g5.2xlarge by Amazon Web Services

g5.2xlarge is a Graphics intensive Gen5 2xlarge server offered by Amazon Web Services with 8 vCPUs, 32 GiB of memory and 450 GB of storage. The pricing starts at 0.1696 USD per hour.
8 vCPU
32 GiB Memory
450 GB Storage
1 GPU
Spare SCore
3,072
(All-cores)
768
(Single-core)

Specifications

Server Metadata

Vendor ID
aws
Name
g5.2xlarge
Description
Graphics intensive Gen5 2xlarge
Family
g5
Hw Virt
Status
active
Observed At
2026-10-27T06:58:17.797250

Availability

REGION / IDSPOTONDEMAND
Oregon (US) / us-west-2
0.509167 USD/h 1.212 USD/h
Northern Virgina (US) / us-east-1
0.86664 USD/h 1.212 USD/h
Ohio (US) / us-east-2
0.6499 USD/h 1.212 USD/h
Aragón (ES) / eu-south-2
1.1778 USD/h 1.2771 USD/h
Stockholm (SE) / eu-north-1
0.2783 USD/h 1.2855 USD/h
Quebec (CA) / ca-central-1
1.3457 USD/h 1.3457 USD/h
Dublin (IE) / eu-west-1
0.918 USD/h 1.353 USD/h
Tel Aviv (IL) / il-central-1
0.767333 USD/h 1.4206 USD/h
Mumbai (IN) / ap-south-1
0.7442 USD/h 1.4554 USD/h
United Arab Emirates / me-central-1
0.6086 USD/h 1.4867 USD/h
Seoul (KR) / ap-northeast-2
0.6812 USD/h 1.4903 USD/h
Frankfurt (DE) / eu-central-1
0.574933 USD/h 1.5156 USD/h
Paris (FR) / eu-west-3
0.493933 USD/h 1.5385 USD/h
London (GB) / eu-west-2
0.5462 USD/h 1.5385 USD/h
Sydney (AU) / ap-southeast-2
1.5758 USD/h 1.5758 USD/h
Jakarta (ID) / ap-southeast-3
0.4962 USD/h 1.6963 USD/h
Tokyo (JP) / ap-northeast-1
0.8488 USD/h 1.7578 USD/h
Hong Kong (HK) / ap-east-1
0.714 USD/h 1.8659 USD/h
Sao Paulo (BR) / sa-east-1
0.4259 USD/h 2.0602 USD/h

Processor

vCPUs
8
Hypervisor
nitro
CPU Allocation
Dedicated
CPU Cores
4
CPU Speed
3.3 GHz
CPU Architecture
x86_64
CPU Manufacturer
AMD
CPU Family
EPYC
CPU Model
7R32
CPU L1D Cache
32 KiB
CPU L1D Cache Total
128 KiB
CPU L1I Cache
32 KiB
CPU L1I Cache Total
128 KiB
CPU L2 Cache
512 KiB
CPU L2 Cache Total
2 MiB
CPU L3 Cache
16 MiB
CPU L3 Cache Total
16 MiB
CPU Flags
fpu, vme, de, pse, tsc, msr, pae, mce, cx8, apic, sep, mtrr, pge, mca, cmov, pat, pse36, clflush, mmx, fxsr, sse, sse2, ht, syscall, nx, mmxext, fxsr_opt, pdpe1gb, rdtscp, lm, constant_tsc, rep_good, nopl, nonstop_tsc, cpuid, extd_apicid, aperfmperf, tsc_known_freq, pni, pclmulqdq, ssse3, fma, cx16, sse4_1, sse4_2, movbe, popcnt, aes, xsave, avx, f16c, rdrand, hypervisor, lahf_lm, cmp_legacy, cr8_legacy, abm, sse4a, misalignsse, 3dnowprefetch, topoext, ssbd, ibrs, ibpb, stibp, vmmcall, fsgsbase, bmi1, avx2, smep, bmi2, rdseed, adx, smap, clflushopt, clwb, sha_ni, xsaveopt, xsavec, xgetbv1, clzero, xsaveerptr, rdpru, wbnoinvd, arat, npt, nrip_save, rdpid
Ecpus
4
Scalability
50

System Resources and Accelerators

MEMORY
Memory Amount
32 GiB
Memory Amount Actual
32 GiB
Memory Generation
DDR4
Memory Speed
3200 Mhz
ACCELERATORS
Accelerator Count
1
Accelerator Memory Min
22 GiB
Accelerator Memory Total
22 GiB
Accelerator Type
GPU
Accelerator Manufacturer
NVIDIA
Accelerator Family
Ampere
Accelerator Model
A10G
Accelerators
  • NVIDIA Ampere NVIDIA A10G (Memory amount: 23028, Firmware version: 590.48.01, BIOS version: 94.02.75.00.01, Clock rate: 1710)
STORAGE
Storage Size
450 GB
Storage Type
nvme ssd
Storages
  • 450 GB nvme ssd
NETWORK
Network Speed Baseline
5 Gbps
Network Speed Max
10 Gbps
Network Storage Speed Baseline
0.85 Gbps
Network Storage Speed Max
3.5 Gbps
Inbound Traffic
0 GB/month
Outbound Traffic
0 GB/month
IPv4
0

CPU and System Topology

Server Description

A GPU-accelerated server featuring dedicated NVIDIA Ampere graphics and local NVMe storage, optimized for machine learning inference and graphics workloads.

GPU AcceleratedGeneral Purpose

Amazon Web Services g5.2xlarge is a GPU-accelerated server instance featuring an AMD EPYC 7R32 processor with 8 vCPUs, 4 physical cores, and a 3.3 GHz clock speed. It is equipped with 32.0 GB of DDR4 memory, a 450 GB local NVMe SSD, and a baseline network bandwidth of 5 Gbps. Acceleration is driven by a single NVIDIA Ampere A10G GPU with 22 GB of VRAM. Benchmark results show top-tier performance in LLM inference for small and medium models, while general CPU compute and large-block memory bandwidth remain weak to poor. This hardware profile makes the instance highly suitable for machine learning inference, graphics-intensive applications, and video rendering, though it is less efficient for CPU-heavy or memory-bandwidth-intensive workloads.

Economics

Average Price per Region

Prices per Zone

Lowest Prices

Performance

Workload Profiles

0.43Score
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) 659,542 ops/sec 1,918,343 ops/sec 0.35 50.00% 50.00% -41.00%
Redis RPS (pipeline=16, SET) 4,331,575 ops/sec 11,704,734 ops/sec 0.36 20.00% 20.00% -18.00%
PassMark Memory Mark (composite) 2,544 2,416 1 10.00% 10.00% 0.00%
Memory bandwidth (read, 16 MB ~ L3) 36,860 MB/sec 108,420 MB/sec 0.34 10.00% 10.00% -10.00%
PassMark single-thread CPU 2,139 Mops/s 2,365 Mops/s 0.88 10.00% 10.00% -1.30%

GPU Bandwidth and Latency

Node (all GPUs)
Node baseline directional bandwidth
11.0 GB/s
Node GPU-to-host bandwidth
13.2 GB/s
Node host-to-GPU bandwidth
13.5 GB/s
Single GPU slot
Per-GPU GPU-to-host bandwidth
13.2 GB/s
Per-GPU host-to-GPU bandwidth
13.5 GB/s
Per-GPU host-device latency
912 ns
Benchmark Environment
Driver Version
 535.183.01
CUDA Runtime Version
 12.6
nvbandwidth Version
 v0.5
GPU Count
 1
P2P Supported
No
Efficiency
SM/CE copy efficiency
1.00

Memory Bandwidth

Compression

OpenSSL

Geekbench Single-Core

Score: 1,341

Geekbench Multi-Core

Score: 5,187

Passmark CPU Scores

BENCHMARKSCORE
Mark
 10,578
Compression
 121,963
Encryption
 8,509
Extended Instructions
 7,315
Floating Point Maths
 16,095
Integer Maths
 27,006
Physics
 991
Prime Numbers
 44
Single Threaded
 2,139
String Sorting
 17,212

Passmark Memory Scores

BENCHMARKSCORE
Memory Mark
 2,544
Database Operations
 2,859
Memory Latency
 50
Memory Read Cached
 24,723
Memory Read Uncached
 12,760
Memory Write
 13,515

Stress-ng Stressors

CPU integerSingleMultiScaling
atomic
 387.45 368.57 0.95×
bitops
 120,812 609,312 5.04×
branch
 152,942 1,062,344 6.95×
funccall
 502.47 2,710 5.39×
funcret
 14,058 85,443 6.08×
hash
 131,871 821,490 6.23×
intmath
 19,594 84,741 4.32×
nop
 236,716 955,974 4.04×
prime
 346,853 1,831,034 5.28×
rotate
 59,834 249,489 4.17×
str
 612,406 3,216,425 5.25×
wcs
 480,278 2,143,185 4.46×
CryptographySingleMultiScaling
crypt
 2,817 12,343 4.38×
CompressionSingleMultiScaling
jpeg
 641.32 3,496 5.45×
zlib
 52.55 328.93 6.26×
HypervisorSingleMultiScaling
clock
 19,771 106,009 5.36×
clone
 211.06 733.09 3.47×
context
 3,378 23,015 6.81×
cpu-sched
 59.65 510.95 8.57×
eventfd
 24,770 327,804 13.23×
fault
 8,333 46,732 5.61×
fork
 3,752 21,892 5.84×
getrandom
 36,386 195,143 5.36×
hrtimers
 711,345 671,834 0.94×
mprotect
 1,076,369 1,486,425 1.38×
mtx
 15,315,653 391,902,707 25.59×
munmap
 2,480 15,043 6.07×
null
 415,259 2,676,992 6.45×
pipe
 1,695,951 2,671,157 1.58×
priv-instr
 426,728 2,891,602 6.78×
resched
 1,344,847 455,092 0.34×
switch
 35,711 1,072,675 30.04×
timermix
 138,565 924,913 6.67×
tlb-shootdown
 11,169 19,710 1.76×
tsc
 653,690 5,088,612 7.78×
usersyscall
 290,645 1,968,480 6.77×
vdso
 79,298,750 431,210,771 5.44×
vfork
 20,660 104,500 5.06×
x86syscall
 5,276,422 34,245,384 6.49×
yield
 3,609,514 3,479,242 0.96×
zero
 2,106,702 13,330,181 6.33×
CPU float/SIMDSingleMultiScaling
eigen
 12,505 58,488 4.68×
expmath
 2,781 17,519 6.30×
fma
 739,963 4,669,871 6.31×
fp
 208.56 1,214 5.82×
fractal
 6.4 40.2 6.28×
hyperbolic
 3,642 21,875 6.01×
logmath
 4,104 23,962 5.84×
matrix
 2,729 19,974 7.32×
matrix-3d
 517.56 1,259 2.43×
monte-carlo
 183.48 1,202 6.55×
mpfr
 1,267 6,551 5.17×
powmath
 1,341 8,658 6.46×
trig
 1,559 9,548 6.12×
veccmp
 17,106 82,959 4.85×
vecfp
 1,388 8,351 6.02×
vecmath
 3,366 21,732 6.46×
vecshuf
 482.66 2,393 4.96×
vecwide
 26,536 127,726 4.81×
AlgorithmsSingleMultiScaling
bitonicsort
 6.8 33.6 4.94×
bsearch
 224.17 1,004 4.48×
bubblesort
 0.8 3.2 4.00×
fibsearch
 262.22 1,324 5.05×
heapsort
 8.6 48.2 5.60×
hsearch
 5,370 26,289 4.90×
insertionsort
 6.8 31 4.56×
malloc
 2,631,173 16,010,451 6.08×
mcontend
 226.72 1,467 6.47×
mergesort
 26 156.99 6.04×
qsort
 16.6 100.4 6.05×
skiplist
 5,507 35,196 6.39×
tree
 1.8 4.8 2.67×
MemorySingleMultiScaling
cache
 517,362 1,115,444 2.16×
cachehammer
 265,280 2,041,655 7.70×
cacheline
 54.28 441.32 8.13×
l1cache
 197.78 770.67 3.90×
memcpy
 353.5 1,479 4.18×
memrate
 1.95 4.62 2.37×
memthrash
 76.49 278.67 3.64×
ptr-chase
 29,092,953 69,785,828 2.40×
spinmem
 229.59 354.96 1.55×
stream
 175.27 2,233 12.74×
vm
 55,708 291,516 5.23×

Stress-ng Raw Scores

Stress-ng Relative Multicore Performance

LLM Inference Speed for Prompt Processing

LLM Inference Speed for Text Generation

Static Web Server

Redis

Alternatives

Servers of the Same Family

INSTANCEvCPUsMEMORYAccelerators
g5.xlarge 416 GiB1
g5.4xlarge 1664 GiB1
g5.8xlarge 32128 GiB1
g5.12xlarge 48192 GiB4
g5.16xlarge 64256 GiB1
g5.24xlarge 96384 GiB4
g5.48xlarge 192768 GiB8

Similar Servers

INSTANCEVENDORvCPUsMEMORYAccelerators

g5.2xlarge FAQs