GPU-32xCPU-384GB-2xL4 by UpCloud

GPU-32xCPU-384GB-2xL4 is a GPU (32 vCPUs, 384 GiB RAM, 2x L4) server offered by UpCloud with 32 vCPUs, 384 GiB of memory and 0 GB of storage. The last known price was 1.8245 USD per hour.
32 vCPU
384 GiB Memory
2 GPU
Spare SCore
112,877
(All-cores)
3,831
(Single-core)

Specifications

Server Metadata

Vendor ID
upcloud
Name
GPU-32xCPU-384GB-2xL4
Description
GPU (32 vCPUs, 384 GiB RAM, 2x L4)
Family
GPU
Hw Virt
Average Time To Start
64
Status
inactive
Observed At
2026-10-26T04:15:07.620267

Availability

REGION / IDSPOTONDEMAND
Sydney #1 (AU) / au-syd1
1.8245 USD/h 1.8355 USD/h
Frankfurt #1 (DE) / de-fra1
1.8245 USD/h 1.8355 USD/h
Copenhagen #1 (DK) / dk-cph1
1.8245 USD/h 1.8355 USD/h
Madrid #1 (ES) / es-mad1
1.8245 USD/h 1.8355 USD/h
Helsinki #1 (FI) / fi-hel1
1.8245 USD/h 1.8355 USD/h
Helsinki #2 (FI) / fi-hel2
1.8245 USD/h 1.8355 USD/h
Amsterdam #1 (NL) / nl-ams1
1.8245 USD/h 1.8355 USD/h
Stavanger #1 (NO) / no-svg1
1.8245 USD/h 1.8355 USD/h
Warsaw #1 (PL) / pl-waw1
1.8245 USD/h 1.8355 USD/h
Stockholm #1 (SE) / se-sto1
1.8245 USD/h 1.8355 USD/h
Singapore #1 (SG) / sg-sin1
1.8245 USD/h 1.8355 USD/h
London #1 (GB) / uk-lon1
1.8245 USD/h 1.8355 USD/h
Chicago #1 (US) / us-chi1
1.8245 USD/h 1.8355 USD/h
New York #1 (US) / us-nyc1
1.8245 USD/h 1.8355 USD/h
San Jose #1 (US) / us-sjo1
1.8245 USD/h 1.8355 USD/h

Processor

vCPUs
32
Hypervisor
KVM
CPU Allocation
Shared
CPU Cores
32
CPU Speed
2 GHz
CPU Architecture
x86_64
CPU Manufacturer
AMD
CPU Family
EPYC
CPU Model
9575F
CPU L1D Cache
64 KiB
CPU L1D Cache Total
2 MiB
CPU L1I Cache
64 KiB
CPU L1I Cache Total
2 MiB
CPU L2 Cache
512 KiB
CPU L2 Cache Total
16 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, ht, syscall, nx, mmxext, fxsr_opt, pdpe1gb, rdtscp, lm, rep_good, nopl, cpuid, extd_apicid, tsc_known_freq, pni, pclmulqdq, ssse3, fma, cx16, pcid, sse4_1, sse4_2, x2apic, movbe, popcnt, tsc_deadline_timer, aes, xsave, avx, f16c, rdrand, hypervisor, lahf_lm, cmp_legacy, cr8_legacy, abm, sse4a, misalignsse, 3dnowprefetch, osvw, perfctr_core, ssbd, ibrs, ibpb, stibp, vmmcall, fsgsbase, tsc_adjust, bmi1, avx2, smep, bmi2, invpcid, avx512f, avx512dq, rdseed, adx, smap, avx512ifma, clflushopt, clwb, avx512cd, sha_ni, avx512bw, avx512vl, xsaveopt, xsavec, xgetbv1, xsaves, avx_vnni, avx512_bf16, clzero, xsaveerptr, wbnoinvd, arat, avx512vbmi, umip, pku, ospke, avx512_vbmi2, gfni, vaes, vpclmulqdq, avx512_vnni, avx512_bitalg, avx512_vpopcntdq, la57, rdpid, movdiri, movdir64b, avx512_vp2intersect, arch_capabilities
Ecpus
29.5
Scalability
92.19

System Resources and Accelerators

MEMORY
Memory Amount
384 GiB
Memory Amount Actual
384 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: 595.84, BIOS version: 95.04.65.00.13, Clock rate: 2040)
  • NVIDIA Ada Lovelace NVIDIA L4 (Memory amount: 23034, Firmware version: 595.84, BIOS version: 95.04.65.00.13, Clock rate: 2040)
STORAGE
Storage Size
0 GB
Storages
    NETWORK
    Inbound Traffic
    0 GB/month
    Outbound Traffic
    45056 GB/month
    IPv4
    1

    CPU and System Topology

    Server Description

    A GPU-accelerated server featuring dual NVIDIA L4 graphics processors, 384 GB of memory, and 32 AMD EPYC processor cores.

    GPU AcceleratedMemory OptimizedCompute Optimized

    UpCloud GPU-32xCPU-384GB-2xL4 is a GPU-accelerated server built on the KVM hypervisor, featuring an AMD EPYC 9575F processor with 32 vCPUs and 32 physical cores. It is equipped with 384.0 GB of system memory and two NVIDIA L4 GPUs from the Ada Lovelace architecture, delivering 48 GB of total VRAM. The server does not include local storage, requiring external storage provisioning. In benchmark testing, the instance demonstrates top-tier single-core and multi-core CPU performance, alongside top-tier LLM inference capabilities, processing 2805.36 tokens per second on a 7B model. However, its Passmark memory score remains in the average tier. This hardware configuration is suited for machine learning inference, data science, and GPU-accelerated compute workloads.

    Economics

    Average Price per Region

    Prices per Zone

    Lowest Prices

    Performance

    Workload Profiles

    3.1Score
    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) 8,261,090 ops/sec 1,919,289 ops/sec 4.4 50.00% 50.00% +110.00%
    Redis RPS (pipeline=16, SET) 43,569,351 ops/sec 11,728,058 ops/sec 3.7 20.00% 20.00% +30.00%
    PassMark Memory Mark (composite) 2,334 2,416 0.96 10.00% 10.00% -0.41%
    Memory bandwidth (read, 16 MB ~ L3) 244,809 MB/sec 108,466 MB/sec 2.2 10.00% 10.00% +8.20%
    PassMark single-thread CPU 3,745 Mops/s 2,365 Mops/s 1.5 10.00% 10.00% +4.10%

    GPU Bandwidth and Latency

    Node (all GPUs)
    Node baseline directional bandwidth
    		45.3 GB/s
    Node GPU-to-host bandwidth
    		52.7 GB/s
    Node host-to-GPU bandwidth
    		53.7 GB/s
    Single GPU slot
    Per-GPU GPU-to-host bandwidth
    		26.4 GB/s
    Per-GPU host-to-GPU bandwidth
    		26.9 GB/s
    Per-GPU host-device latency
    		590 ns
    Benchmark Environment
    Driver Version
    		 595.84
    CUDA Runtime Version
    		 13.2
    nvbandwidth Version
    		 v0.8
    GPU Count
    		 2
    P2P Supported
    No
    Efficiency
    SM/CE copy efficiency
    		1.00

    Memory Bandwidth

    Compression

    OpenSSL

    Geekbench Single-Core

    Score: 2,514

    Geekbench Multi-Core

    Score: 22,510

    Passmark CPU Scores

    BENCHMARKSCORE
    Mark
    		 71,253
    Compression
    		 865,613
    Encryption
    		 46,935
    Extended Instructions
    		 76,007
    Floating Point Maths
    		 211,671
    Integer Maths
    		 229,981
    Physics
    		 9,980
    Prime Numbers
    		 667
    Single Threaded
    		 3,745
    String Sorting
    		 115,624

    Passmark Memory Scores

    BENCHMARKSCORE
    Memory Mark
    		 2,334
    Database Operations
    		 30,935
    Memory Latency
    		 109
    Memory Read Cached
    		 32,999
    Memory Read Uncached
    		 31,068
    Memory Write
    		 31,141

    Stress-ng Stressors

    CPU integerSingleMultiScaling
    atomic
    		 478.46 396.35 0.83×
    bitops
    		 222,866 6,035,457 27.08×
    branch
    		 175,432 5,141,863 29.31×
    funccall
    		 828.5 22,871 27.61×
    funcret
    		 27,158 710,841 26.17×
    hash
    		 213,611 6,136,395 28.73×
    intmath
    		 80,683 2,324,498 28.81×
    nop
    		 397,525 11,748,806 29.55×
    prime
    		 588,929 16,011,348 27.19×
    rotate
    		 138,277 3,702,722 26.78×
    str
    		 1,556,503 37,878,023 24.34×
    wcs
    		 968,885 23,519,846 24.28×
    CryptographySingleMultiScaling
    crypt
    		 5,094 128,117 25.15×
    CompressionSingleMultiScaling
    jpeg
    		 1,182 32,274 27.29×
    zlib
    		 83.15 2,187 26.30×
    HypervisorSingleMultiScaling
    clock
    		 76,317 254,457 3.33×
    clone
    		 382.01 9,403 24.62×
    context
    		 8,650 255,876 29.58×
    cpu-sched
    		 69.29 2,096 30.24×
    eventfd
    		 82,196 1,440,070 17.52×
    fault
    		 33,621 177,183 5.27×
    fork
    		 5,512 76,172 13.82×
    getrandom
    		 91,359 2,238,597 24.50×
    hrtimers
    		 723,457 6,257,721 8.65×
    mprotect
    		 708,746 12,674,582 17.88×
    mtx
    		 36,564,130 2,506,452,254 68.55×
    munmap
    		 3,737 62,942 16.84×
    null
    		 1,387,118 40,139,944 28.94×
    pipe
    		 1,102,090 18,668,198 16.94×
    priv-instr
    		 1,351,476 16,968,093 12.56×
    resched
    		 13,901,747 14,557,071 1.05×
    switch
    		 150,240 12,411,493 82.61×
    timermix
    		 331,754 8,447,887 25.46×
    tlb-shootdown
    		 22,389 190,758 8.52×
    tsc
    		 910,628 28,494,605 31.29×
    usersyscall
    		 1,051,611 14,771,970 14.05×
    vdso
    		 98,690,695 2,858,003,114 28.96×
    vfork
    		 37,696 167,587 4.45×
    x86syscall
    		 17,852,968 536,268,523 30.04×
    yield
    		 40,835,142 53,175,894 1.30×
    zero
    		 7,750,397 213,466,396 27.54×
    CPU float/SIMDSingleMultiScaling
    eigen
    		 17,437 480,629 27.56×
    expmath
    		 3,952 106,016 26.83×
    fma
    		 1,531,188 40,655,272 26.55×
    fp
    		 195.85 5,954 30.40×
    fractal
    		 9 251.44 27.94×
    hyperbolic
    		 5,254 152,314 28.99×
    logmath
    		 5,344 158,023 29.57×
    matrix
    		 8,759 192,810 22.01×
    matrix-3d
    		 1,322 8,100 6.13×
    monte-carlo
    		 462.69 12,631 27.30×
    mpfr
    		 2,312 59,454 25.72×
    powmath
    		 1,780 52,534 29.51×
    trig
    		 2,311 66,103 28.60×
    veccmp
    		 28,996 768,392 26.50×
    vecfp
    		 2,456 70,723 28.80×
    vecmath
    		 4,704 133,445 28.37×
    vecshuf
    		 972.13 25,554 26.29×
    vecwide
    		 39,393 1,037,301 26.33×
    AlgorithmsSingleMultiScaling
    bitonicsort
    		 12 317.79 26.48×
    bsearch
    		 356.93 9,996 28.00×
    bubblesort
    		 1.4 34.2 24.43×
    fibsearch
    		 417.36 10,708 25.66×
    heapsort
    		 14.6 373.59 25.59×
    hsearch
    		 12,505 305,295 24.41×
    insertionsort
    		 15.2 357.4 23.51×
    malloc
    		 3,703,033 53,889,945 14.55×
    mcontend
    		 119.74 5,253 43.87×
    mergesort
    		 40.2 1,073 26.70×
    qsort
    		 26.8 709.99 26.49×
    skiplist
    		 8,143 224,535 27.57×
    tree
    		 2 44.75 22.38×
    MemorySingleMultiScaling
    cache
    		 1,076,880 1,891,130 1.76×
    cachehammer
    		 262,535 7,414,065 28.24×
    cacheline
    		 56.78 2,836 49.94×
    l1cache
    		 675.63 15,765 23.33×
    memcpy
    		 740.3 17,835 24.09×
    memrate
    		 1.07 13.47 12.59×
    memthrash
    		 180.27 1,320 7.32×
    ptr-chase
    		 23,678,300 257,586,164 10.88×
    spinmem
    		 407.78 3,120 7.65×
    stream
    		 420.82 49,491 117.61×
    vm
    		 235,720 296,067 1.26×

    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

    PostgreSQL heavy read-only throughput

    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-80GB-1xRTXPRO6000 1680 GiB1
    GPU-16xCPU-192GB-1xL40S 16192 GiB1
    GPU-16xCPU-192GB-2xL40S 16192 GiB2
    GPU-16xCPU-192GB-3xL40S 16192 GiB3

    Similar Servers

    INSTANCEVENDORvCPUsMEMORYAccelerators

    GPU-32xCPU-384GB-2xL4 FAQs