a10-90 by OVHcloud

a10-90 is a Cloud GPU (60 vCPUs, 85 GiB RAM, 400 GB SSD, 2xA10 24 GiB VRAM) server offered by OVHcloud with 60 vCPUs, 85 GiB of memory and 400 GB of storage. The pricing starts at 1.6981 USD per hour.
60 vCPU
85 GiB Memory
400 GB Storage
2 GPU
Spare SCore
187,019
(All-cores)
3,091
(Single-core)

Specifications

Server Metadata

Vendor ID
ovh
Name
a10-90
Description
Cloud GPU (60 vCPUs, 85 GiB RAM, 400 GB SSD, 2xA10 24 GiB VRAM)
Family
Cloud GPU
Hw Virt
Status
active
Observed At
2026-10-31T19:15:36.093485

Availability

REGION / IDSPOTONDEMAND
GRA11 (FR) / GRA11
- 1.6981 USD/h
GRA9 (FR) / GRA9
- 1.6981 USD/h

Processor

vCPUs
60
Hypervisor
KVM
CPU Allocation
Dedicated
CPU Cores
60
CPU Speed
3.3 GHz
CPU Architecture
x86_64
CPU Manufacturer
AMD
CPU Family
EPYC
CPU Model
9554
CPU L1D Cache
64 KiB
CPU L1D Cache Total
4 MiB
CPU L1I Cache
64 KiB
CPU L1I Cache Total
4 MiB
CPU L2 Cache
512 KiB
CPU L2 Cache Total
30 MiB
CPU L3 Cache
16 MiB
CPU L3 Cache Total
960 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, 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
60.5
Scalability
100.83

System Resources and Accelerators

MEMORY
Memory Amount
85 GiB
Memory Amount Actual
88 GiB
ACCELERATORS
Accelerator Count
2
Accelerator Memory Min
24 GiB
Accelerator Memory Total
48 GiB
Accelerator Type
GPU
Accelerator Manufacturer
NVIDIA
Accelerator Family
Ampere
Accelerator Model
A10
Accelerators
  • NVIDIA Ampere NVIDIA A10 (Memory amount: 23028, Firmware version: 590.48.01, BIOS version: 94.02.5C.00.04, Clock rate: 1695)
  • NVIDIA Ampere NVIDIA A10 (Memory amount: 23028, Firmware version: 590.48.01, BIOS version: 94.02.5C.00.04, Clock rate: 1695)
STORAGE
Storage Size
400 GB
Storage Type
ssd
Storages
  • 400 GB ssd
NETWORK
Network Speed Baseline
16 Gbps
Network Speed Max
16 Gbps
Inbound Traffic
0 GB/month
Outbound Traffic
0 GB/month
IPv4
1

CPU and System Topology

Server Description

A high-density GPU server featuring dual accelerators and dedicated AMD EPYC cores optimized for parallel computing and LLM inference.

GPU AcceleratedCompute Optimized

The OVHcloud a10-90 is a GPU-accelerated server designed for compute-intensive and parallel processing workloads. It features 60 dedicated AMD EPYC 9554 physical cores running at 3.3 GHz, paired with 85 GB of system memory and 400 GB of local SSD storage. Acceleration is driven by two NVIDIA A10 Ampere GPUs delivering 48 GB of combined VRAM. Benchmarks show top-tier performance in multi-core CPU execution, Redis operations, and LLM inference across various model sizes. However, with only 1.42 GB of RAM per core, the server exhibits a low memory-to-vCPU ratio, presenting a tradeoff for memory-heavy applications. It is highly cost-efficient for GPU-bound tasks, machine learning, and parallel compilation, but less suited for large in-memory databases.

Economics

Average Price per Region

Prices per Zone

Lowest Prices

Performance

Workload Profiles

2.9Score
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) 6,377,804 ops/sec 1,918,343 ops/sec 3.4 50.00% 50.00% +84.00%
Redis RPS (pipeline=16, SET) 65,602,630 ops/sec 11,704,734 ops/sec 5.5 20.00% 20.00% +41.00%
PassMark Memory Mark (composite) 3,031 2,416 1.2 10.00% 10.00% +1.80%
Memory bandwidth (read, 16 MB ~ L3) 225,855 MB/sec 108,466 MB/sec 2.1 10.00% 10.00% +7.70%
PassMark single-thread CPU 2,935 Mops/s 2,365 Mops/s 1.2 10.00% 10.00% +1.80%

GPU Bandwidth and Latency

Node (all GPUs)
Node baseline directional bandwidth
22.0 GB/s
Node GPU-to-host bandwidth
26.4 GB/s
Node host-to-GPU bandwidth
26.7 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
671 ns
Benchmark Environment
Driver Version
 535.274.02
CUDA Version (driver)
 12.6
nvbandwidth Version
 v0.5
GPU Count
 2
P2P Supported
No
Efficiency
SM/CE copy efficiency
1.00

Memory Bandwidth

Compression

OpenSSL

Geekbench Single-Core

Score: 2,028

Geekbench Multi-Core

Score: 16,254

Passmark CPU Scores

BENCHMARKSCORE
Mark
 92,462
Compression
 1,561,968
Encryption
 79,726
Extended Instructions
 126,529
Floating Point Maths
 323,408
Integer Maths
 367,170
Physics
 12,990
Prime Numbers
 1,182
Single Threaded
 2,935
String Sorting
 225,537

Passmark Memory Scores

BENCHMARKSCORE
Memory Mark
 3,031
Database Operations
 29,381
Memory Latency
 69
Memory Read Cached
 27,945
Memory Read Uncached
 26,837
Memory Write
 25,288

Stress-ng Stressors

CPU integerSingleMultiScaling
atomic
 854.04 319.98 0.37×
bitops
 161,747 9,367,879 57.92×
branch
 215,806 11,627,530 53.88×
funccall
 627.63 36,899 58.79×
funcret
 21,256 1,203,118 56.60×
hash
 165,533 9,759,664 58.96×
intmath
 62,594 2,985,773 47.70×
nop
 616,290 34,083,328 55.30×
prime
 483,424 27,838,910 57.59×
rotate
 89,382 5,029,063 56.26×
str
 1,106,941 60,677,790 54.82×
wcs
 709,953 36,007,233 50.72×
CryptographySingleMultiScaling
crypt
 3,704 218,095 58.87×
CompressionSingleMultiScaling
jpeg
 728.54 17,825 24.47×
zlib
 59.75 3,432 57.44×
HypervisorSingleMultiScaling
clock
 22,825 235,701 10.33×
clone
 243.31 2,407 9.89×
context
 3,144 44,811 14.25×
cpu-sched
 70.33 3,884 55.22×
eventfd
 34,220 1,528,095 44.66×
fault
 8,895 153,145 17.22×
fork
 1,407 79,107 56.20×
getrandom
 53,206 3,112,614 58.50×
hrtimers
 185,499 7,070,836 38.12×
mprotect
 588,280 11,042,708 18.77×
mtx
 65,248,641 6,143,234,236 94.15×
munmap
 1,396 68,023 48.71×
null
 368,466 21,721,993 58.95×
pipe
 359,890 15,879,585 44.12×
priv-instr
 384,744 14,117,005 36.69×
resched
 6,007,382 8,721,044 1.45×
switch
 88,681 5,229,065 58.96×
timermix
 121,570 7,239,538 59.55×
tlb-shootdown
 14,740 196,623 13.34×
tsc
 793,512 47,421,139 59.76×
usersyscall
 288,851 13,427,446 46.49×
vdso
 85,047,613 4,933,984,826 58.01×
vfork
 14,723 132,723 9.01×
x86syscall
 4,568,789 271,311,356 59.38×
yield
 33,955,449 41,637,372 1.23×
zero
 2,715,421 158,945,069 58.53×
CPU float/SIMDSingleMultiScaling
eigen
 16,113 977,676 60.67×
expmath
 3,075 174,652 56.80×
fma
 932,673 55,633,775 59.65×
fp
 255.2 15,136 59.31×
fractal
 7.6 426.82 56.16×
hyperbolic
 3,863 227,127 58.79×
logmath
 4,610 269,280 58.41×
matrix
 4,943 266,806 53.98×
matrix-3d
 968.13 5,452 5.63×
monte-carlo
 192.77 8,997 46.67×
mpfr
 1,419 60,657 42.75×
powmath
 1,349 80,067 59.36×
trig
 1,862 109,594 58.84×
veccmp
 19,230 1,118,573 58.17×
vecfp
 1,836 106,431 57.97×
vecmath
 4,061 235,952 58.10×
vecshuf
 602.72 29,551 49.03×
vecwide
 26,521 1,089,383 41.08×
AlgorithmsSingleMultiScaling
bitonicsort
 10.4 625.58 60.15×
bsearch
 310.47 13,983 45.04×
bubblesort
 0.8 47.6 59.50×
fibsearch
 386.5 22,935 59.34×
heapsort
 11.2 648.58 57.91×
hsearch
 11,523 633,305 54.96×
insertionsort
 13 391.8 30.14×
malloc
 3,784,389 109,742,161 29.00×
mcontend
 81.65 9,508 116.45×
mergesort
 34.4 1,932 56.17×
qsort
 22.2 1,224 55.12×
skiplist
 7,126 420,159 58.96×
tree
 2 87.21 43.60×
MemorySingleMultiScaling
cache
 589,216 1,057,232 1.79×
cachehammer
 288,172 11,127,015 38.61×
cacheline
 27.38 3,989 145.70×
l1cache
 416.72 22,202 53.28×
memcpy
 452.78 26,073 57.58×
memrate
 3.91 13.98 3.58×
memthrash
 212.39 1,040 4.89×
ptr-chase
 43,420,736 514,325,036 11.85×
spinmem
 255.99 3,622 14.15×
stream
 280.88 42,714 152.07×
vm
 89,983 300,110 3.34×

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

a10-90 FAQs