a2-highgpu-1g by Google Cloud Platform

a2-highgpu-1g is a Accelerator Optimized: 1 NVIDIA Tesla A100 GPU, 12 vCPUs, 85GB RAM server offered by Google Cloud Platform with 12 vCPUs, 85 GiB of memory and 0 GB of storage. The pricing starts at 1.7397 USD per hour.
12 vCPU
85 GiB Memory
1 GPU
Spare SCore
9,483
(All-cores)
1,199
(Single-core)

Specifications

Server Metadata

Vendor ID
gcp
Server ID
1000012
Name
a2-highgpu-1g
Description
Accelerator Optimized: 1 NVIDIA Tesla A100 GPU, 12 vCPUs, 85GB RAM
Family
a2
Hw Virt
Average Time To Start
26
Status
active
Observed At
2026-11-05T01:32:27.791202

Availability

REGION / IDSPOTONDEMAND
Council Bluffs (US) / us-central1
1.9891 USD/h 3.6734 USD/h
The Dalles (US) / us-west1
1.9891 USD/h 3.6734 USD/h
Moncks Corner (US) / us-east1
1.9891 USD/h 3.6734 USD/h
Eemshaven (NL) / europe-west4
1.7397 USD/h 3.748 USD/h
Las Vegas (US) / us-west4
1.8265 USD/h 3.9344 USD/h
Jurong West (SG) / asia-southeast1
2.4023 USD/h 4.0138 USD/h
Tel Aviv (IL) / me-west1
2.4244 USD/h 4.0407 USD/h
Seoul (KR) / asia-northeast3
2.412 USD/h 4.0496 USD/h
Tokyo (JP) / asia-northeast1
2.4175 USD/h 4.0496 USD/h
Salt Lake City (US) / us-west3
2.3304 USD/h 4.3251 USD/h

Processor

vCPUs
12
CPU Allocation
Dedicated
CPU Cores
6
CPU Speed
2.2 GHz
CPU Architecture
x86_64
CPU Manufacturer
Intel
CPU Family
Xeon
CPU L1D Cache
32 KiB
CPU L1D Cache Total
192 KiB
CPU L1I Cache
32 KiB
CPU L1I Cache Total
192 KiB
CPU L2 Cache
1 MiB
CPU L2 Cache Total
6 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
7.9
Scalability
65.83

System Resources and Accelerators

MEMORY
Memory Amount
85 GiB
Memory Amount Actual
85 GiB
ACCELERATORS
Accelerator Count
1
Accelerator Memory Min
40 GiB
Accelerator Memory Total
40 GiB
Accelerator Type
GPU
Accelerator Manufacturer
NVIDIA
Accelerator Family
Ampere
Accelerator Model
A100
Accelerators
  • NVIDIA Ampere NVIDIA A100-SXM4-40GB (Memory amount: 40960, Firmware version: 590.48.01, BIOS version: 92.00.45.00.03, Clock rate: 1410)
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

    An accelerator-optimized server featuring an NVIDIA A100 GPU and 85GB RAM, designed for high-efficiency machine learning and parallel computing workloads.

    GPU AcceleratedMemory Optimized

    Google Cloud Platform a2-highgpu-1g is an accelerator-optimized server designed for GPU-intensive workloads. It features an Intel Xeon x86_64 processor with 12 dedicated vCPUs, 6 physical cores, and 85.0 GB of system memory. The primary hardware highlight is a single NVIDIA Ampere A100 GPU with 40 GB of VRAM. No local storage is bundled. While single-core CPU benchmarks and general text processing show weak performance, the server delivers top-tier results in LLM inference for small and medium models. Multi-core performance, memory bandwidth, and database operations remain in the average tier. This server is qualitatively cost-effective for workloads that heavily leverage GPU acceleration, such as machine learning inference, data science, and parallel computing, whereas standard CPU-bound tasks will underutilize its specialized hardware.

    Economics

    Average Price per Region

    Prices per Zone

    Lowest Prices

    Performance

    Workload Profiles

    0.55Score
    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) 921,933 ops/sec 1,933,252 ops/sec 0.48 50.00% 50.00% -31.00%
    Redis RPS (pipeline=16, SET) 5,646,966 ops/sec 11,794,699 ops/sec 0.47 20.00% 20.00% -14.00%
    PassMark Memory Mark (composite) 2,333 2,410 0.96 10.00% 10.00% -0.41%
    Memory bandwidth (read, 16 MB ~ L3) 75,589 MB/sec 108,490 MB/sec 0.69 10.00% 10.00% -3.60%
    PassMark single-thread CPU 1,698 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
    		5.3 GB/s
    Node GPU-to-host bandwidth
    		13.2 GB/s
    Node host-to-GPU bandwidth
    		12.3 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
    		1,239 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: 1,060

    Geekbench Multi-Core

    Score: 5,399

    Passmark CPU Scores

    BENCHMARKSCORE
    Mark
    		 10,529
    Compression
    		 141,009
    Encryption
    		 4,354
    Extended Instructions
    		 8,926
    Floating Point Maths
    		 17,048
    Integer Maths
    		 31,697
    Physics
    		 1,217
    Prime Numbers
    		 64
    Single Threaded
    		 1,698
    String Sorting
    		 16,898

    Passmark Memory Scores

    BENCHMARKSCORE
    Memory Mark
    		 2,333
    Database Operations
    		 3,886
    Memory Latency
    		 55
    Memory Read Cached
    		 20,650
    Memory Read Uncached
    		 9,599
    Memory Write
    		 9,200

    Stress-ng Stressors

    CPU integerSingleMultiScaling
    atomic
    		 193.98 180.96 0.93×
    bitops
    		 92,188 634,021 6.88×
    branch
    		 131,172 1,190,959 9.08×
    funccall
    		 462.57 2,840 6.14×
    funcret
    		 9,627 82,509 8.57×
    hash
    		 107,339 926,734 8.63×
    intmath
    		 19,080 170,013 8.91×
    nop
    		 162,683 844,490 5.19×
    prime
    		 271,507 1,818,363 6.70×
    rotate
    		 82,348 680,316 8.26×
    str
    		 632,002 3,973,369 6.29×
    wcs
    		 403,151 2,575,713 6.39×
    CryptographySingleMultiScaling
    crypt
    		 2,080 12,137 5.84×
    CompressionSingleMultiScaling
    jpeg
    		 513.81 3,638 7.08×
    zlib
    		 49.05 382.43 7.80×
    HypervisorSingleMultiScaling
    clock
    		 26,789 158,876 5.93×
    clone
    		 297.03 1,559 5.25×
    context
    		 2,246 15,656 6.97×
    cpu-sched
    		 77.44 785.93 10.15×
    eventfd
    		 46,268 307,302 6.64×
    fault
    		 9,244 70,395 7.62×
    fork
    		 3,619 24,360 6.73×
    getrandom
    		 38,145 278,558 7.30×
    hrtimers
    		 369,344 1,031,467 2.79×
    mprotect
    		 883,826 1,659,925 1.88×
    mtx
    		 8,219,749 299,783,439 36.47×
    munmap
    		 2,260 18,153 8.03×
    null
    		 253,563 1,566,740 6.18×
    pipe
    		 824,890 3,747,824 4.54×
    priv-instr
    		 482,504 3,274,408 6.79×
    resched
    		 2,008,788 530,750 0.26×
    switch
    		 74,360 1,759,879 23.67×
    timermix
    		 161,337 1,436,419 8.90×
    tlb-shootdown
    		 10,058 22,465 2.23×
    tsc
    		 849,412 7,093,413 8.35×
    usersyscall
    		 331,969 2,223,308 6.70×
    vdso
    		 70,506,522 596,099,050 8.45×
    vfork
    		 25,362 133,686 5.27×
    x86syscall
    		 2,905,810 17,921,514 6.17×
    yield
    		 5,769,255 5,724,241 0.99×
    zero
    		 1,727,191 11,183,406 6.47×
    CPU float/SIMDSingleMultiScaling
    eigen
    		 13,757 100,190 7.28×
    expmath
    		 2,265 20,873 9.22×
    fma
    		 510,332 4,970,993 9.74×
    fp
    		 100.8 785.99 7.80×
    fractal
    		 4.2 37 8.81×
    hyperbolic
    		 2,389 22,303 9.34×
    logmath
    		 3,859 34,103 8.84×
    matrix
    		 2,600 25,087 9.65×
    matrix-3d
    		 570.31 2,522 4.42×
    monte-carlo
    		 118.41 826.19 6.98×
    mpfr
    		 1,030 6,624 6.43×
    powmath
    		 1,352 12,339 9.13×
    trig
    		 1,457 11,562 7.93×
    veccmp
    		 13,245 82,713 6.25×
    vecfp
    		 1,150 8,557 7.44×
    vecmath
    		 1,860 17,594 9.46×
    vecshuf
    		 480.48 3,089 6.43×
    vecwide
    		 21,938 143,094 6.52×
    AlgorithmsSingleMultiScaling
    bitonicsort
    		 6.2 43 6.94×
    bsearch
    		 203.77 1,397 6.86×
    bubblesort
    		 0.6 2.4 4.00×
    fibsearch
    		 208.04 1,428 6.87×
    heapsort
    		 7.4 53 7.16×
    hsearch
    		 4,259 32,276 7.58×
    insertionsort
    		 7 40.8 5.83×
    malloc
    		 1,882,006 18,719,033 9.95×
    mcontend
    		 126.91 2,346 18.49×
    mergesort
    		 19.8 159.99 8.08×
    qsort
    		 12.6 86.6 6.87×
    skiplist
    		 4,153 35,239 8.49×
    tree
    		 1.2 7.2 6.00×
    MemorySingleMultiScaling
    cache
    		 303,105 504,729 1.67×
    cachehammer
    		 361,917 3,807,717 10.52×
    cacheline
    		 31.51 363.62 11.54×
    l1cache
    		 185.72 1,263 6.80×
    memcpy
    		 296.01 1,903 6.43×
    memrate
    		 1.17 6.87 5.87×
    memthrash
    		 143.19 417.93 2.92×
    ptr-chase
    		 15,359,965 122,643,209 7.98×
    spinmem
    		 194.19 447.94 2.31×
    stream
    		 25.83 2,304 89.19×
    vm
    		 52,494 292,349 5.57×

    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
    a2-ultragpu-1g 12170 GiB1
    a2-highgpu-2g 24170 GiB2
    a2-ultragpu-2g 24340 GiB2
    a2-highgpu-4g 48340 GiB4
    a2-ultragpu-4g 48680 GiB4
    a2-highgpu-8g 96680 GiB8
    a2-ultragpu-8g 961360 GiB8

    Similar Servers

    INSTANCEVENDORvCPUsMEMORYAccelerators

    a2-highgpu-1g FAQs