g2-standard-24 by Google Cloud Platform

g2-standard-24 is a Graphics Optimized: 2 NVIDIA L4 GPUs, 24 vCPUs, 96GB RAM server offered by Google Cloud Platform with 24 vCPUs, 96 GiB of memory and 0 GB of storage. The pricing starts at 0.8251 USD per hour.
24 vCPU
96 GiB Memory
2 GPU
Spare SCore
18,985
(All-cores)
1,201
(Single-core)

Specifications

Server Metadata

Vendor ID
gcp
Server ID
1600024
Name
g2-standard-24
Description
Graphics Optimized: 2 NVIDIA L4 GPUs, 24 vCPUs, 96GB RAM
Family
g2
Hw Virt
Average Time To Start
20
Status
active
Observed At
2026-10-27T15:26:20.225018

Availability

REGION / IDSPOTONDEMAND
Ashburn (US) / us-east4
1.0109 USD/h 1.9943 USD/h
Council Bluffs (US) / us-central1
1.1973 USD/h 2.0008 USD/h
The Dalles (US) / us-west1
1.1973 USD/h 2.0008 USD/h
Moncks Corner (US) / us-east1
1.1973 USD/h 2.0008 USD/h
Mumbai (IN) / asia-south1
1.1137 USD/h 2.0828 USD/h
Eemshaven (NL) / europe-west4
1.2617 USD/h 2.103 USD/h
Montréal (CA) / northamerica-northeast1
1.3118 USD/h 2.2031 USD/h
Toronto (CA) / northamerica-northeast2
1.3149 USD/h 2.2031 USD/h
St. Ghislain (BE) / europe-west1
1.3219 USD/h 2.2031 USD/h
Las Vegas (US) / us-west4
1.3488 USD/h 2.2535 USD/h
London (GB) / europe-west2
1.3669 USD/h 2.2781 USD/h
Changhua County (TW) / asia-east1
1.388 USD/h 2.3168 USD/h
Frankfurt (DE) / europe-west3
1.4078 USD/h 2.3581 USD/h
Jurong West (SG) / asia-southeast1
1.4812 USD/h 2.4684 USD/h
Seoul (KR) / asia-northeast3
0.8251 USD/h 2.5695 USD/h
Tokyo (JP) / asia-northeast1
1.2525 USD/h 2.5695 USD/h
Zurich (CH) / europe-west6
1.6283 USD/h 2.7207 USD/h
Dammam (SA) / me-central2
1.8967 USD/h 3.2013 USD/h

Processor

vCPUs
24
CPU Allocation
Dedicated
CPU Cores
12
CPU Speed
2.2 GHz
CPU Architecture
x86_64
CPU Manufacturer
Intel
CPU Family
Xeon
CPU L1D Cache
32 KiB
CPU L1D Cache Total
384 KiB
CPU L1I Cache
32 KiB
CPU L1I Cache Total
384 KiB
CPU L2 Cache
1 MiB
CPU L2 Cache Total
12 MiB
CPU L3 Cache
39 MiB
CPU L3 Cache Total
39 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, ss, ht, syscall, nx, pdpe1gb, rdtscp, lm, constant_tsc, rep_good, nopl, xtopology, nonstop_tsc, cpuid, tsc_known_freq, pni, pclmulqdq, ssse3, fma, cx16, pcid, sse4_1, sse4_2, x2apic, movbe, popcnt, aes, xsave, avx, f16c, rdrand, hypervisor, lahf_lm, abm, 3dnowprefetch, ssbd, ibrs, ibpb, stibp, ibrs_enhanced, fsgsbase, tsc_adjust, bmi1, hle, avx2, smep, bmi2, erms, invpcid, rtm, mpx, avx512f, avx512dq, rdseed, adx, smap, clflushopt, clwb, avx512cd, avx512bw, avx512vl, xsaveopt, xsavec, xgetbv1, xsaves, arat, avx512_vnni, md_clear, arch_capabilities
Ecpus
15.8
Scalability
65.83

System Resources and Accelerators

MEMORY
Memory Amount
96 GiB
Memory Amount Actual
96 GiB
ACCELERATORS
Accelerator Count
2
Accelerator Memory Min
24 GiB
Accelerator Memory Total
48 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: 95.04.29.00.07, Clock rate: 2040)
  • NVIDIA Ada Lovelace NVIDIA L4 (Memory amount: 23034, Firmware version: 590.48.01, BIOS version: 95.04.29.00.07, Clock rate: 2040)
STORAGE
Storage Size
0 GB
Storages
    NETWORK
    Inbound Traffic
    0 GB/month
    Outbound Traffic
    0 GB/month
    IPv4
    0

    CPU and System Topology

    Server Description

    A graphics-optimized server featuring dual GPUs and balanced memory, delivering top-tier performance for large language model inference workloads.

    GPU AcceleratedGeneral Purpose

    Google Cloud Platform g2-standard-24 is a graphics-optimized server featuring 24 vCPUs, 96.0 GB of RAM, and two NVIDIA L4 GPUs with 48 GB of total VRAM. Powered by an Intel Xeon x86_64 processor running at 2.2 GHz, the instance allocates 12 physical cores with hyperthreading. It does not include local storage. Performance benchmarks indicate a significant contrast between CPU and GPU capabilities. Single-core CPU performance is weak, ranking in the bottom 25% for tasks like single-threaded compression and Geekbench single-core tests. However, multi-core CPU performance and memory bandwidth remain average, while LLM inference speeds for prompt processing and text generation rank in the top 10% and top 25% respectively. This server is optimized for GPU-accelerated workloads, such as machine learning inference and graphics rendering, where the dual NVIDIA L4 GPUs can be fully leveraged, rather than single-threaded CPU-bound applications.

    Economics

    Average Price per Region

    Prices per Zone

    Lowest Prices

    Performance

    Workload Profiles

    0.95Score
    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) 1,877,652 ops/sec 1,918,343 ops/sec 1 50.00% 50.00% 0.00%
    Redis RPS (pipeline=16, SET) 11,452,540 ops/sec 11,704,734 ops/sec 0.92 20.00% 20.00% -1.70%
    PassMark Memory Mark (composite) 2,543 2,416 1 10.00% 10.00% 0.00%
    Memory bandwidth (read, 16 MB ~ L3) 110,082 MB/sec 108,420 MB/sec 1 10.00% 10.00% 0.00%
    PassMark single-thread CPU 1,713 Mops/s 2,365 Mops/s 0.71 10.00% 10.00% -3.40%

    GPU Bandwidth and Latency

    Node (all GPUs)
    Node baseline directional bandwidth
    		6.5 GB/s
    Node GPU-to-host bandwidth
    		13.2 GB/s
    Node host-to-GPU bandwidth
    		12.4 GB/s
    Peer-to-peer
    GPU-to-GPU P2P duplex bandwidth
    		5.0 GB/s
    GPU-to-GPU P2P gather bandwidth
    		9.9 GB/s
    GPU-to-GPU P2P latency
    		919 ns
    GPU-to-GPU P2P bandwidth
    		9.9 GB/s
    Single GPU slot
    Per-GPU GPU-to-host bandwidth
    		13.2 GB/s
    Per-GPU host-to-GPU bandwidth
    		12.3 GB/s
    Per-GPU host-device latency
    		919 ns
    Benchmark Environment
    Driver Version
    		 535.183.01
    CUDA Runtime Version
    		 12.6
    nvbandwidth Version
    		 v0.5
    GPU Count
    		 2
    P2P Supported
    Yes
    Efficiency
    SM/CE copy efficiency
    		1.00

    Memory Bandwidth

    Compression

    OpenSSL

    Geekbench Single-Core

    Score: 1,059

    Geekbench Multi-Core

    Score: 8,145

    Passmark CPU Scores

    BENCHMARKSCORE
    Mark
    		 19,045
    Compression
    		 278,913
    Encryption
    		 8,663
    Extended Instructions
    		 18,063
    Floating Point Maths
    		 34,111
    Integer Maths
    		 63,470
    Physics
    		 1,945
    Prime Numbers
    		 102
    Single Threaded
    		 1,713
    String Sorting
    		 32,680

    Passmark Memory Scores

    BENCHMARKSCORE
    Memory Mark
    		 2,543
    Database Operations
    		 7,776
    Memory Latency
    		 52
    Memory Read Cached
    		 20,669
    Memory Read Uncached
    		 10,575
    Memory Write
    		 9,969

    Stress-ng Stressors

    CPU integerSingleMultiScaling
    atomic
    		 179.17 173.87 0.97×
    bitops
    		 92,330 1,270,448 13.76×
    branch
    		 131,370 2,385,597 18.16×
    funccall
    		 463.19 5,714 12.34×
    funcret
    		 9,638 164,806 17.10×
    hash
    		 107,357 1,836,705 17.11×
    intmath
    		 19,180 341,521 17.81×
    nop
    		 163,170 1,689,645 10.36×
    prime
    		 270,959 3,639,194 13.43×
    rotate
    		 82,374 1,361,364 16.53×
    str
    		 642,976 8,199,527 12.75×
    wcs
    		 411,037 5,155,955 12.54×
    CryptographySingleMultiScaling
    crypt
    		 2,082 24,289 11.67×
    CompressionSingleMultiScaling
    jpeg
    		 514.04 7,064 13.74×
    zlib
    		 48.83 782.07 16.02×
    HypervisorSingleMultiScaling
    clock
    		 25,266 255,591 10.12×
    clone
    		 294.87 2,874 9.75×
    context
    		 2,239 30,829 13.77×
    cpu-sched
    		 55.2 1,536 27.82×
    eventfd
    		 46,894 556,072 11.86×
    fault
    		 9,373 136,143 14.53×
    fork
    		 3,539 43,211 12.21×
    getrandom
    		 38,171 558,008 14.62×
    hrtimers
    		 321,905 1,974,611 6.13×
    mprotect
    		 1,022,129 3,308,475 3.24×
    mtx
    		 9,090,550 526,760,914 57.95×
    munmap
    		 2,313 30,522 13.20×
    null
    		 251,774 3,078,618 12.23×
    pipe
    		 831,296 7,326,661 8.81×
    priv-instr
    		 480,964 6,439,518 13.39×
    resched
    		 3,781,128 1,286,965 0.34×
    switch
    		 78,834 3,461,064 43.90×
    timermix
    		 168,722 2,844,643 16.86×
    tlb-shootdown
    		 9,237 39,252 4.25×
    tsc
    		 849,368 14,186,229 16.70×
    usersyscall
    		 330,016 4,404,703 13.35×
    vdso
    		 70,522,276 1,189,409,350 16.87×
    vfork
    		 25,507 201,559 7.90×
    x86syscall
    		 2,909,928 35,787,514 12.30×
    yield
    		 11,393,234 11,718,304 1.03×
    zero
    		 1,730,031 22,074,796 12.76×
    CPU float/SIMDSingleMultiScaling
    eigen
    		 13,801 198,244 14.36×
    expmath
    		 2,279 41,800 18.34×
    fma
    		 508,230 9,975,382 19.63×
    fp
    		 107.47 1,585 14.75×
    fractal
    		 4.2 73.81 17.57×
    hyperbolic
    		 2,384 44,620 18.72×
    logmath
    		 3,859 68,175 17.67×
    matrix
    		 2,592 50,171 19.36×
    matrix-3d
    		 608.13 3,318 5.46×
    monte-carlo
    		 118.2 1,581 13.37×
    mpfr
    		 1,031 13,259 12.86×
    powmath
    		 1,350 24,733 18.32×
    trig
    		 1,458 23,171 15.89×
    veccmp
    		 13,319 165,361 12.42×
    vecfp
    		 1,151 17,037 14.80×
    vecmath
    		 1,864 35,241 18.91×
    vecshuf
    		 481.3 6,191 12.86×
    vecwide
    		 21,168 286,606 13.54×
    AlgorithmsSingleMultiScaling
    bitonicsort
    		 6.2 86.2 13.90×
    bsearch
    		 203.71 2,813 13.81×
    bubblesort
    		 0.6 4.8 8.00×
    fibsearch
    		 208.39 2,871 13.78×
    heapsort
    		 7.4 105.6 14.27×
    hsearch
    		 4,273 63,074 14.76×
    insertionsort
    		 7 81.6 11.66×
    malloc
    		 1,996,413 27,424,492 13.74×
    mcontend
    		 109.06 4,736 43.43×
    mergesort
    		 19.8 316.59 15.99×
    qsort
    		 12.6 172.79 13.71×
    skiplist
    		 4,154 69,920 16.83×
    tree
    		 1.2 14.39 11.99×
    MemorySingleMultiScaling
    cache
    		 333,142 575,249 1.73×
    cachehammer
    		 369,178 7,281,267 19.72×
    cacheline
    		 31.77 798.66 25.14×
    l1cache
    		 188.79 2,492 13.20×
    memcpy
    		 295.75 3,807 12.87×
    memrate
    		 1.17 8.99 7.68×
    memthrash
    		 139.39 710.28 5.10×
    ptr-chase
    		 26,905,369 237,075,144 8.81×
    spinmem
    		 190.98 855.11 4.48×
    stream
    		 28.69 6,174 215.19×
    vm
    		 49,687 295,317 5.94×

    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
    g2-standard-4 416 GiB1
    g2-standard-8 832 GiB1
    g2-standard-12 1248 GiB1
    g2-standard-16 1664 GiB1
    g2-standard-32 32128 GiB1
    g2-standard-48 48192 GiB4
    g2-standard-96 96384 GiB8

    Similar Servers

    INSTANCEVENDORvCPUsMEMORYAccelerators

    g2-standard-24 FAQs