GPU-48xCPU-960GB-4xH100 by UpCloud

GPU-48xCPU-960GB-4xH100 is a GPU (48 vCPUs, 960 GiB RAM, 4x H100) server offered by UpCloud with 48 vCPUs, 960 GiB of memory and 0 GB of storage. The last known price was 7.8421 USD per hour.
48 vCPU
960 GiB Memory
4 GPU
Spare SCore
96,294
(All-cores)
2,196
(Single-core)

Specifications #

Server Metadata #

Vendor ID
upcloud
Name
GPU-48xCPU-960GB-4xH100
Description
GPU (48 vCPUs, 960 GiB RAM, 4x H100)
Family
GPU
Series
H100
Hw Virt
Status
inactive
Observed At
2026-11-18T15:26:59.372097

Availability #

Region / IDSpotOndemand
Sydney #1 (AU) / au-syd1
7.8421 USD/h
7.8539 USD/h
Frankfurt #1 (DE) / de-fra1
7.8421 USD/h
7.8539 USD/h
Copenhagen #1 (DK) / dk-cph1
7.8421 USD/h
7.8539 USD/h
Madrid #1 (ES) / es-mad1
7.8421 USD/h
7.8539 USD/h
Helsinki #1 (FI) / fi-hel1
7.8421 USD/h
7.8539 USD/h
Helsinki #2 (FI) / fi-hel2
7.8421 USD/h
7.8539 USD/h
Amsterdam #1 (NL) / nl-ams1
7.8421 USD/h
7.8539 USD/h
Stavanger #1 (NO) / no-svg1
7.8421 USD/h
7.8539 USD/h
Warsaw #1 (PL) / pl-waw1
7.8421 USD/h
7.8539 USD/h
Stockholm #1 (SE) / se-sto1
7.8421 USD/h
7.8539 USD/h
Singapore #1 (SG) / sg-sin1
7.8421 USD/h
7.8539 USD/h
London #1 (GB) / uk-lon1
7.8421 USD/h
7.8539 USD/h
Chicago #1 (US) / us-chi1
7.8421 USD/h
7.8539 USD/h
New York #1 (US) / us-nyc1
7.8421 USD/h
7.8539 USD/h
San Jose #1 (US) / us-sjo1
7.8421 USD/h
7.8539 USD/h

Processor #

vCPUs
48
Hypervisor
KVM
CPU Allocation
Shared
CPU Cores
48
CPU Speed
2 GHz
CPU Architecture
x86_64
CPU Manufacturer
Intel
CPU Family
Xeon
CPU Model
8462Y+
CPU L1D Cache
32 KiB
CPU L1D Cache Total
2 MiB
CPU L1I Cache
32 KiB
CPU L1I Cache Total
2 MiB
CPU L2 Cache
4 MiB
CPU L2 Cache Total
192 MiB
CPU L3 Cache
16 MiB
CPU L3 Cache Total
16 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, cpuid, tsc_known_freq, pni, pclmulqdq, ssse3, fma, cx16, pdcm, pcid, sse4_1, sse4_2, x2apic, movbe, popcnt, tsc_deadline_timer, aes, xsave, avx, f16c, rdrand, hypervisor, lahf_lm, abm, 3dnowprefetch, cpuid_fault, ssbd, ibrs, ibpb, stibp, 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, avx_vnni, avx512_bf16, wbnoinvd, arat, avx512vbmi, umip, pku, ospke, waitpkg, avx512_vbmi2, gfni, vaes, vpclmulqdq, avx512_vnni, avx512_bitalg, avx512_vpopcntdq, la57, rdpid, bus_lock_detect, cldemote, movdiri, movdir64b, fsrm, md_clear, serialize, tsxldtrk, amx_bf16, avx512_fp16, amx_tile, amx_int8, arch_capabilities
Ecpus
43.8
Scalability
91.25

System Resources and Accelerators #

Memory
Memory Amount
960 GiB
Memory Amount Actual
960 GiB
Accelerators
Accelerator Count
4
Accelerator Memory Min
80 GiB
Accelerator Memory Total
320 GiB
Accelerator Type
GPU
Accelerator Manufacturer
NVIDIA
Accelerator Family
Hopper
Accelerator Model
H100
Accelerators
  • NVIDIA Hopper NVIDIA H100 80GB HBM3 (Memory amount: 81559, Firmware version: 595.71.05, BIOS version: 96.00.DA.00.0C, Clock rate: 1980)
  • NVIDIA Hopper NVIDIA H100 80GB HBM3 (Memory amount: 81559, Firmware version: 595.71.05, BIOS version: 96.00.DA.00.0C, Clock rate: 1980)
  • NVIDIA Hopper NVIDIA H100 80GB HBM3 (Memory amount: 81559, Firmware version: 595.71.05, BIOS version: 96.00.DA.00.0C, Clock rate: 1980)
  • NVIDIA Hopper NVIDIA H100 80GB HBM3 (Memory amount: 81559, Firmware version: 595.71.05, BIOS version: 96.00.DA.00.0C, Clock rate: 1980)
Storage
Storage Size
0 GB
Storages
    Compatible Storage IDs
    maxiops, standard, hdd
    Network
    Inbound Traffic
    0 GB/month
    Outbound Traffic
    71680 GB/month
    IPv4
    1

    CPU and System Topology #

    Server Description #

    A GPU-accelerated infrastructure option featuring four NVIDIA H100 GPUs and 960 GB of RAM for intensive machine learning workloads.

    UpCloud GPU-48xCPU-960GB-4xH100 is a GPU-accelerated server utilizing the KVM hypervisor and x86_64 architecture. It features 48 physical cores from Intel Xeon 8462Y+ processors running at 2.0 GHz with a shared CPU allocation, paired with 960.0 GB of system memory. The primary acceleration is provided by four NVIDIA Hopper H100 GPUs with a total of 320 GB VRAM. No local storage is included. Benchmarks indicate top-tier memory bandwidth of 7649.82 GB/sec and strong multi-core CPU performance, with a PassMark CPU score of 85426.66. The instance also demonstrates top-tier LLM inference capabilities, processing 10308.98 tokens per second for prompt processing on a 135M model. This server is suited for large-scale machine learning, LLM inference, and data science workloads.

    Economics #

    Average Price per Region #

    Prices per Zone #

    Lowest Prices #

    Performance #

    Workload Profiles #

    4.6Score
    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.
    Component Score Weight Impact
    Raw Reference Normalized Target Actual
    Redis RPS (pipeline=1, SET) 14,950,797 ops/sec 1,933,252 ops/sec 7.9 50.00% 50.00% +180.00%
    Redis RPS (pipeline=16, SET) 75,252,071 ops/sec 11,814,424 ops/sec 6.2 20.00% 20.00% +44.00%
    PassMark Memory Mark (composite) 2,129 2,408 0.88 10.00% 10.00% -1.30%
    Memory bandwidth (read, 16 MB ~ L3) 351,343 MB/sec 108,674 MB/sec 3.2 10.00% 10.00% +12.00%
    PassMark single-thread CPU 3,275 Mops/s 2,365 Mops/s 1.4 10.00% 10.00% +3.40%

    GPU Bandwidth and Latency #

    Node (all GPUs)
    Node baseline directional bandwidth
    77.7 GB/s
    Node GPU-to-host bandwidth
    128.0 GB/s
    Node host-to-GPU bandwidth
    160.8 GB/s
    Peer-to-peer
    GPU-to-GPU P2P duplex bandwidth
    393.1 GB/s
    GPU-to-GPU P2P gather bandwidth
    399.2 GB/s
    GPU-to-GPU P2P latency
    828 ns
    GPU-to-GPU P2P bandwidth
    397.5 GB/s
    Single GPU slot
    Per-GPU GPU-to-host bandwidth
    55.2 GB/s
    Per-GPU host-to-GPU bandwidth
    55.3 GB/s
    Per-GPU host-device latency
    1,169 ns
    Per-GPU device memory copy bandwidth
    1490.5 GB/s
    Benchmark Environment
    Driver Version
    595.71.05
    CUDA Version (driver)
    13.2
    nvbandwidth Version
    v0.8
    GPU Count
    4
    P2P Supported
    Yes
    Efficiency
    SM/CE copy efficiency
    0.62

    Memory Bandwidth#

    Compression #

    OpenSSL #

    Passmark CPU Scores #

    BENCHMARKSCORE
    Mark
    85,427
    Compression
    1,188,329
    Encryption
    68,257
    Extended Instructions
    117,918
    Floating Point Maths
    321,361
    Integer Maths
    275,561
    Physics
    12,429
    Prime Numbers
    895
    Single Threaded
    3,275
    String Sorting
    163,786

    Passmark Memory Scores #

    BENCHMARKSCORE
    Memory Mark
    2,129
    Database Operations
    31,709
    Memory Latency
    85
    Memory Read Cached
    30,506
    Memory Read Uncached
    9,105
    Memory Write
    8,028

    Stress-ng Stressors #

    CPU integerSingleMultiScaling
    atomic
    237.09 225.07 0.95×
    bitops
    169,704 7,380,624 43.49×
    branch
    158,352 7,041,911 44.47×
    funccall
    856.3 37,600 43.91×
    funcret
    14,360 621,171 43.26×
    hash
    202,938 9,092,991 44.81×
    intmath
    72,696 3,219,850 44.29×
    nop
    300,996 13,039,523 43.32×
    prime
    532,549 23,052,277 43.29×
    rotate
    113,859 4,938,632 43.38×
    str
    965,499 45,822,817 47.46×
    wcs
    785,096 33,138,766 42.21×
    CryptographySingleMultiScaling
    crypt
    3,728 168,635 45.24×
    CompressionSingleMultiScaling
    jpeg
    928.72 41,773 44.98×
    zlib
    67.71 2,935 43.35×
    HypervisorSingleMultiScaling
    clock
    46,763 213,364 4.56×
    clone
    397.51 14,000 35.22×
    context
    5,228 231,102 44.20×
    cpu-sched
    61.4 2,979 48.52×
    eventfd
    68,288 1,461,614 21.40×
    fault
    30,907 164,780 5.33×
    fork
    4,282 77,094 18.00×
    getrandom
    67,981 2,970,552 43.70×
    hrtimers
    374,589 7,080,342 18.90×
    mprotect
    756,255 16,048,375 21.22×
    mtx
    17,990,400 3,095,244,222 172.05×
    munmap
    3,003 63,347 21.09×
    null
    817,327 35,371,300 43.28×
    pipe
    634,843 23,140,088 36.45×
    priv-instr
    847,305 9,935,250 11.73×
    resched
    12,598,600 15,718,550 1.25×
    switch
    111,228 15,355,689 138.06×
    timermix
    257,810 11,386,725 44.17×
    tlb-shootdown
    12,604 192,725 15.29×
    tsc
    662,214 28,807,954 43.50×
    usersyscall
    733,698 10,095,564 13.76×
    vdso
    90,095,870 3,924,633,857 43.56×
    vfork
    29,267 179,920 6.15×
    x86syscall
    9,072,898 402,050,713 44.31×
    yield
    52,942,856 70,985,079 1.34×
    zero
    5,115,316 223,068,064 43.61×
    CPU float/SIMDSingleMultiScaling
    eigen
    18,661 814,323 43.64×
    expmath
    4,032 176,741 43.83×
    fma
    1,579,073 70,197,724 44.46×
    fp
    350.4 15,784 45.04×
    fractal
    8 348.29 43.54×
    hyperbolic
    4,618 202,111 43.76×
    logmath
    5,903 264,727 44.85×
    matrix
    5,335 228,077 42.75×
    matrix-3d
    845.45 10,226 12.09×
    monte-carlo
    291.47 12,916 44.31×
    mpfr
    1,991 86,601 43.50×
    powmath
    2,218 95,994 43.28×
    trig
    3,026 127,918 42.27×
    veccmp
    29,481 1,281,228 43.46×
    vecfp
    2,740 113,539 41.44×
    vecmath
    4,468 198,005 44.31×
    vecshuf
    4,357 193,666 44.45×
    vecwide
    95,292 3,971,739 41.68×
    AlgorithmsSingleMultiScaling
    bitonicsort
    10 436.59 43.66×
    bsearch
    329.61 14,379 43.62×
    bubblesort
    1.2 48 40.00×
    fibsearch
    356.98 15,486 43.38×
    heapsort
    11.4 504.39 44.24×
    hsearch
    9,669 413,514 42.77×
    insertionsort
    14.4 634 44.03×
    malloc
    2,940,518 61,841,354 21.03×
    mcontend
    72.07 7,928 110.01×
    mergesort
    34.8 1,535 44.12×
    qsort
    25 1,060 42.41×
    skiplist
    6,238 271,077 43.46×
    tree
    2 61.91 30.95×
    MemorySingleMultiScaling
    cache
    766,374 1,498,963 1.96×
    cachehammer
    235,034 8,705,806 37.04×
    cacheline
    33.38 3,758 112.59×
    l1cache
    356.1 15,945 44.78×
    memcpy
    506.39 22,033 43.51×
    memrate
    0.12 5.77 48.08×
    memthrash
    226.68 1,660 7.32×
    ptr-chase
    36,688,482 420,105,516 11.45×
    spinmem
    248.58 2,637 10.61×
    stream
    154.21 96,912 628.44×
    vm
    143,054 298,485 2.09×

    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
    GPU-8xCPU-64GB-1xL40S 864 GiB1
    GPU-12xCPU-128GB-1xL40S 12128 GiB1
    GPU-12xCPU-128GB-2xL40S 12128 GiB2
    GPU-16xCPU-192GB-1xL40S 16192 GiB1
    GPU-16xCPU-192GB-2xL40S 16192 GiB2
    GPU-20xCPU-256GB-1xL40S 20256 GiB1
    GPU-20xCPU-256GB-2xL40S 20256 GiB2

    Similar Servers #

    INSTANCEVENDORvCPUsMEMORYAccelerators

    GPU-48xCPU-960GB-4xH100 FAQs #