FX96mds_v2 by Microsoft Azure

FX96mds_v2 is a FXmdsv2 family [Memory Intensive] [Local Disk] [Premium Storage capable] [Intel processor] 96 vCPUs server offered by Microsoft Azure with 96 vCPUs, 1832 GiB of memory and 5.67 TB of storage. The pricing starts at 2.4264 USD per hour.
96 vCPU
1832 GiB Memory
5.67 TB Storage
Spare SCore
97,946
(All-cores)
2,170
(Single-core)

Specifications

Server Metadata

Vendor ID
azure
Server ID
Standard_FX96mds_v2
Name
FX96mds_v2
API Reference
Standard_FX96mds_v2
Description
FXmdsv2 family [Memory Intensive] [Local Disk] [Premium Storage capable] [Intel processor] 96 vCPUs
Family
FXmdsv2
Hw Virt
Average Time To Start
116
Status
active
Observed At
2026-10-24T05:10:25.325109

Availability

REGION / IDSPOTONDEMAND
East US 2 (US) / eastus2
2.4264 USD/h 13.13 USD/h
West US 3 (US) / westus3
2.4264 USD/h 13.13 USD/h
East US (US) / eastus
2.756 USD/h 13.13 USD/h
West US 2 (US) / westus2
2.427 USD/h 13.133 USD/h
Central India (IN) / centralindia
2.5042 USD/h 13.551 USD/h
Canada Central (CA) / canadacentral
2.6698 USD/h 14.447 USD/h
Central US (US) / centralus
2.7389 USD/h 14.837 USD/h
Sweden Central (SE) / swedencentral
2.7677 USD/h 14.977 USD/h
Indonesia Central (ID) / indonesiacentral
2.8389 USD/h 15.362 USD/h
South Central US (US) / southcentralus
2.9117 USD/h 15.756 USD/h
West Central US (US) / westcentralus
2.9117 USD/h 15.756 USD/h
North Europe (IE) / northeurope
2.9124 USD/h 15.76 USD/h
Spain Central (ES) / spaincentral
2.9124 USD/h 15.76 USD/h
UK West (GB) / ukwest
2.9359 USD/h 15.887 USD/h
Qatar Central (QA) / qatarcentral
2.9856 USD/h 16.156 USD/h
UAE North (AE) / uaenorth
2.9856 USD/h 16.156 USD/h
France Central (FR) / francecentral
3.0581 USD/h 16.548 USD/h
UK South (GB) / uksouth
3.0581 USD/h 16.548 USD/h
Australia Southeast (AU) / australiasoutheast
3.1544 USD/h 17.069 USD/h
Austria East (AT) / austriaeast
3.1544 USD/h 17.069 USD/h
Belgium Central (BE) / belgiumcentral
3.1544 USD/h 17.069 USD/h
West US (US) / westus
3.1553 USD/h 17.074 USD/h
Chile Central (CL) / chilecentral
3.397 USD/h 18.382 USD/h
Germany West Central (DE) / germanywestcentral
3.3979 USD/h 18.387 USD/h
West Europe (NL) / westeurope
3.3979 USD/h 18.387 USD/h
South India (IN) / southindia
3.4212 USD/h 18.513 USD/h
South Africa North (ZA) / southafricanorth
3.4661 USD/h 18.756 USD/h
Southeast Asia (SG) / southeastasia
3.5184 USD/h 19.039 USD/h
Australia East (AU) / australiaeast
3.5193 USD/h 19.044 USD/h
Japan East (JP) / japaneast
3.5193 USD/h 19.044 USD/h
Korea Central (KR) / koreacentral
3.5193 USD/h 19.044 USD/h
New Zealand North (NZ) / newzealandnorth
3.6942 USD/h 19.99 USD/h
Switzerland North (CH) / switzerlandnorth
3.7378 USD/h 20.226 USD/h
Germany North (DE) / germanynorth
4.4173 USD/h 23.903 USD/h
Brazil South (BR) / brazilsouth
4.7329 USD/h 25.611 USD/h

Processor

vCPUs
96
Hypervisor
Microsoft Hyper-V
CPU Allocation
Dedicated
CPU Cores
48
CPU Speed
4.2 GHz
CPU Architecture
x86_64
CPU Manufacturer
Intel
CPU Family
Xeon
CPU Model
8573C
CPU L1D Cache
48 KiB
CPU L1D Cache Total
2 MiB
CPU L1I Cache
32 KiB
CPU L1I Cache Total
2 MiB
CPU L2 Cache
2 MiB
CPU L2 Cache Total
96 MiB
CPU L3 Cache
260 MiB
CPU L3 Cache Total
520 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, ss, ht, syscall, nx, pdpe1gb, rdtscp, lm, constant_tsc, rep_good, nopl, xtopology, tsc_reliable, nonstop_tsc, cpuid, aperfmperf, tsc_known_freq, pni, pclmulqdq, vmx, ssse3, fma, cx16, pcid, sse4_1, sse4_2, x2apic, movbe, popcnt, tsc_deadline_timer, aes, xsave, avx, f16c, rdrand, hypervisor, lahf_lm, abm, 3dnowprefetch, tpr_shadow, ept, vpid, ept_ad, fsgsbase, tsc_adjust, bmi1, hle, avx2, smep, bmi2, erms, invpcid, rtm, avx512f, avx512dq, rdseed, adx, smap, avx512ifma, clflushopt, clwb, avx512cd, sha_ni, avx512bw, avx512vl, xsaveopt, xsavec, xgetbv1, xsaves, user_shstk, avx_vnni, avx512_bf16, vnmi, avx512vbmi, umip, waitpkg, avx512_vbmi2, gfni, vaes, vpclmulqdq, avx512_vnni, avx512_bitalg, avx512_vpopcntdq, la57, rdpid, cldemote, movdiri, movdir64b, fsrm, serialize, tsxldtrk, ibt, amx_bf16, avx512_fp16, amx_tile, amx_int8, arch_capabilities
Ecpus
45.1
Scalability
46.98

System Resources and Accelerators

MEMORY
Memory Amount
2 TB
Memory Amount Actual
2 TB
ACCELERATORS
Accelerator Count
0
Accelerator Memory Min
0 MiB
Accelerator Memory Total
0 MiB
Accelerators
    STORAGE
    Storage Size
    5669 GB
    Storage Type
    nvme ssd
    Storages
    • 5669 GB nvme ssd
    NETWORK
    Inbound Traffic
    0 GB/month
    Outbound Traffic
    0 GB/month
    IPv4
    0

    CPU and System Topology

    Server Description

    A high-frequency, memory-intensive server with local NVMe storage optimized for enterprise databases and demanding in-memory transactional workloads.

    Memory OptimizedStorage & DatabaseCompute Optimized

    Microsoft Azure FX96mds_v2 is a memory-optimized server utilizing the x86_64 architecture with a dedicated 96-vCPU Intel Xeon 8573C processor running at 4.2 GHz. It features 1832.0 GB of RAM and 5669 GB of local NVMe SSD storage, delivering a high memory-to-core ratio of 19.08 GB. The server demonstrates top-tier performance in memory bandwidth, database operations, and transactional workloads, processing 130.32 million Redis SET operations per second and achieving a PassMark CPU score of 103,610.4. While lacking a dedicated GPU, its support for Intel Advanced Matrix Extensions (AMX) enables strong performance in medium-sized LLM inference. This configuration is highly suited for memory-intensive enterprise databases, in-memory caching, and high-throughput data processing.

    Economics

    Average Price per Region

    Prices per Zone

    Lowest Prices

    Performance

    Workload Profiles

    5.4Score
    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) 16,497,910 ops/sec 1,919,289 ops/sec 8.4 50.00% 50.00% +190.00%
    Redis RPS (pipeline=16, SET) 90,225,947 ops/sec 11,728,058 ops/sec 7.5 20.00% 20.00% +50.00%
    PassMark Memory Mark (composite) 2,712 2,416 1.1 10.00% 10.00% +0.96%
    Memory bandwidth (read, 16 MB ~ L3) 617,085 MB/sec 108,466 MB/sec 5.6 10.00% 10.00% +19.00%
    PassMark single-thread CPU 3,252 Mops/s 2,365 Mops/s 1.4 10.00% 10.00% +3.40%

    Memory Bandwidth

    Compression

    OpenSSL

    Passmark CPU Scores

    BENCHMARKSCORE
    Mark
    		 103,610
    Compression
    		 1,673,508
    Encryption
    		 103,380
    Extended Instructions
    		 137,555
    Floating Point Maths
    		 332,950
    Integer Maths
    		 469,021
    Physics
    		 19,568
    Prime Numbers
    		 951
    Single Threaded
    		 3,252
    String Sorting
    		 228,076

    Passmark Memory Scores

    BENCHMARKSCORE
    Memory Mark
    		 2,712
    Database Operations
    		 37,423
    Memory Latency
    		 58
    Memory Read Cached
    		 29,220
    Memory Read Uncached
    		 10,617
    Memory Write
    		 9,884

    Stress-ng Stressors

    CPU integerSingleMultiScaling
    atomic
    		 215.94 117.51 0.54×
    bitops
    		 165,376 8,820,127 53.33×
    branch
    		 156,024 10,681,254 68.46×
    funccall
    		 841.66 48,024 57.06×
    funcret
    		 13,747 968,892 70.48×
    hash
    		 198,180 12,492,402 63.04×
    intmath
    		 72,956 3,720,212 50.99×
    nop
    		 294,145 10,210,795 34.71×
    prime
    		 529,097 28,455,302 53.78×
    rotate
    		 111,373 5,207,527 46.76×
    str
    		 1,061,115 66,782,548 62.94×
    wcs
    		 736,012 42,762,611 58.10×
    CryptographySingleMultiScaling
    crypt
    		 3,798 192,536 50.69×
    CompressionSingleMultiScaling
    jpeg
    		 937.89 51,999 55.44×
    zlib
    		 96.48 4,824 49.99×
    HypervisorSingleMultiScaling
    clock
    		 29,268 177,586 6.07×
    clone
    		 1,426 33,830 23.73×
    context
    		 5,956 434,957 73.02×
    cpu-sched
    		 63.49 6,008 94.62×
    eventfd
    		 35,853 2,486,456 69.35×
    fault
    		 26,148 187,813 7.18×
    fork
    		 5,575 53,678 9.63×
    getrandom
    		 65,417 4,241,168 64.83×
    hrtimers
    		 507,797 7,338,134 14.45×
    mprotect
    		 573,940 19,388,891 33.78×
    mtx
    		 16,706,790 2,884,701,890 172.67×
    munmap
    		 4,321 56,087 12.98×
    null
    		 967,523 67,356,218 69.62×
    pipe
    		 748,703 59,259,255 79.15×
    priv-instr
    		 952,258 14,016,944 14.72×
    resched
    		 10,549,718 8,809,415 0.84×
    switch
    		 44,659 18,055,423 404.30×
    timermix
    		 288,993 18,478,266 63.94×
    tlb-shootdown
    		 7,109 131,219 18.46×
    tsc
    		 1,016,977 68,859,437 67.71×
    usersyscall
    		 808,537 14,041,894 17.37×
    vdso
    		 118,704,944 7,735,653,622 65.17×
    vfork
    		 27,659 156,443 5.66×
    x86syscall
    		 11,657,412 875,186,994 75.08×
    yield
    		 132,634,313 135,575,692 1.02×
    zero
    		 6,467,897 468,927,883 72.50×
    CPU float/SIMDSingleMultiScaling
    eigen
    		 17,791 969,731 54.51×
    expmath
    		 3,894 250,695 64.37×
    fma
    		 1,432,957 82,057,301 57.26×
    fp
    		 234.14 14,770 63.08×
    fractal
    		 7.2 452.08 62.79×
    hyperbolic
    		 4,500 283,834 63.07×
    logmath
    		 5,805 398,606 68.66×
    matrix
    		 5,722 301,889 52.76×
    matrix-3d
    		 768.14 16,825 21.90×
    monte-carlo
    		 334.01 23,639 70.77×
    mpfr
    		 1,962 97,508 49.69×
    powmath
    		 2,202 147,145 66.83×
    trig
    		 2,983 174,071 58.36×
    veccmp
    		 27,910 1,431,558 51.29×
    vecfp
    		 1,729 97,445 56.35×
    vecmath
    		 4,324 273,037 63.14×
    vecshuf
    		 921.22 46,793 50.79×
    vecwide
    		 40,187 2,213,419 55.08×
    AlgorithmsSingleMultiScaling
    bitonicsort
    		 10 522.79 52.28×
    bsearch
    		 312.2 18,989 60.82×
    bubblesort
    		 1.2 57.6 48.00×
    fibsearch
    		 339.54 18,078 53.24×
    heapsort
    		 11.2 663.19 59.21×
    hsearch
    		 9,142 486,745 53.24×
    insertionsort
    		 14 542.8 38.77×
    malloc
    		 3,079,965 82,763,518 26.87×
    mcontend
    		 59.71 15,255 255.49×
    mergesort
    		 34.2 2,120 62.00×
    qsort
    		 24.4 1,537 62.99×
    skiplist
    		 6,038 428,550 70.97×
    tree
    		 2 114.78 57.39×
    MemorySingleMultiScaling
    cache
    		 231,535 9,552,472 41.26×
    cachehammer
    		 208,786 7,905,615 37.86×
    cacheline
    		 46.17 3,122 67.62×
    l1cache
    		 1,217 18,871 15.50×
    memcpy
    		 489.59 21,402 43.71×
    memrate
    		 0.13 11.13 85.62×
    memthrash
    		 196.92 1,899 9.65×
    ptr-chase
    		 36,087,026 1,033,243,551 28.63×
    spinmem
    		 417.79 1,911 4.57×
    stream
    		 3.33 7,414 2226.45×
    vm
    		 206,372 307,905 1.49×

    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
    FX2mds_v2 242 GiB0
    FX4-2mds_v2 284 GiB0
    FX8-2mds_v2 2168 GiB0
    FX4mds_v2 484 GiB0
    FX8-4mds_v2 4168 GiB0
    FX16-4mds_v2 4336 GiB0
    FX12-6mds_v2 6252 GiB0

    Similar Servers

    INSTANCEVENDORvCPUsMEMORYAccelerators

    FX96mds_v2 FAQs