GPU-16xCPU-192GB-1xL40S by UpCloud

GPU-16xCPU-192GB-1xL40S is a GPU (16 vCPUs, 192 GiB RAM, 1x L40S) server offered by UpCloud with 16 vCPUs, 192 GiB of memory and 0 GB of storage. The pricing starts at 1.2872 USD per hour.
16 vCPU
192 GiB Memory
1 GPU
Spare SCore
67,221
(All-cores)
4,173
(Single-core)

Specifications

Server Metadata

Vendor ID
upcloud
Name
GPU-16xCPU-192GB-1xL40S
Description
GPU (16 vCPUs, 192 GiB RAM, 1x L40S)
Family
GPU
Hw Virt
Average Time To Start
49
Status
active
Observed At
2026-10-24T05:41:07.739140

Availability

REGION / IDSPOTONDEMAND
Helsinki #2 (FI) / fi-hel2
1.2872 USD/h 1.7106 USD/h

Processor

vCPUs
16
Hypervisor
KVM
CPU Allocation
Shared
CPU Cores
16
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
1 MiB
CPU L1I Cache
64 KiB
CPU L1I Cache Total
1 MiB
CPU L2 Cache
512 KiB
CPU L2 Cache Total
8 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
16.1
Scalability
100.63

System Resources and Accelerators

MEMORY
Memory Amount
192 GiB
Memory Amount Actual
192 GiB
ACCELERATORS
Accelerator Count
1
Accelerator Memory Min
48 GiB
Accelerator Memory Total
48 GiB
Accelerator Type
GPU
Accelerator Manufacturer
NVIDIA
Accelerator Family
Ada Lovelace
Accelerator Model
L40S
Accelerators
  • NVIDIA Ada Lovelace NVIDIA L40S (Memory amount: 46068, Firmware version: 595.84, BIOS version: 95.02.66.00.15, Clock rate: 2520)
STORAGE
Storage Size
0 GB
Storages
    NETWORK
    Inbound Traffic
    0 GB/month
    Outbound Traffic
    24576 GB/month
    IPv4
    1

    CPU and System Topology

    Server Description

    A GPU-accelerated server combining AMD EPYC processing and NVIDIA L40S graphics memory for demanding AI inference and data-intensive workloads.

    GPU AcceleratedMemory Optimized

    UpCloud GPU-16xCPU-192GB-1xL40S is a GPU-accelerated server featuring 16 vCPUs and 192 GB of RAM on a KVM hypervisor. Powered by an AMD EPYC 9575F x86_64 processor with a shared CPU allocation, the server integrates a single NVIDIA L40S GPU with 48 GB of VRAM. It lacks local storage but includes one public IPv4 address. Benchmarks show top-tier CPU performance, with a Geekbench single-core score of 2959.0 and a multi-core score of 22269.0. Memory bandwidth is strong, reaching 329.44 GB/sec for large-block reads. The server excels in LLM inference, processing 3898.64 tokens per second on a 135M model and 36.41 tokens per second on a 7B model. This configuration is suited for AI inference, image processing, and database operations, offering high resource density despite the shared CPU tradeoff.

    Economics

    Average Price per Region

    Prices per Zone

    Lowest Prices

    Performance

    Workload Profiles

    3.2Score
    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) 6,469,900 ops/sec 1,919,289 ops/sec 3.4 50.00% 50.00% +84.00%
    Redis RPS (pipeline=16, SET) 42,156,418 ops/sec 11,728,058 ops/sec 3.5 20.00% 20.00% +28.00%
    PassMark Memory Mark (composite) 2,865 2,416 1.2 10.00% 10.00% +1.80%
    Memory bandwidth (read, 16 MB ~ L3) 941,064 MB/sec 108,466 MB/sec 8.5 10.00% 10.00% +24.00%
    PassMark single-thread CPU 4,268 Mops/s 2,365 Mops/s 1.8 10.00% 10.00% +6.10%

    GPU Bandwidth and Latency

    Node (all GPUs)
    Node baseline directional bandwidth
    		22.6 GB/s
    Node GPU-to-host bandwidth
    		26.4 GB/s
    Node host-to-GPU bandwidth
    		26.9 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
    		581 ns
    Benchmark Environment
    Driver Version
    		 595.84
    CUDA Runtime Version
    		 13.2
    nvbandwidth Version
    		 v0.8
    GPU Count
    		 1
    P2P Supported
    No
    Efficiency
    SM/CE copy efficiency
    		1.00

    Memory Bandwidth

    Compression

    OpenSSL

    Geekbench Single-Core

    Score: 2,959

    Geekbench Multi-Core

    Score: 22,269

    Passmark CPU Scores

    BENCHMARKSCORE
    Mark
    		 52,860
    Compression
    		 593,722
    Encryption
    		 28,379
    Extended Instructions
    		 52,916
    Floating Point Maths
    		 136,461
    Integer Maths
    		 136,322
    Physics
    		 8,898
    Prime Numbers
    		 535
    Single Threaded
    		 4,268
    String Sorting
    		 82,061

    Passmark Memory Scores

    BENCHMARKSCORE
    Memory Mark
    		 2,865
    Database Operations
    		 22,382
    Memory Latency
    		 87
    Memory Read Cached
    		 38,094
    Memory Read Uncached
    		 35,342
    Memory Write
    		 35,422

    Stress-ng Stressors

    CPU integerSingleMultiScaling
    atomic
    		 589.18 446.02 0.76×
    bitops
    		 254,639 4,048,631 15.90×
    branch
    		 191,479 3,161,395 16.51×
    funccall
    		 1,001 15,678 15.66×
    funcret
    		 30,634 469,875 15.34×
    hash
    		 240,558 3,824,284 15.90×
    intmath
    		 88,336 1,409,991 15.96×
    nop
    		 473,587 7,486,259 15.81×
    prime
    		 675,440 10,640,928 15.75×
    rotate
    		 157,271 2,512,670 15.98×
    str
    		 1,769,694 27,213,855 15.38×
    wcs
    		 1,179,258 18,003,647 15.27×
    CryptographySingleMultiScaling
    crypt
    		 6,023 93,907 15.59×
    CompressionSingleMultiScaling
    jpeg
    		 1,423 22,206 15.61×
    zlib
    		 88.38 1,283 14.51×
    HypervisorSingleMultiScaling
    clock
    		 86,733 257,280 2.97×
    clone
    		 434.71 5,973 13.74×
    context
    		 9,695 155,343 16.02×
    cpu-sched
    		 66.67 1,058 15.87×
    eventfd
    		 80,096 800,373 9.99×
    fault
    		 44,230 172,780 3.91×
    fork
    		 7,760 65,462 8.44×
    getrandom
    		 110,464 1,723,902 15.61×
    hrtimers
    		 310,583 3,551,592 11.44×
    mprotect
    		 723,149 8,099,241 11.20×
    mtx
    		 59,801,336 1,517,825,922 25.38×
    munmap
    		 5,678 49,997 8.81×
    null
    		 1,619,814 25,226,150 15.57×
    pipe
    		 698,630 12,433,449 17.80×
    priv-instr
    		 1,370,449 9,230,235 6.74×
    resched
    		 12,022,676 7,654,342 0.64×
    switch
    		 104,816 7,459,746 71.17×
    timermix
    		 367,116 5,556,891 15.14×
    tlb-shootdown
    		 21,014 147,299 7.01×
    tsc
    		 986,764 15,713,205 15.92×
    usersyscall
    		 1,224,231 9,240,512 7.55×
    vdso
    		 110,506,922 1,748,676,999 15.82×
    vfork
    		 41,985 187,206 4.46×
    x86syscall
    		 20,954,061 331,749,751 15.83×
    yield
    		 26,398,974 28,714,502 1.09×
    zero
    		 8,818,415 139,145,743 15.78×
    CPU float/SIMDSingleMultiScaling
    eigen
    		 19,516 307,466 15.75×
    expmath
    		 4,457 70,407 15.80×
    fma
    		 1,737,461 27,358,307 15.75×
    fp
    		 223.2 3,576 16.02×
    fractal
    		 9.8 153.21 15.63×
    hyperbolic
    		 6,087 96,636 15.87×
    logmath
    		 6,248 99,991 16.00×
    matrix
    		 9,978 158,203 15.85×
    matrix-3d
    		 592 9,517 16.08×
    monte-carlo
    		 506.56 7,913 15.62×
    mpfr
    		 2,644 41,733 15.78×
    powmath
    		 2,028 32,339 15.94×
    trig
    		 2,715 42,723 15.74×
    veccmp
    		 33,017 525,461 15.92×
    vecfp
    		 2,652 42,853 16.16×
    vecmath
    		 5,277 83,682 15.86×
    vecshuf
    		 1,105 17,490 15.82×
    vecwide
    		 44,837 694,471 15.49×
    AlgorithmsSingleMultiScaling
    bitonicsort
    		 14 222.6 15.90×
    bsearch
    		 400.28 6,347 15.86×
    bubblesort
    		 1.4 22.8 16.29×
    fibsearch
    		 479.75 7,523 15.68×
    heapsort
    		 15.6 245.99 15.77×
    hsearch
    		 14,845 219,614 14.79×
    insertionsort
    		 17.6 272.6 15.49×
    malloc
    		 5,537,997 37,390,426 6.75×
    mcontend
    		 113.03 2,594 22.95×
    mergesort
    		 42.8 670.39 15.66×
    qsort
    		 28.8 461.79 16.03×
    skiplist
    		 9,421 148,125 15.72×
    tree
    		 2.8 43.58 15.56×
    MemorySingleMultiScaling
    cache
    		 1,353,068 1,376,193 1.02×
    cachehammer
    		 291,066 4,128,302 14.18×
    cacheline
    		 65.85 995.23 15.11×
    l1cache
    		 759.78 11,872 15.63×
    memcpy
    		 871.13 14,141 16.23×
    memrate
    		 1.12 12.69 11.33×
    memthrash
    		 165.41 981.32 5.93×
    ptr-chase
    		 84,977,190 535,773,079 6.30×
    spinmem
    		 453.58 1,716 3.78×
    stream
    		 488.16 27,083 55.48×
    vm
    		 259,694 293,824 1.13×

    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-2xL40S 16192 GiB2
    GPU-16xCPU-192GB-3xL40S 16192 GiB3
    GPU-20xCPU-256GB-1xL40S 20256 GiB1

    Similar Servers

    INSTANCEVENDORvCPUsMEMORYAccelerators

    GPU-16xCPU-192GB-1xL40S FAQs