g6.4xlarge by Amazon Web Services

g6.4xlarge is a Graphics intensive Gen6 4xlarge server offered by Amazon Web Services with 16 vCPUs, 64 GiB of memory and 600 GB of storage. The pricing starts at 0.1589 USD per hour.
16 vCPU
64 GiB Memory
600 GB Storage
1 GPU
Spare SCore
23,942
(All-cores)
2,993
(Single-core)

Specifications

Server Metadata

Vendor ID
aws
Name
g6.4xlarge
Description
Graphics intensive Gen6 4xlarge
Family
g6
Hw Virt
Status
active
Observed At
2026-11-02T08:25:35.989621

Availability

REGION / IDSPOTONDEMAND
Ohio (US) / us-east-2
0.6843 USD/h 1.3232 USD/h
Oregon (US) / us-west-2
0.852075 USD/h 1.3232 USD/h
Northern Virgina (US) / us-east-1
0.9809 USD/h 1.3232 USD/h
Aragón (ES) / eu-south-2
0.427067 USD/h 1.3942 USD/h
Stockholm (SE) / eu-north-1
0.3422 USD/h 1.4034 USD/h
Quebec (CA) / ca-central-1
1.071333 USD/h 1.4692 USD/h
Hyderabad (IN) / ap-south-2
0.7012 USD/h 1.5889 USD/h
Mumbai (IN) / ap-south-1
0.5821 USD/h 1.5889 USD/h
United Arab Emirates / me-central-1
0.6652 USD/h 1.6248 USD/h
Seoul (KR) / ap-northeast-2
0.748033 USD/h 1.627 USD/h
Frankfurt (DE) / eu-central-1
0.903233 USD/h 1.6547 USD/h
(MY) / ap-southeast-5
0.5619 USD/h 1.6668 USD/h
Paris (FR) / eu-west-3
0.51345 USD/h 1.6797 USD/h
London (GB) / eu-west-2
0.599933 USD/h 1.6797 USD/h
Sydney (AU) / ap-southeast-2
0.496867 USD/h 1.7204 USD/h
Zurich (CH) / eu-central-2
0.6099 USD/h 1.7839 USD/h
Tokyo (JP) / ap-northeast-1
0.91675 USD/h 1.919 USD/h
Sao Paulo (BR) / sa-east-1
0.2859 USD/h 2.2492 USD/h

Processor

vCPUs
16
Hypervisor
nitro
CPU Allocation
Dedicated
CPU Cores
8
CPU Speed
3.4 GHz
CPU Architecture
x86_64
CPU Manufacturer
AMD
CPU Family
EPYC
CPU Model
7R13
CPU L1D Cache
32 KiB
CPU L1D Cache Total
256 KiB
CPU L1I Cache
32 KiB
CPU L1I Cache Total
256 KiB
CPU L2 Cache
512 KiB
CPU L2 Cache Total
4 MiB
CPU L3 Cache
32 MiB
CPU L3 Cache Total
32 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, pcid, sse4_1, sse4_2, x2apic, 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, invpcid, rdseed, adx, smap, clflushopt, clwb, sha_ni, xsaveopt, xsavec, xgetbv1, clzero, xsaveerptr, rdpru, wbnoinvd, arat, npt, nrip_save, vaes, vpclmulqdq, rdpid
Ecpus
8
Scalability
50

System Resources and Accelerators

MEMORY
Memory Amount
64 GiB
Memory Amount Actual
64 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
Ada Lovelace
Accelerator Model
L4
Accelerators
  • NVIDIA Ada Lovelace NVIDIA L4 (Memory amount: 23034, Firmware version: 590.48.01, BIOS version: 95.04.65.00.37, Clock rate: 2040)
STORAGE
Storage Size
600 GB
Storage Type
nvme ssd
Storages
  • 600 GB nvme ssd
NETWORK
Network Speed Baseline
10 Gbps
Network Speed Max
25 Gbps
Network Storage Speed Baseline
8 Gbps
Network Storage Speed Max
8 Gbps
Inbound Traffic
0 GB/month
Outbound Traffic
0 GB/month
IPv4
0

CPU and System Topology

Server Description

An NVIDIA L4 GPU-accelerated instance with AMD EPYC processors and local NVMe storage for efficient machine learning inference and graphics workloads.

GPU AcceleratedGeneral PurposeStorage & Database

Amazon Web Services g6.4xlarge is a graphics-intensive instance featuring an AMD EPYC 7R13 processor with 16 vCPUs, 64.0 GB of DDR4 memory, and a dedicated NVIDIA Ada Lovelace L4 GPU with 22 GB of VRAM. Built on the AWS Nitro hypervisor, it includes 600 GB of local NVMe SSD storage and 10 Gbps baseline network bandwidth. Performance benchmarks show strong single-core CPU capabilities and top-tier execution in LLM inference tasks for small and medium models, though large-block memory write speeds represent a relative bottleneck. This instance is optimized for workloads such as machine learning inference, graphics rendering, and video transcoding, offering a balanced hardware profile for GPU-accelerated tasks.

Economics

Average Price per Region

Prices per Zone

Lowest Prices

Performance

Workload Profiles

0.76Score
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) 1,398,465 ops/sec 1,918,343 ops/sec 0.74 50.00% 50.00% -14.00%
Redis RPS (pipeline=16, SET) 9,966,691 ops/sec 11,704,734 ops/sec 0.83 20.00% 20.00% -3.70%
PassMark Memory Mark (composite) 2,567 2,416 1.1 10.00% 10.00% +0.96%
Memory bandwidth (read, 16 MB ~ L3) 40,993 MB/sec 108,466 MB/sec 0.37 10.00% 10.00% -9.50%
PassMark single-thread CPU 2,641 Mops/s 2,365 Mops/s 1.1 10.00% 10.00% +0.96%

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
779 ns
Benchmark Environment
Driver Version
 535.183.01
CUDA Version (driver)
 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: 1,741

Geekbench Multi-Core

Score: 10,077

Passmark CPU Scores

BENCHMARKSCORE
Mark
 23,623
Compression
 281,371
Encryption
 18,989
Extended Instructions
 16,501
Floating Point Maths
 40,348
Integer Maths
 73,648
Physics
 2,085
Prime Numbers
 103
Single Threaded
 2,641
String Sorting
 37,580

Passmark Memory Scores

BENCHMARKSCORE
Memory Mark
 2,567
Database Operations
 7,177
Memory Latency
 68
Memory Read Cached
 26,527
Memory Read Uncached
 17,144
Memory Write
 17,503

Stress-ng Stressors

CPU integerSingleMultiScaling
atomic
 277.27 226.2 0.82×
bitops
 156,789 1,449,434 9.24×
branch
 182,274 2,673,169 14.67×
funccall
 599.05 6,163 10.29×
funcret
 19,015 197,329 10.38×
hash
 151,376 1,845,217 12.19×
intmath
 64,503 700,592 10.86×
nop
 314,188 2,529,126 8.05×
prime
 433,455 4,255,262 9.82×
rotate
 85,622 684,966 8.00×
str
 951,091 9,387,256 9.87×
wcs
 624,029 5,844,669 9.37×
CryptographySingleMultiScaling
crypt
 3,594 28,704 7.99×
CompressionSingleMultiScaling
jpeg
 812.46 8,468 10.42×
zlib
 52.73 846.65 16.06×
HypervisorSingleMultiScaling
clock
 16,801 174,557 10.39×
clone
 288.33 1,923 6.67×
context
 2,124 24,026 11.31×
cpu-sched
 70.11 1,027 14.65×
eventfd
 33,309 399,035 11.98×
fault
 8,323 105,096 12.63×
fork
 4,134 42,037 10.17×
getrandom
 47,437 482,420 10.17×
hrtimers
 244,490 1,405,446 5.75×
mprotect
 1,258,310 2,421,199 1.92×
mtx
 36,814,789 723,730,693 19.66×
munmap
 2,380 29,065 12.21×
null
 226,945 2,477,577 10.92×
pipe
 1,299,392 5,218,442 4.02×
priv-instr
 374,664 4,452,068 11.88×
resched
 3,006,552 861,817 0.29×
switch
 60,889 2,257,826 37.08×
timermix
 143,942 1,783,701 12.39×
tlb-shootdown
 14,609 48,220 3.30×
tsc
 741,667 11,409,775 15.38×
usersyscall
 247,050 2,911,670 11.79×
vdso
 87,160,217 1,148,111,667 13.17×
vfork
 26,793 208,617 7.79×
x86syscall
 2,654,751 28,559,792 10.76×
yield
 7,412,273 7,519,568 1.01×
zero
 1,709,016 18,595,070 10.88×
CPU float/SIMDSingleMultiScaling
eigen
 15,805 129,927 8.22×
expmath
 2,862 36,581 12.78×
fma
 864,611 10,839,417 12.54×
fp
 252 2,880 11.43×
fractal
 7.4 99.01 13.38×
hyperbolic
 3,647 44,340 12.16×
logmath
 4,296 49,309 11.48×
matrix
 2,995 35,305 11.79×
matrix-3d
 337.82 1,722 5.10×
monte-carlo
 128.18 1,373 10.71×
mpfr
 1,549 15,017 9.69×
powmath
 1,352 17,624 13.03×
trig
 1,650 20,349 12.33×
veccmp
 17,729 186,692 10.53×
vecfp
 1,635 18,016 11.02×
vecmath
 3,788 48,924 12.92×
vecshuf
 591.59 5,511 9.32×
vecwide
 26,240 266,825 10.17×
AlgorithmsSingleMultiScaling
bitonicsort
 9.6 80 8.33×
bsearch
 292.42 2,995 10.24×
bubblesort
 0.8 9.6 12.00×
fibsearch
 355.78 3,414 9.60×
heapsort
 10.2 102.4 10.04×
hsearch
 8,485 76,096 8.97×
insertionsort
 12.8 102.4 8.00×
malloc
 3,492,233 38,435,027 11.01×
mcontend
 236.29 3,398 14.38×
mergesort
 30.8 356.59 11.58×
qsort
 21.2 231.19 10.91×
skiplist
 6,820 83,756 12.28×
tree
 1.8 12.79 7.11×
MemorySingleMultiScaling
cache
 668,275 1,611,680 2.41×
cachehammer
 250,749 3,753,859 14.97×
cacheline
 42.58 672.92 15.80×
l1cache
 177.68 1,377 7.75×
memcpy
 406.49 3,096 7.62×
memrate
 2.53 6.08 2.40×
memthrash
 163.51 475.58 2.91×
ptr-chase
 48,547,166 130,935,358 2.70×
spinmem
 220.99 608.95 2.76×
stream
 94.45 2,583 27.35×
vm
 58,442 292,988 5.01×

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
g6.xlarge 416 GiB1
g6.2xlarge 832 GiB1
g6.8xlarge 32128 GiB1
g6.12xlarge 48192 GiB4
g6.16xlarge 64256 GiB1
g6.24xlarge 96384 GiB4
g6.48xlarge 192768 GiB8

Similar Servers

INSTANCEVENDORvCPUsMEMORYAccelerators

g6.4xlarge FAQs