n4d-highmem-96 by Google Cloud Platform

n4d-highmem-96 is a 96 vCPUs, 768 GB RAM server offered by Google Cloud Platform with 96 vCPUs, 768 GiB of memory and 0 GB of storage. The pricing starts at 1.623 USD per hour.
96 vCPU
768 GiB Memory
Spare SCore
164,007
(All-cores)
3,423
(Single-core)

Specifications

Server Metadata

Vendor ID
gcp
Server ID
1751108
Name
n4d-highmem-96
Description
96 vCPUs, 768 GB RAM
Family
n4d
Hw Virt
Average Time To Start
21
Status
active
Observed At
2026-09-13T05:02:37.852658

Availability

REGION / IDSPOTONDEMAND
Mumbai (IN) / asia-south1
1.863 USD/h 4.0593 USD/h
Moncks Corner (US) / us-east1
1.7432 USD/h 5.3366 USD/h
The Dalles (US) / us-west1
2.0185 USD/h 5.3366 USD/h
Columbus (US) / us-east5
2.583 USD/h 5.3366 USD/h
Ashburn (US) / us-east4
2.7126 USD/h 5.3366 USD/h
Council Bluffs (US) / us-central1
2.8479 USD/h 5.3366 USD/h
Delhi (IN) / asia-south2
1.9079 USD/h 5.5482 USD/h
Eemshaven (NL) / europe-west4
2.4515 USD/h 5.607 USD/h
St. Ghislain (BE) / europe-west1
1.7622 USD/h 5.8726 USD/h
Hamina (FI) / europe-north1
1.9217 USD/h 5.8726 USD/h
Las Vegas (US) / us-west4
1.623 USD/h 6.0122 USD/h
London (GB) / europe-west2
2.6646 USD/h 6.0811 USD/h
Frankfurt (DE) / europe-west3
3.0461 USD/h 6.3005 USD/h
Salt Lake City (US) / us-west3
3.8473 USD/h 6.416 USD/h
Jurong West (SG) / asia-southeast1
3.3483 USD/h 6.5811 USD/h

Processor

vCPUs
96
CPU Allocation
Dedicated
CPU Cores
48
CPU Architecture
x86_64
CPU Manufacturer
AMD
CPU Family
EPYC
CPU Model
9B45
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
1 MiB
CPU L2 Cache Total
48 MiB
CPU L3 Cache
32 MiB
CPU L3 Cache Total
192 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, constant_tsc, rep_good, nopl, xtopology, nonstop_tsc, cpuid, extd_apicid, tsc_known_freq, pni, pclmulqdq, ssse3, fma, cx16, pcid, sse4_1, sse4_2, x2apic, movbe, popcnt, aes, xsave, avx, f16c, rdrand, hypervisor, lahf_lm, cmp_legacy, cr8_legacy, abm, sse4a, misalignsse, 3dnowprefetch, osvw, topoext, ssbd, ibrs, ibpb, stibp, ibrs_enhanced, 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, rdpid, movdiri, movdir64b, avx512_vp2intersect, flush_l1d
Ecpus
47.9
Scalability
99.79

System Resources and Accelerators

MEMORY
Memory Amount
768 GiB
Memory Amount Actual
768 GiB
GPU
GPU Count
0
GPU Memory Min
0 MiB
GPU Memory Total
0 MiB
GPUs
    STORAGE
    Storage Size
    0 GB
    Storages
      NETWORK
      Inbound Traffic
      0 GB/month
      Outbound Traffic
      0 GB/month
      IPv4
      0

      CPU and System Topology

      Server Description

      An x86_64 server featuring ninety-six vCPUs and 768 GB RAM for memory-intensive databases and multi-threaded enterprise workloads.

      Memory OptimizedGeneral Purpose

      Google Cloud Platform n4d-highmem-96 is an x86_64 memory-optimized server featuring 96 vCPUs and 768 GB of RAM, powered by a dedicated AMD EPYC 9B45 processor. With 48 physical cores and a 1:8 vCPU-to-RAM ratio, it delivers top-tier CPU and memory performance. Benchmarks show top 10% results in multi-core processing, Redis operations, and static web serving, alongside strong LLM inference capabilities. However, the server lacks local storage and GPU accelerators, requiring network-attached storage. It is designed for memory-intensive databases, software compilation, and heavy multi-threaded workloads.

      Economics

      Average Price per Region

      Prices per Zone

      Lowest Prices

      Performance

      Workload Profiles

      5.85Score
      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) 19,732,157 ops/sec 1,917,397 ops/sec 10.3 50.00% 50.00% +221.00%
      Redis RPS (pipeline=16, SET) 105,355,522 ops/sec 11,681,410 ops/sec 9.02 20.00% 20.00% +55.30%
      PassMark Memory Mark (composite) 2,900 2,418 1.2 10.00% 10.00% +1.84%
      Memory bandwidth (read, 16 MB ~ L3) 312,024 MB/sec 108,140 MB/sec 2.89 10.00% 10.00% +11.20%
      PassMark single-thread CPU 3,419 Mops/s 2,365 Mops/s 1.45 10.00% 10.00% +3.79%

      Memory Bandwidth

      Compression

      OpenSSL

      Geekbench Single-Core

      Score: 2,542

      Geekbench Multi-Core

      Score: 26,419

      Passmark CPU Scores

      BENCHMARKSCORE
      Mark
      		 111595
      Compression
      		 1961732
      Encryption
      		 111910
      Extended Instructions
      		 151166
      Floating Point Maths
      		 343606
      Integer Maths
      		 575609
      Physics
      		 17612
      Prime Numbers
      		 850
      Single Threaded
      		 3419
      String Sorting
      		 243836

      Passmark Memory Scores

      BENCHMARKSCORE
      Memory Mark
      		 2900
      Database Operations
      		 37266
      Memory Latency
      		 78
      Memory Read Cached
      		 31105
      Memory Read Uncached
      		 28679
      Memory Write
      		 22531

      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

      INSTANCEvCPUsMEMORYGPUs
      n4d-highcpu-2 24 GiB0
      n4d-standard-2 28 GiB0
      n4d-highmem-2 216 GiB0
      n4d-highcpu-4 48 GiB0
      n4d-standard-4 416 GiB0
      n4d-highmem-4 432 GiB0
      n4d-highcpu-8 816 GiB0

      Similar Servers

      INSTANCEVENDORvCPUsMEMORYGPUs

      n4d-highmem-96 FAQs