g2-standard-48 by Google Cloud Platform

g2-standard-48 is a Graphics Optimized: 4 NVIDIA L4 GPUs, 48 vCPUs, 192GB RAM server offered by Google Cloud Platform with 48 vCPUs, 192 GiB of memory and 0 GB of storage. The pricing starts at 1.7322 USD per hour.
48 vCPU
192 GiB Memory
4 GPU
Spare SCore
37,916
(All-cores)
1,196
(Single-core)

Specifications

Server Metadata

Vendor ID
gcp
Server ID
1600048
Name
g2-standard-48
Description
Graphics Optimized: 4 NVIDIA L4 GPUs, 48 vCPUs, 192GB RAM
Family
g2
Hw Virt
Average Time To Start
23
Status
active
Observed At
2026-11-05T11:58:09.227745

Availability

REGION / IDSPOTONDEMAND
Ashburn (US) / us-east4
2.1227 USD/h 3.9886 USD/h
Council Bluffs (US) / us-central1
2.3946 USD/h 4.0017 USD/h
The Dalles (US) / us-west1
2.3946 USD/h 4.0017 USD/h
Moncks Corner (US) / us-east1
2.3946 USD/h 4.0017 USD/h
Mumbai (IN) / asia-south1
2.1157 USD/h 4.1655 USD/h
Eemshaven (NL) / europe-west4
2.5233 USD/h 4.206 USD/h
Montréal (CA) / northamerica-northeast1
2.6237 USD/h 4.4062 USD/h
Toronto (CA) / northamerica-northeast2
2.6298 USD/h 4.4062 USD/h
St. Ghislain (BE) / europe-west1
2.6438 USD/h 4.4062 USD/h
Las Vegas (US) / us-west4
2.6975 USD/h 4.5071 USD/h
London (GB) / europe-west2
2.7338 USD/h 4.5562 USD/h
Changhua County (TW) / asia-east1
2.7761 USD/h 4.6335 USD/h
Frankfurt (DE) / europe-west3
2.8156 USD/h 4.7161 USD/h
Jurong West (SG) / asia-southeast1
2.9625 USD/h 4.9369 USD/h
Seoul (KR) / asia-northeast3
1.7322 USD/h 5.1389 USD/h
Tokyo (JP) / asia-northeast1
2.7555 USD/h 5.1389 USD/h
Zurich (CH) / europe-west6
3.2566 USD/h 5.4415 USD/h
Dammam (SA) / me-central2
3.6035 USD/h 6.4027 USD/h

Processor

vCPUs
48
CPU Allocation
Dedicated
CPU Cores
24
CPU Speed
2.2 GHz
CPU Architecture
x86_64
CPU Manufacturer
Intel
CPU Family
Xeon
CPU L1D Cache
32 KiB
CPU L1D Cache Total
768 KiB
CPU L1I Cache
32 KiB
CPU L1I Cache Total
768 KiB
CPU L2 Cache
1 MiB
CPU L2 Cache Total
24 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
31.7
Scalability
66.04

System Resources and Accelerators

MEMORY
Memory Amount
192 GiB
Memory Amount Actual
192 GiB
ACCELERATORS
Accelerator Count
4
Accelerator Memory Min
24 GiB
Accelerator Memory Total
96 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)
  • 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 quad-GPU accelerated server featuring 96 GB of VRAM and 48 vCPUs, optimized for high-throughput machine learning inference and parallel computing.

    GPU AcceleratedCompute Optimized

    Google Cloud Platform g2-standard-48 is a graphics-optimized server designed for GPU-accelerated workloads. It features 48 vCPUs powered by an Intel Xeon processor with 24 physical cores running at 2.2 GHz, paired with 192.0 GB of system memory. The server is equipped with four NVIDIA L4 GPUs from the Ada Lovelace architecture, providing a total of 96 GB of VRAM, though it lacks local storage. Benchmark data highlights a significant performance divergence: while single-core CPU performance is weak, multi-core capabilities are average, and GPU-driven LLM inference speeds rank in the top tier. This resource profile makes the instance highly efficient for parallel processing tasks. It is best suited for machine learning inference, large language model deployment, and graphics rendering workloads where GPU acceleration offsets the modest single-core CPU performance.

    Economics

    Average Price per Region

    Prices per Zone

    Lowest Prices

    Performance

    Workload Profiles

    1.5Score
    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) 3,421,975 ops/sec 1,937,246 ops/sec 1.8 50.00% 50.00% +34.00%
    Redis RPS (pipeline=16, SET) 20,098,406 ops/sec 11,803,344 ops/sec 1.7 20.00% 20.00% +11.00%
    PassMark Memory Mark (composite) 2,596 2,409 1.1 10.00% 10.00% +0.96%
    Memory bandwidth (read, 16 MB ~ L3) 110,051 MB/sec 108,490 MB/sec 1 10.00% 10.00% 0.00%
    PassMark single-thread CPU 1,703 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
    		12.9 GB/s
    Node GPU-to-host bandwidth
    		26.4 GB/s
    Node host-to-GPU bandwidth
    		24.8 GB/s
    Peer-to-peer
    GPU-to-GPU P2P duplex bandwidth
    		8.2 GB/s
    GPU-to-GPU P2P gather bandwidth
    		9.4 GB/s
    GPU-to-GPU P2P latency
    		921 ns
    GPU-to-GPU P2P bandwidth
    		10.1 GB/s
    Single GPU slot
    Per-GPU GPU-to-host bandwidth
    		13.2 GB/s
    Per-GPU host-to-GPU bandwidth
    		12.4 GB/s
    Per-GPU host-device latency
    		921 ns
    Benchmark Environment
    Driver Version
    		 535.183.01
    CUDA Version (driver)
    		 12.6
    nvbandwidth Version
    		 v0.5
    GPU Count
    		 4
    P2P Supported
    Yes
    Efficiency
    SM/CE copy efficiency
    		1.00

    Memory Bandwidth

    Compression

    OpenSSL

    Geekbench Single-Core

    Score: 1,051

    Geekbench Multi-Core

    Score: 10,460

    Passmark CPU Scores

    BENCHMARKSCORE
    Mark
    		 31,607
    Compression
    		 548,320
    Encryption
    		 16,611
    Extended Instructions
    		 34,890
    Floating Point Maths
    		 67,905
    Integer Maths
    		 126,890
    Physics
    		 2,474
    Prime Numbers
    		 166
    Single Threaded
    		 1,703
    String Sorting
    		 61,440

    Passmark Memory Scores

    BENCHMARKSCORE
    Memory Mark
    		 2,596
    Database Operations
    		 15,027
    Memory Latency
    		 52
    Memory Read Cached
    		 20,570
    Memory Read Uncached
    		 10,704
    Memory Write
    		 10,150

    Stress-ng Stressors

    CPU integerSingleMultiScaling
    atomic
    		 189.34 156.58 0.83×
    bitops
    		 92,330 2,539,585 27.51×
    branch
    		 131,327 4,770,541 36.33×
    funccall
    		 462.78 11,400 24.63×
    funcret
    		 9,687 328,859 33.95×
    hash
    		 107,506 3,689,941 34.32×
    intmath
    		 19,070 681,208 35.72×
    nop
    		 163,126 3,377,625 20.71×
    prime
    		 271,613 7,283,461 26.82×
    rotate
    		 82,300 2,715,079 32.99×
    str
    		 626,393 15,326,129 24.47×
    wcs
    		 402,600 10,168,603 25.26×
    CryptographySingleMultiScaling
    crypt
    		 2,082 48,579 23.33×
    CompressionSingleMultiScaling
    jpeg
    		 499.03 14,122 28.30×
    zlib
    		 42.55 1,566 36.80×
    HypervisorSingleMultiScaling
    clock
    		 24,266 323,230 13.32×
    clone
    		 295.5 5,262 17.81×
    context
    		 2,247 60,130 26.77×
    cpu-sched
    		 60.49 3,162 52.27×
    eventfd
    		 46,517 980,409 21.08×
    fault
    		 9,408 225,994 24.02×
    fork
    		 3,505 69,014 19.69×
    getrandom
    		 38,205 1,114,859 29.18×
    hrtimers
    		 374,306 4,807,367 12.84×
    mprotect
    		 1,052,845 6,596,894 6.27×
    mtx
    		 7,764,587 1,064,380,583 137.08×
    munmap
    		 2,330 55,445 23.79×
    null
    		 254,796 5,892,863 23.13×
    pipe
    		 832,064 13,531,967 16.26×
    priv-instr
    		 484,347 11,404,382 23.55×
    resched
    		 5,402,439 1,991,627 0.37×
    switch
    		 76,059 6,985,551 91.84×
    timermix
    		 165,870 5,739,022 34.60×
    tlb-shootdown
    		 10,426 63,161 6.06×
    tsc
    		 850,206 28,375,415 33.37×
    usersyscall
    		 333,347 8,657,147 25.97×
    vdso
    		 70,567,877 2,373,605,730 33.64×
    vfork
    		 23,292 213,245 9.16×
    x86syscall
    		 2,896,113 66,797,266 23.06×
    yield
    		 20,972,641 26,402,507 1.26×
    zero
    		 1,731,912 43,852,394 25.32×
    CPU float/SIMDSingleMultiScaling
    eigen
    		 13,765 397,994 28.91×
    expmath
    		 2,292 83,613 36.49×
    fma
    		 510,469 19,889,493 38.96×
    fp
    		 104.72 3,165 30.22×
    fractal
    		 4.2 147.5 35.12×
    hyperbolic
    		 2,375 89,247 37.58×
    logmath
    		 3,862 136,352 35.30×
    matrix
    		 2,594 100,243 38.65×
    matrix-3d
    		 606.75 3,184 5.25×
    monte-carlo
    		 118.61 3,090 26.06×
    mpfr
    		 1,031 26,519 25.73×
    powmath
    		 1,351 49,439 36.60×
    trig
    		 1,459 45,449 31.15×
    veccmp
    		 13,274 329,205 24.80×
    vecfp
    		 1,153 34,043 29.52×
    vecmath
    		 1,863 70,470 37.83×
    vecshuf
    		 481.18 12,359 25.69×
    vecwide
    		 22,310 572,471 25.66×
    AlgorithmsSingleMultiScaling
    bitonicsort
    		 6.2 169.39 27.32×
    bsearch
    		 189.34 5,639 29.78×
    bubblesort
    		 0.6 9.6 16.00×
    fibsearch
    		 208.22 5,721 27.47×
    heapsort
    		 7.4 211.19 28.54×
    hsearch
    		 4,282 124,450 29.06×
    insertionsort
    		 7 163.2 23.31×
    malloc
    		 2,030,632 47,692,288 23.49×
    mcontend
    		 103.75 9,363 90.24×
    mergesort
    		 19.8 608.38 30.73×
    qsort
    		 12.6 345.79 27.44×
    skiplist
    		 4,147 140,930 33.98×
    tree
    		 1.2 19.56 16.30×
    MemorySingleMultiScaling
    cache
    		 361,944 788,779 2.18×
    cachehammer
    		 371,335 12,956,345 34.89×
    cacheline
    		 37.94 2,157 56.85×
    l1cache
    		 188.57 4,963 26.32×
    memcpy
    		 296.3 7,550 25.48×
    memrate
    		 1.17 8.59 7.34×
    memthrash
    		 201.05 893.75 4.45×
    ptr-chase
    		 30,139,625 426,244,991 14.14×
    spinmem
    		 190.79 1,784 9.35×
    stream
    		 29.98 12,331 411.31×
    vm
    		 52,703 300,508 5.70×

    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-24 2496 GiB2
    g2-standard-32 32128 GiB1
    g2-standard-96 96384 GiB8

    Similar Servers

    INSTANCEVENDORvCPUsMEMORYAccelerators

    g2-standard-48 FAQs