ecs.gn7i-c32g1.32xlarge by Alibaba Cloud

ecs.gn7i-c32g1.32xlarge is a ecs.gn7i family (128 vCPUs, 752 GiB RAM, 0 GB storage, 4xA10 24576 GiB VRAM) server offered by Alibaba Cloud with 128 vCPUs, 752 GiB of memory and 0 GB of storage. The pricing starts at 1.9721 USD per hour.
128 vCPU
752 GiB Memory
4 GPU
Spare SCore
123,936
(All-cores)
1,883
(Single-core)

Specifications

Server Metadata

Vendor ID
alicloud
Name
ecs.gn7i-c32g1.32xlarge
Description
ecs.gn7i family (128 vCPUs, 752 GiB RAM, 0 GB storage, 4xA10 24576 GiB VRAM)
Family
ecs.gn7i
Hw Virt
Average Time To Start
23
Status
active
Observed At
2026-10-24T04:54:00.950908

Availability

REGION / IDSPOTONDEMAND
Heyuan (CN) / cn-heyuan
2.4792 USD/h 7.0833 USD/h
Guangzhou (CN) / cn-guangzhou
2.7546 USD/h 7.8703 USD/h
Seoul (KR) / ap-northeast-2
1.9721 USD/h 7.8882 USD/h
Virginia (US) / us-east-1
3.8612 USD/h 11.032 USD/h
Tokyo (JP) / ap-northeast-1
4.2903 USD/h 12.258 USD/h
Singapore (SG) / ap-southeast-1
4.312 USD/h 12.32 USD/h
Frankfurt (DE) / eu-central-1
4.7159 USD/h 13.474 USD/h

Processor

vCPUs
128
Hypervisor
KVM
CPU Allocation
Dedicated
CPU Cores
64
CPU Speed
2.9 GHz
CPU Architecture
x86_64
CPU Manufacturer
Intel
CPU Family
Xeon
CPU Model
8369B
CPU L1D Cache
48 KiB
CPU L1D Cache Total
3 MiB
CPU L1I Cache
32 KiB
CPU L1I Cache Total
2 MiB
CPU L2 Cache
1 MiB
CPU L2 Cache Total
80 MiB
CPU L3 Cache
48 MiB
CPU L3 Cache Total
96 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, arch_perfmon, rep_good, nopl, xtopology, nonstop_tsc, cpuid, tsc_known_freq, pni, pclmulqdq, monitor, ssse3, fma, cx16, pcid, sse4_1, sse4_2, x2apic, movbe, popcnt, aes, xsave, avx, f16c, rdrand, hypervisor, lahf_lm, abm, 3dnowprefetch, cpuid_fault, ibrs_enhanced, fsgsbase, tsc_adjust, bmi1, avx2, smep, bmi2, erms, invpcid, avx512f, avx512dq, rdseed, adx, smap, avx512ifma, clflushopt, clwb, avx512cd, sha_ni, avx512bw, avx512vl, xsaveopt, xsavec, xgetbv1, xsaves, wbnoinvd, arat, avx512vbmi, pku, ospke, avx512_vbmi2, gfni, vaes, vpclmulqdq, avx512_vnni, avx512_bitalg, avx512_vpopcntdq, rdpid, fsrm, arch_capabilities
Ecpus
65.8
Scalability
51.41

System Resources and Accelerators

MEMORY
Memory Amount
752 GiB
Memory Amount Actual
764 GiB
ACCELERATORS
Accelerator Count
4
Accelerator Memory Min
24 GiB
Accelerator Memory Total
96 GiB
Accelerator Type
GPU
Accelerator Manufacturer
NVIDIA
Accelerator Family
Ampere
Accelerator Model
A10
Accelerators
  • NVIDIA Ampere NVIDIA A10 (Memory amount: 23028, Firmware version: 595.71.05, BIOS version: 94.02.91.00.02, Clock rate: 1695)
  • NVIDIA Ampere NVIDIA A10 (Memory amount: 23028, Firmware version: 595.71.05, BIOS version: 94.02.91.00.02, Clock rate: 1695)
  • NVIDIA Ampere NVIDIA A10 (Memory amount: 23028, Firmware version: 595.71.05, BIOS version: 94.02.91.00.02, Clock rate: 1695)
  • NVIDIA Ampere NVIDIA A10 (Memory amount: 23028, Firmware version: 595.71.05, BIOS version: 94.02.91.00.02, Clock rate: 1695)
STORAGE
Storage Size
0 GB
Storages
    NETWORK
    Network Speed Baseline
    64 Gbps
    Inbound Traffic
    0 GB/month
    Outbound Traffic
    0 GB/month
    IPv4
    0

    CPU and System Topology

    Server Description

    A high-density GPU server combining 128 dedicated vCPUs, 752 GB RAM, and four accelerators for parallel processing workloads.

    GPU AcceleratedCompute OptimizedMemory Optimized

    Alibaba Cloud ecs.gn7i-c32g1.32xlarge is a high-performance, GPU-accelerated server designed for parallel computing workloads. Powered by an Intel Xeon 8369B processor with 128 dedicated vCPUs and 752.0 GB of RAM, it integrates four NVIDIA A10 Ampere GPUs delivering a total of 96 GB of VRAM. The instance features a 64 Gbps baseline network bandwidth but lacks local storage. In benchmarks, it exhibits strong multi-core CPU performance and top-tier multi-threaded decompression capabilities, alongside high memory bandwidth. This resource-dense profile is well-suited for deep learning, video processing, and high-performance computing tasks.

    Economics

    Average Price per Region

    Prices per Zone

    Lowest Prices

    Performance

    Workload Profiles

    5.6Score
    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.Partial coverage: missing component benchmark(s): PassMark Memory Mark (composite), PassMark single-thread CPU.
    Component Score Weight Impact
    Raw Reference Normalized Target Actual
    Redis RPS (pipeline=1, SET) 12,523,657 ops/sec 1,919,289 ops/sec 6.8 50.00% 62.00% +230.00%
    Redis RPS (pipeline=16, SET) 61,161,081 ops/sec 11,728,058 ops/sec 5.1 20.00% 25.00% +50.00%
    PassMark Memory Mark (composite) — 2,416 — 10.00% 0.00% —
    Memory bandwidth (read, 16 MB ~ L3) 266,729 MB/sec 108,466 MB/sec 2.5 10.00% 13.00% +13.00%
    PassMark single-thread CPU — 2,365 Mops/s — 10.00% 0.00% —

    GPU Bandwidth and Latency

    Node (all GPUs)
    Node baseline directional bandwidth
    		36.6 GB/s
    Node GPU-to-host bandwidth
    		72.5 GB/s
    Node host-to-GPU bandwidth
    		95.3 GB/s
    Single GPU slot
    Per-GPU GPU-to-host bandwidth
    		26.2 GB/s
    Per-GPU host-to-GPU bandwidth
    		24.1 GB/s
    Per-GPU host-device latency
    		610 ns
    Benchmark Environment
    Driver Version
    		 595.71.05
    CUDA Runtime Version
    		 13.2
    nvbandwidth Version
    		 v0.8
    GPU Count
    		 4
    P2P Supported
    No
    Efficiency
    SM/CE copy efficiency
    		0.59

    Memory Bandwidth

    Compression

    OpenSSL

    Stress-ng Stressors

    CPU integerSingleMultiScaling
    atomic
    		 213.9 235.18 1.10×
    bitops
    		 129,673 8,471,440 65.33×
    branch
    		 137,482 13,952,389 101.49×
    funccall
    		 635.79 40,866 64.28×
    funcret
    		 11,798 1,031,268 87.41×
    hash
    		 150,936 13,861,517 91.84×
    intmath
    		 60,968 4,757,372 78.03×
    nop
    		 188,944 10,557,484 55.88×
    prime
    		 393,719 26,212,675 66.58×
    rotate
    		 97,889 6,649,555 67.93×
    str
    		 867,686 56,023,375 64.57×
    wcs
    		 606,369 31,771,233 52.40×
    CryptographySingleMultiScaling
    crypt
    		 2,842 166,250 58.49×
    CompressionSingleMultiScaling
    jpeg
    		 722.8 50,583 69.98×
    zlib
    		 51.16 5,550 108.48×
    HypervisorSingleMultiScaling
    clock
    		 20,562 172,261 8.38×
    clone
    		 565.75 11,162 19.73×
    context
    		 4,814 457,459 95.03×
    cpu-sched
    		 56.84 8,077 142.11×
    eventfd
    		 108,722 3,713,746 34.16×
    fault
    		 17,237 178,648 10.36×
    fork
    		 5,230 59,375 11.35×
    getrandom
    		 51,211 3,683,101 71.92×
    hrtimers
    		 122,065 1,847,259 15.13×
    mprotect
    		 641,225 19,182,998 29.92×
    mtx
    		 13,516,442 4,000,605,837 295.98×
    munmap
    		 3,775 74,830 19.82×
    null
    		 798,307 59,715,209 74.80×
    pipe
    		 1,058,186 45,527,627 43.02×
    priv-instr
    		 821,035 18,041,421 21.97×
    resched
    		 10,868,697 16,533,148 1.52×
    switch
    		 189,552 7,281,873 38.42×
    timermix
    		 343,231 36,451,915 106.20×
    tlb-shootdown
    		 11,344 134,137 11.82×
    tsc
    		 767,379 72,474,800 94.44×
    usersyscall
    		 637,556 17,918,519 28.11×
    vdso
    		 105,724,259 7,219,263,580 68.28×
    vfork
    		 37,018 98,327 2.66×
    x86syscall
    		 8,873,306 767,681,397 86.52×
    yield
    		 95,445,291 116,370,374 1.22×
    zero
    		 4,606,646 342,739,609 74.40×
    CPU float/SIMDSingleMultiScaling
    eigen
    		 14,797 1,145,457 77.41×
    expmath
    		 2,882 258,502 89.71×
    fma
    		 660,839 72,030,519 109.00×
    fp
    		 192 12,999 67.70×
    fractal
    		 5 463.03 92.61×
    hyperbolic
    		 3,075 287,364 93.46×
    logmath
    		 4,854 443,399 91.34×
    matrix
    		 3,530 323,149 91.55×
    matrix-3d
    		 795.3 9,249 11.63×
    monte-carlo
    		 258.31 21,180 81.99×
    mpfr
    		 1,413 91,221 64.58×
    powmath
    		 1,688 153,022 90.67×
    trig
    		 1,964 150,213 76.47×
    veccmp
    		 20,326 1,199,293 59.00×
    vecfp
    		 1,529 118,443 77.46×
    vecmath
    		 3,231 285,112 88.24×
    vecshuf
    		 665.65 45,659 68.59×
    vecwide
    		 31,929 2,048,736 64.17×
    AlgorithmsSingleMultiScaling
    bitonicsort
    		 8 486.58 60.82×
    bsearch
    		 250.21 20,317 81.20×
    bubblesort
    		 1 51.2 51.20×
    fibsearch
    		 288.45 22,744 78.85×
    heapsort
    		 9.4 665.57 70.81×
    hsearch
    		 7,110 478,917 67.36×
    insertionsort
    		 10.4 384 36.92×
    malloc
    		 2,624,712 81,488,883 31.05×
    mcontend
    		 69.86 21,777 311.73×
    mergesort
    		 26.8 2,115 78.93×
    qsort
    		 20.8 1,300 62.48×
    skiplist
    		 5,460 503,966 92.30×
    tree
    		 1.6 76.72 47.95×
    MemorySingleMultiScaling
    cache
    		 251,627 2,695,353 10.71×
    cachehammer
    		 365,404 13,546,163 37.07×
    cacheline
    		 62.75 3,451 54.99×
    l1cache
    		 898.78 20,648 22.97×
    memcpy
    		 405.94 24,540 60.45×
    memrate
    		 1.77 28.85 16.30×
    memthrash
    		 402.69 833.47 2.07×
    ptr-chase
    		 33,529,765 983,722,345 29.34×
    spinmem
    		 353.99 8,767 24.77×
    stream
    		 21.03 36,181 1720.46×
    vm
    		 114,903 605,130 5.27×

    Stress-ng Raw Scores

    Stress-ng Relative Multicore Performance

    Static Web Server

    Redis

    Alternatives

    Servers of the Same Family

    INSTANCEvCPUsMEMORYAccelerators
    ecs.gn7i-c8g1.2xlarge 830 GiB1
    ecs.gn7i-c16g1.4xlarge 1660 GiB1
    ecs.gn7i-c32g1.8xlarge 32188 GiB1
    ecs.gn7i-2x.8xlarge 32128 GiB2
    ecs.gn7i-4x.8xlarge 32128 GiB4
    ecs.gn7i-c48g1.12xlarge 48310 GiB1
    ecs.gn7i-c56g1.14xlarge 56346 GiB1

    Similar Servers

    INSTANCEVENDORvCPUsMEMORYAccelerators

    ecs.gn7i-c32g1.32xlarge FAQs