cax21 by Hetzner Cloud

cax21 is a CAX21 server offered by Hetzner Cloud with 4 vCPUs, 8 GiB of memory and 80 GB of storage.
4 vCPU
8 GiB Memory
80 GB Storage
Spare SCore
7,056
(All-cores)
1,778
(Single-core)

Specifications

Server Metadata

Vendor ID
hcloud
Server ID
93
Name
cax21
Description
CAX21
Family
Cost Optimized
Series
cax
Hw Virt
Average Time To Start
32.25
Status
active
Observed At
2026-11-14T01:46:52.933092

Availability

Region / IDSpotOndemand

Processor

vCPUs
4
Hypervisor
QEMU
CPU Allocation
Shared
CPU Cores
4
CPU Speed
2 GHz
CPU Architecture
arm64
CPU Manufacturer
Ampere
CPU Family
Ampere Altra
CPU Model
Neoverse-N1
CPU Flags
fp, asimd, evtstrm, aes, pmull, sha1, sha2, crc32, atomics, fphp, asimdhp, cpuid, asimdrdm, lrcpc, dcpop, asimddp, ssbs
Ecpus
4
Scalability
100

System Resources and Accelerators

Memory
Memory Amount
8 GiB
Memory Amount Actual
8 GiB
Accelerators
Accelerator Count
0
Accelerator Memory Min
0 MiB
Accelerator Memory Total
0 MiB
Accelerators
    Storage
    Storage Size
    80 GB
    Storage Type
    nvme ssd
    Storages
      Compatible Storage IDs
      block
      Network
      Inbound Traffic
      0 GB/month
      Outbound Traffic
      20480 GB/month
      IPv4
      0

      CPU and System Topology

      Server Description

      An entry-level ARM-based virtual server featuring shared Ampere Altra cores and local NVMe storage for budget-conscious development and lightweight hosting.

      Hetzner Cloud cax21 is an arm64 virtual server featuring 4 shared vCPUs powered by an Ampere Altra Neoverse-N1 processor running at 2.0 GHz. It is configured with 8.0 GB of RAM and 80 GB of local NVMe SSD storage. Benchmarks show average performance in general CPU execution, with stress-ng single-core and multi-core scores of 1777.99 and 7055.54, though memory bandwidth and cryptographic operations perform weakly. The shared CPU allocation presents a tradeoff in performance consistency, making it a highly cost-efficient option for budget-conscious deployments rather than high-performance computing. The cax21 is best suited for lightweight web serving, development environments, and basic database operations.

      Economics

      Average Price per Region

      Prices per Zone

      Lowest Prices

      Performance

      Workload Profiles

      0.25Score
      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) 343,330 ops/sec 1,933,252 ops/sec 0.18 50.00% 50.00% -58.00%
      Redis RPS (pipeline=16, SET) 2,441,762 ops/sec 11,814,424 ops/sec 0.2 20.00% 20.00% -28.00%
      PassMark Memory Mark (composite) 1,482 2,408 0.62 10.00% 10.00% -4.70%
      Memory bandwidth (read, 16 MB ~ L3) 47,770 MB/sec 108,674 MB/sec 0.44 10.00% 10.00% -7.90%
      PassMark single-thread CPU 1,255 Mops/s 2,365 Mops/s 0.54 10.00% 10.00% -6.00%

      Memory Bandwidth

      Compression

      OpenSSL

      Geekbench Single-Core

      Score: 1,005

      Geekbench Multi-Core

      Score: 3,335

      Passmark CPU Scores

      BENCHMARKSCORE
      Mark
      		 3,112
      Compression
      		 21,369
      Encryption
      		 637
      Extended Instructions
      		 2,930
      Floating Point Maths
      		 14,099
      Integer Maths
      		 18,658
      Physics
      		 572
      Prime Numbers
      		 29
      Single Threaded
      		 1,255
      String Sorting
      		 13,033

      Passmark Memory Scores

      BENCHMARKSCORE
      Memory Mark
      		 1,482
      Database Operations
      		 2,685
      Memory Latency
      		 75
      Memory Read Cached
      		 10,993
      Memory Read Uncached
      		 8,240
      Memory Write
      		 14,638

      Stress-ng Stressors

      CPU integerSingleMultiScaling
      atomic
      		 104.41 103.74 0.99×
      bitops
      		 107,542 427,878 3.98×
      branch
      		 218,352 867,177 3.97×
      funccall
      		 1,327 5,256 3.96×
      funcret
      		 20,645 81,026 3.92×
      hash
      		 123,104 497,957 4.05×
      intmath
      		 30,907 123,553 4.00×
      nop
      		 171,834 682,527 3.97×
      prime
      		 333,828 1,322,940 3.96×
      rotate
      		 95,609 378,305 3.96×
      str
      		 511,143 2,024,220 3.96×
      wcs
      		 118,284 467,799 3.95×
      CryptographySingleMultiScaling
      crypt
      		 2,881 11,378 3.95×
      CompressionSingleMultiScaling
      jpeg
      		 496.69 1,953 3.93×
      zlib
      		 31.11 134.25 4.32×
      HypervisorSingleMultiScaling
      clock
      		 22,758 101,881 4.48×
      clone
      		 276.86 685.33 2.48×
      context
      		 2,753 10,951 3.98×
      cpu-sched
      		 68.44 283.86 4.15×
      eventfd
      		 58,260 225,659 3.87×
      fault
      		 10,549 30,982 2.94×
      fork
      		 2,359 11,248 4.77×
      getrandom
      		 48,597 194,248 4.00×
      hrtimers
      		 342,606 439,616 1.28×
      mprotect
      		 847,329 1,202,340 1.42×
      mtx
      		 6,046,218 160,119,277 26.48×
      munmap
      		 1,434 7,604 5.30×
      null
      		 478,081 1,893,293 3.96×
      pipe
      		 613,611 1,946,527 3.17×
      priv-instr
      		 720,375 2,586,820 3.59×
      resched
      		 1,649,752 705,430 0.43×
      switch
      		 115,586 667,898 5.78×
      timermix
      		 318,165 1,261,587 3.97×
      tlb-shootdown
      		 3,944 5,264 1.33×
      vdso
      		 45,509,055 179,567,215 3.95×
      vfork
      		 17,447 59,223 3.39×
      yield
      		 2,848,809 3,054,041 1.07×
      zero
      		 2,457,420 9,745,937 3.97×
      CPU float/SIMDSingleMultiScaling
      eigen
      		 1,318 5,244 3.98×
      expmath
      		 237.93 942.78 3.96×
      fma
      		 635,604 2,515,815 3.96×
      fp
      		 102.41 406.17 3.97×
      fractal
      		 6.6 25.2 3.82×
      hyperbolic
      		 242.05 959.49 3.96×
      logmath
      		 291.15 1,150 3.95×
      matrix
      		 5,293 20,964 3.96×
      matrix-3d
      		 471.73 1,833 3.88×
      monte-carlo
      		 193.84 774.11 3.99×
      mpfr
      		 631.55 2,519 3.99×
      powmath
      		 174.91 693.81 3.97×
      trig
      		 327.25 1,297 3.96×
      veccmp
      		 17,046 67,714 3.97×
      vecfp
      		 738.23 2,924 3.96×
      vecmath
      		 3,055 12,150 3.98×
      vecshuf
      		 665.23 2,643 3.97×
      vecwide
      		 18,227 72,255 3.96×
      AlgorithmsSingleMultiScaling
      bitonicsort
      		 5.2 21 4.04×
      bsearch
      		 255.88 1,016 3.97×
      bubblesort
      		 0.8 3 3.75×
      fibsearch
      		 271.25 1,072 3.95×
      heapsort
      		 8 32 4.00×
      hsearch
      		 4,763 19,071 4.00×
      insertionsort
      		 4.6 17.6 3.83×
      malloc
      		 2,277,329 10,638,986 4.67×
      mcontend
      		 285.15 7,358 25.80×
      mergesort
      		 27.6 108.99 3.95×
      qsort
      		 19 75.6 3.98×
      skiplist
      		 5,319 21,021 3.95×
      tree
      		 1 4 4.00×
      MemorySingleMultiScaling
      cache
      		 249,692 410,988 1.65×
      cachehammer
      		 186,726 499,430 2.67×
      cacheline
      		 44.28 168.82 3.81×
      memcpy
      		 438.74 1,745 3.98×
      memrate
      		 3.15 10.79 3.43×
      memthrash
      		 84.32 114.15 1.35×
      ptr-chase
      		 9,972,337 41,740,276 4.19×
      spinmem
      		 951.15 437.33 0.46×
      stream
      		 1,165 23,902 20.51×
      vm
      		 56,188 258,983 4.61×

      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

      PostgreSQL heavy read-only throughput

      Alternatives

      Servers of the Same Family

      INSTANCEvCPUsMEMORYAccelerators
      cx23 24 GiB0
      cax11 24 GiB0
      cx33 48 GiB0
      cx43 816 GiB0
      cax31 816 GiB0
      cx53 1631 GiB0
      cax41 1631 GiB0

      Similar Servers

      INSTANCEVENDORvCPUsMEMORYAccelerators

      cax21 FAQs