l4-360 by OVHcloud

l4-360 is a Cloud GPU (90 vCPUs, 360 GiB RAM, 400 GB NVMe, 4xL4 24 GiB VRAM) server offered by OVHcloud with 90 vCPUs, 360 GiB of memory and 400 GB of storage. The pricing starts at 3.3579 USD per hour.
90 vCPU
360 GiB Memory
400 GB Storage
4 GPU
Spare SCore
146,192
(All-cores)
3,139
(Single-core)

Specifications

Server Metadata

Vendor ID
ovh
Name
l4-360
Description
Cloud GPU (90 vCPUs, 360 GiB RAM, 400 GB NVMe, 4xL4 24 GiB VRAM)
Family
Cloud GPU
Hw Virt
Status
active
Observed At
2026-10-25T21:22:14.684205

Availability

REGION / IDSPOTONDEMAND
GRA11 (FR) / GRA11
- 3.3579 USD/h

Processor

vCPUs
90
Hypervisor
KVM
CPU Allocation
Dedicated
CPU Cores
90
CPU Speed
2.75 GHz
CPU Architecture
x86_64
CPU Manufacturer
AMD
CPU Family
EPYC
CPU Model
9454
CPU L1D Cache
64 KiB
CPU L1D Cache Total
6 MiB
CPU L1I Cache
64 KiB
CPU L1I Cache Total
6 MiB
CPU L2 Cache
512 KiB
CPU L2 Cache Total
45 MiB
CPU L3 Cache
16 MiB
CPU L3 Cache Total
1 GiB
CPU Flags
fpu, vme, de, pse, tsc, msr, pae, mce, cx8, apic, sep, mtrr, pge, mca, cmov, pat, pse36, clflush, mmx, fxsr, sse, sse2, 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, svm, cr8_legacy, abm, sse4a, misalignsse, 3dnowprefetch, osvw, perfctr_core, ssbd, ibrs, ibpb, stibp, vmmcall, 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, avx512_bf16, clzero, xsaveerptr, wbnoinvd, arat, npt, lbrv, nrip_save, tsc_scale, vmcb_clean, pausefilter, pfthreshold, v_vmsave_vmload, vgif, avx512vbmi, umip, pku, ospke, avx512_vbmi2, gfni, vaes, vpclmulqdq, avx512_vnni, avx512_bitalg, avx512_vpopcntdq, la57, rdpid, overflow_recov, succor, fsrm, arch_capabilities
Ecpus
46.6
Scalability
51.78

System Resources and Accelerators

MEMORY
Memory Amount
360 GiB
Memory Amount Actual
352 GiB
ACCELERATORS
Accelerator Count
4
Accelerator Memory Min
24 GiB
Accelerator Memory Total
96 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: 590.48.01, BIOS version: 95.04.65.00.13, Clock rate: 2040)
  • NVIDIA Ada Lovelace NVIDIA L4 (Memory amount: 23034, Firmware version: 590.48.01, BIOS version: 95.04.65.00.13, Clock rate: 2040)
  • NVIDIA Ada Lovelace NVIDIA L4 (Memory amount: 23034, Firmware version: 590.48.01, BIOS version: 95.04.65.00.13, Clock rate: 2040)
  • NVIDIA Ada Lovelace NVIDIA L4 (Memory amount: 23034, Firmware version: 590.48.01, BIOS version: 95.04.65.00.13, Clock rate: 2040)
STORAGE
Storage Size
400 GB
Storage Type
nvme ssd
Storages
  • 400 GB nvme ssd
NETWORK
Network Speed Baseline
25 Gbps
Network Speed Max
25 Gbps
Inbound Traffic
0 GB/month
Outbound Traffic
0 GB/month
IPv4
1

CPU and System Topology

Server Description

A dedicated multi-GPU server combining ninety physical AMD EPYC cores with four NVIDIA L4 accelerators for intensive parallel computing and machine learning.

GPU AcceleratedCompute OptimizedStorage & Database

OVHcloud l4-360 is a GPU-accelerated server instance featuring 90 dedicated AMD EPYC 9454 physical cores running at 2.75 GHz, paired with 360.0 GB of system memory. The hardware configuration includes four NVIDIA L4 GPUs based on the Ada Lovelace architecture, providing a total of 96 GB of VRAM, alongside a 400 GB NVMe SSD and 25 Gbps baseline network bandwidth. Benchmark data shows top-tier performance in multi-core processing, cryptography, and LLM inference, with the system processing 2772.24 tokens/sec for 7B model prompt processing. While the 400 GB local storage is modest relative to the compute capacity, the high density of dedicated cores and GPUs makes this instance highly efficient for parallelized tasks. The server is suited for machine learning inference, data science, software compilation, and cryptographic workloads.

Economics

Average Price per Region

Prices per Zone

Lowest Prices

Performance

Workload Profiles

3Score
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): Memory bandwidth (read, 16 MB ~ L3).
Component Score Weight Impact
Raw Reference Normalized Target Actual
Redis RPS (pipeline=1, SET) 6,803,367 ops/sec 1,919,289 ops/sec 3.6 50.00% 56.00% +100.00%
Redis RPS (pipeline=16, SET) 58,920,432 ops/sec 11,728,058 ops/sec 4.9 20.00% 22.00% +42.00%
PassMark Memory Mark (composite) 3,052 2,416 1.3 10.00% 11.00% +2.90%
Memory bandwidth (read, 16 MB ~ L3) — 108,466 MB/sec — 10.00% 0.00% —
PassMark single-thread CPU 2,979 Mops/s 2,365 Mops/s 1.2 10.00% 11.00% +2.00%

GPU Bandwidth and Latency

Node (all GPUs)
Node baseline directional bandwidth
44.0 GB/s
Node GPU-to-host bandwidth
52.9 GB/s
Node host-to-GPU bandwidth
53.5 GB/s
Single GPU slot
Per-GPU GPU-to-host bandwidth
13.2 GB/s
Per-GPU host-to-GPU bandwidth
13.4 GB/s
Per-GPU host-device latency
681 ns
Benchmark Environment
Driver Version
 535.274.02
CUDA Runtime Version
 12.6
nvbandwidth Version
 v0.5
GPU Count
 4
P2P Supported
No
Efficiency
SM/CE copy efficiency
1.00

Memory Bandwidth

Compression

OpenSSL

Geekbench Single-Core

Score: 2,070

Geekbench Multi-Core

Score: 16,093

Passmark CPU Scores

BENCHMARKSCORE
Mark
 86,477
Compression
 1,519,258
Encryption
 99,948
Extended Instructions
 96,764
Floating Point Maths
 242,590
Integer Maths
 427,885
Physics
 6,499
Prime Numbers
 923
Single Threaded
 2,979
String Sorting
 156,338

Passmark Memory Scores

BENCHMARKSCORE
Memory Mark
 3,052
Database Operations
 24,309
Memory Latency
 69
Memory Read Cached
 28,550
Memory Read Uncached
 27,418
Memory Write
 24,386

Stress-ng Stressors

CPU integerSingleMultiScaling
atomic
 963.83 184.5 0.19×
bitops
 165,035 8,433,492 51.10×
branch
 219,017 13,837,671 63.18×
funccall
 634.68 36,040 56.78×
funcret
 21,425 1,173,771 54.79×
hash
 167,973 10,922,699 65.03×
intmath
 68,041 4,095,680 60.19×
nop
 644,467 30,116,723 46.73×
prime
 494,279 25,362,115 51.31×
rotate
 90,621 4,199,570 46.34×
str
 1,099,425 52,316,643 47.59×
wcs
 718,811 29,810,985 41.47×
CryptographySingleMultiScaling
crypt
 3,820 169,129 44.27×
CompressionSingleMultiScaling
jpeg
 914.9 38,934 42.56×
zlib
 57.84 4,451 76.95×
HypervisorSingleMultiScaling
clock
 22,900 188,765 8.24×
clone
 254.53 7,398 29.06×
context
 2,972 233,137 78.45×
cpu-sched
 72.59 5,670 78.10×
eventfd
 37,473 2,011,291 53.67×
fault
 10,538 140,099 13.30×
fork
 1,708 54,082 31.66×
getrandom
 54,607 2,890,465 52.93×
hrtimers
 146,548 9,138,083 62.36×
mprotect
 611,822 14,269,225 23.32×
mtx
 45,958,951 4,494,159,510 97.79×
munmap
 1,044 60,616 58.06×
null
 358,673 26,632,924 74.25×
pipe
 362,384 19,572,334 54.01×
priv-instr
 432,009 19,712,238 45.63×
resched
 5,777,942 11,396,020 1.97×
switch
 105,590 3,610,616 34.19×
timermix
 144,392 9,526,764 65.98×
tlb-shootdown
 15,679 170,323 10.86×
tsc
 804,732 68,872,870 85.58×
usersyscall
 322,309 18,603,849 57.72×
vdso
 86,180,050 5,979,174,407 69.38×
vfork
 13,590 72,151 5.31×
x86syscall
 4,080,935 309,353,319 75.80×
yield
 45,366,843 59,627,565 1.31×
zero
 2,825,258 199,890,657 70.75×
CPU float/SIMDSingleMultiScaling
eigen
 16,749 787,601 47.02×
expmath
 3,055 190,204 62.25×
fma
 944,619 55,317,208 58.56×
fp
 256.79 16,064 62.56×
fractal
 7.6 465.93 61.31×
hyperbolic
 3,923 234,765 59.85×
logmath
 4,678 274,613 58.71×
matrix
 5,044 215,365 42.70×
matrix-3d
 989.9 4,501 4.55×
monte-carlo
 172.1 11,777 68.43×
mpfr
 1,739 86,963 50.01×
powmath
 1,351 90,963 67.35×
trig
 1,887 112,484 59.60×
veccmp
 19,520 954,038 48.87×
vecfp
 1,862 102,543 55.09×
vecmath
 4,143 261,579 63.14×
vecshuf
 656.88 34,109 51.93×
vecwide
 29,410 1,328,831 45.18×
AlgorithmsSingleMultiScaling
bitonicsort
 10.8 492.38 45.59×
bsearch
 314.71 16,541 52.56×
bubblesort
 0.8 52 65.00×
fibsearch
 391.11 20,525 52.48×
heapsort
 11.2 592.98 52.94×
hsearch
 11,728 492,404 41.99×
insertionsort
 13.4 492.4 36.75×
malloc
 3,856,799 126,739,529 32.86×
mcontend
 76.07 13,465 177.00×
mergesort
 34.6 1,898 54.85×
qsort
 22.4 1,145 51.10×
skiplist
 7,255 452,583 62.38×
tree
 2 82.61 41.30×
MemorySingleMultiScaling
cache
 531,280 903,386 1.70×
cachehammer
 281,339 12,890,333 45.82×
cacheline
 84.76 - -
l1cache
 442.46 20,917 47.27×
memcpy
 459.41 19,453 42.34×
memrate
 3.93 16.16 4.11×
memthrash
 265.67 564.46 2.12×
ptr-chase
 39,857,167 629,928,350 15.80×
spinmem
 251.79 4,707 18.70×
stream
 279 64,476 231.10×
vm
 84,114 303,403 3.61×

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
rtx5000-28 428 GiB1
t1-le-45 845 GiB1
rtx5000-56 856 GiB2
t2-le-45 1545 GiB1
l40s-90 1590 GiB1
a100-180 15180 GiB1
t1-le-90 1690 GiB2

Similar Servers

INSTANCEVENDORvCPUsMEMORYAccelerators

l4-360 FAQs