512dq, rdseed, adx, smap, clflushopt, clwb, intel_pt, avx512cd, avx512bw, avx512vl, xsaveopt, xsavec, xgetbv1, xsaves, cqm_llc, cqm_occup_llc, cqm_mbm_total, cqm_mbm_local, dtherm, ida, arat, pln, pts, hwp, hwp_act_window, hwp_epp, hwp_pkg_req, vnmi, pku, ospke, avx512_vnni, md_clear, flush_l1d, arch_capabilities
Ecpus
56.2
Scalability
117.08

System Resources and Accelerators

MEMORY
Memory Amount
384 GiB
Memory Amount Actual
384 GiB
Memory Generation
DDR4
Memory Speed
2933 Mhz
GPU
GPU Count
8
GPU Memory Min
16 GiB
GPU Memory Total
128 GiB
GPU Manufacturer
NVIDIA
GPU Family
Turing
GPU Model
T4
GPUs
  • NVIDIA Turing Tesla T4 (Memory amount: 15360, Firmware version: 535.183.01, BIOS version: 90.04.96.00.02, Clock rate: 1590)
  • NVIDIA Turing Tesla T4 (Memory amount: 15360, Firmware version: 535.183.01, BIOS version: 90.04.96.00.02, Clock rate: 1590)
  • NVIDIA Turing Tesla T4 (Memory amount: 15360, Firmware version: 535.183.01, BIOS version: 90.04.96.00.02, Clock rate: 1590)
  • NVIDIA Turing Tesla T4 (Memory amount: 15360, Firmware version: 535.183.01, BIOS version: 90.04.96.00.02, Clock rate: 1590)
  • NVIDIA Turing Tesla T4 (Memory amount: 15360, Firmware version: 535.183.01, BIOS version: 90.04.96.00.02, Clock rate: 1590)
  • NVIDIA Turing Tesla T4 (Memory amount: 15360, Firmware version: 535.183.01, BIOS version: 90.04.96.00.02, Clock rate: 1590)
  • NVIDIA Turing Tesla T4 (Memory amount: 15360, Firmware version: 535.183.01, BIOS version: 90.04.96.00.02, Clock rate: 1590)
  • NVIDIA Turing Tesla T4 (Memory amount: 15360, Firmware version: 535.183.01, BIOS version: 90.04.96.00.02, Clock rate: 1590)
STORAGE
Storage Size
1800 GB
Storage Type
nvme ssd
Storages
  • 900 GB nvme ssd
  • 900 GB nvme ssd
NETWORK
Network Speed Baseline
100 Gbps
Network Speed Max
100 Gbps
Network Storage Speed Baseline
19 Gbps
Network Storage Speed Max
19 Gbps
Inbound Traffic
0 GB/month
Outbound Traffic
0 GB/month
IPv4
0

Server Description

A bare-metal server featuring eight GPUs, high-density multi-core processing, and high-speed local NVMe storage for parallelized workloads.

GPU AcceleratedCompute OptimizedStorage & Database

Amazon Web Services g4dn.metal is a bare-metal, GPU-accelerated server designed for parallel computing and graphics-intensive workloads. It is powered by an Intel Xeon 8259CL processor with 96 vCPUs and 48 physical cores running at 2.5 GHz, paired with 384.0 GB of DDR4 memory. The server features eight NVIDIA T4 GPUs with 128 GB of total VRAM, 1800 GB of local NVMe SSD storage, and a 100 Gbps baseline network bandwidth. Benchmarks show strong multi-core CPU performance and high memory bandwidth, though single-core performance remains average. This hardware profile makes the server highly efficient for multi-stream video transcoding, machine learning inference, graphics rendering, and high-throughput database operations.

Economics

Average Price per Region

Prices per Zone

Lowest Prices

Performance

Workload Profiles

3.24Score
Precomputed compound score for Cache Intensive workloads. A weighted average (geometric mean) of benchmark scores compared to their medians: score = ∏ (x_i / m_i)^(w_i / Σw). The score of 1.0 represents a synthetic baseline server with the median performance of each component benchmark; 0.5 means roughly half the performance; and 2.0 means twice the performance of that reference profile. Component weights: 50% Redis RPS (pipeline=1, SET), 20% Redis RPS (pipeline=16, SET), 10% PassMark Memory Mark (composite), 10% Memory bandwidth (read, 16 MB ~ L3), 10% PassMark single-thread CPU. Rationale for component selection: In-memory key-value store workload, mixing direct Redis performance metrics with memory speed and latency benchmarks, and single-core CPU performance profiles.
Component Score Weight Impact
Raw Reference Normalized Target Actual
Redis RPS (pipeline=1, SET) 9,702,493 ops/sec 1,917,397 ops/sec 5.06 50.00% 50.00% +125.00%
Redis RPS (pipeline=16, SET) 50,096,974 ops/sec 11,681,410 ops/sec 4.29 20.00% 20.00% +33.80%
PassMark Memory Mark (composite) 2,808 2,418 1.16 10.00% 10.00% +1.50%
Memory bandwidth (read, 16 MB ~ L3) 228,904 MB/sec 108,140 MB/sec 2.12 10.00% 10.00% +7.80%
PassMark single-thread CPU 2,039 Mops/s 2,365 Mops/s 0.862 10.00% 10.00% -1.47%

Memory Bandwidth

Compression

OpenSSL

Geekbench Single-Core

Score: 1,336

Geekbench Multi-Core

Score: 12,909

Passmark CPU Scores

BENCHMARKSCORE
Mark
 56876
Compression
 1187897
Encryption
 37685
Extended Instructions
 74889
Floating Point Maths
 147194
Integer Maths
 273893
Physics
 4988
Prime Numbers
 337
Single Threaded
 2039
String Sorting
 138454

Passmark Memory Scores

BENCHMARKSCORE
Memory Mark
 2808
Database Operations
 21849
Memory Latency
 50
Memory Read Cached
 24923
Memory Read Uncached
 10727
Memory Write
 10405

Stress-ng Raw Scores

Stress-ng Relative Multicore Performance

Static Web Server

Redis

Alternatives

Servers of the Same Family

INSTANCEvCPUsMEMORYGPUs
g4dn.xlarge 416 GiB1
g4dn.2xlarge 832 GiB1
g4dn.4xlarge 1664 GiB1
g4dn.8xlarge 32128 GiB1
g4dn.12xlarge 48192 GiB4
g4dn.16xlarge 64256 GiB1

Similar Servers

INSTANCEVENDORvCPUsMEMORYGPUs

g4dn.metal FAQs