a3-highgpu-2g by Google Cloud Platform
(All-cores)
(Single-core)
Specifications
Server Metadata
Vendor ID | gcp |
Server ID | 1720362 |
Name | a3-highgpu-2g |
Description | Accelerator Optimized: 2 NVIDIA H100 GPU, 52 vCPUs, 468GB RAM |
Family | a3 |
Hw Virt | |
Average Time To Start | 35 |
Status | active |
Observed At | 2026-07-21T22:45:10.240053 |
Availability
| REGION / ID | SPOT | ONDEMAND |
|---|---|---|
Ashburn (US) / us-east4 | 0.9663 USD/h | 2.365 USD/h |
Council Bluffs (US) / us-central1 | 1.353 USD/h | 2.365 USD/h |
The Dalles (US) / us-west1 | 1.353 USD/h | 2.365 USD/h |
Mumbai (IN) / asia-south1 | 1.0074 USD/h | 2.4596 USD/h |
Eemshaven (NL) / europe-west4 | 0.9708 USD/h | 2.4833 USD/h |
Hamina (FI) / europe-north1 | 0.2633 USD/h | 2.6041 USD/h |
St. Ghislain (BE) / europe-west1 | 1.0154 USD/h | 2.6041 USD/h |
Las Vegas (US) / us-west4 | 0.9931 USD/h | 2.6637 USD/h |
Columbus (US) / us-east5 | 0.9931 USD/h | 2.6637 USD/h |
Changhua County (TW) / asia-east1 | 1.2826 USD/h | 2.7385 USD/h |
Paris (FR) / europe-west9 | 0.6193 USD/h | 2.7434 USD/h |
Frankfurt (DE) / europe-west3 | 1.1408 USD/h | 2.7907 USD/h |
Delhi (IN) / asia-south2 | 0.829 USD/h | 2.8406 USD/h |
Jurong West (SG) / asia-southeast1 | 1.3074 USD/h | 2.9177 USD/h |
Sydney (AU) / australia-southeast1 | 0.7402 USD/h | 2.9563 USD/h |
Tokyo (JP) / asia-northeast1 | 0.6133 USD/h | 3.0372 USD/h |
Processor
vCPUs | 52 |
CPU Allocation | Dedicated |
CPU Cores | 26 |
CPU Speed | 2.7 GHz |
CPU Architecture | x86_64 |
CPU Manufacturer | Intel |
CPU Family | Xeon |
CPU Model | 8481C |
CPU L1D Cache | 48 KiB |
CPU L1D Cache Total | 1 MiB |
CPU L1I Cache | 32 KiB |
CPU L1I Cache Total | 832 KiB |
CPU L2 Cache | 2 MiB |
CPU L2 Cache Total | 52 MiB |
CPU L3 Cache | 105 MiB |
CPU L3 Cache Total | 105 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, nonstop_tsc, cpuid, tsc_known_freq, pni, pclmulqdq, vmx, ssse3, fma, cx16, pcid, sse4_1, sse4_2, x2apic, movbe, popcnt, aes, xsave, avx, f16c, rdrand, hypervisor, lahf_lm, abm, 3dnowprefetch, ssbd, ibrs, ibpb, stibp, ibrs_enhanced, tpr_shadow, flexpriority, ept, vpid, ept_ad, fsgsbase, tsc_adjust, bmi1, avx2, smep, bmi2, erms, invpcid, rtm, avx512f, avx512dq, rdseed, adx, smap, avx512ifma, clflushopt, clwb, avx512cd, sha_ni, avx512bw, avx512vl, xsaveopt, xsavec, xgetbv1, xsaves, avx_vnni, avx512_bf16, arat, vnmi, avx512vbmi, umip, avx512_vbmi2, gfni, vaes, vpclmulqdq, avx512_vnni, avx512_bitalg, avx512_vpopcntdq, la57, rdpid, cldemote, movdiri, movdir64b, fsrm, md_clear, serialize, tsxldtrk, amx_bf16, avx512_fp16, amx_tile, amx_int8, arch_capabilities |
Ecpus | 22.7 |
Scalability | 87.31 |
System Resources and Accelerators
| MEMORY | |
|---|---|
Memory Amount | 468 GiB |
Memory Amount Actual | 468 GiB |
| GPU | |
|---|---|
GPU Count | 2 |
GPU Memory Min | 80 GiB |
GPU Memory Total | 159 GiB |
GPU Manufacturer | NVIDIA |
GPU Family | Hopper |
GPU Model | H100 |
GPUs |
|
| STORAGE | |
|---|---|
Storage Size | 1608 GB |
Storage Type | nvme ssd |
Storages |
|
| NETWORK | |
|---|---|
Inbound Traffic | 0 GB/month |
Outbound Traffic | 0 GB/month |
IPv4 | 0 |
CPU and System Topology
Server Description
An accelerator-optimized platform combining dual H100 GPUs, dedicated Intel Xeon processors, and high-bandwidth memory for demanding machine learning workloads.
Google Cloud Platform a3-highgpu-2g is an accelerator-optimized server featuring two NVIDIA Hopper H100 GPUs with 159 GB of total VRAM, 52 dedicated Intel Xeon 8481C vCPUs operating at 2.7 GHz, and 468.0 GB of system memory. It includes 1608 GB of local NVMe SSD storage. In benchmark testing, the server demonstrates top-tier performance in LLM inference, processing 5644.37 tokens per second for 7B model prompts and 194.27 tokens per second for 7B model text generation. It also exhibits strong memory bandwidth, reaching 3828.33 GB/sec for cached reads, and a strong Passmark CPU score of 54314.03. The server provides high resource density for GPU-accelerated tasks, though it lacks complimentary public IPv4 addresses. It is designed for machine learning inference, data processing, and database operations.
Economics
Performance
Memory Bandwidth
Compression
OpenSSL
Geekbench Single-Core
Geekbench Multi-Core
Passmark CPU Scores
| BENCHMARK | SCORE |
|---|---|
Mark | 54314 |
Compression | 716947 |
Encryption | 43985 |
Extended Instructions | 46057 |
Floating Point Maths | 142147 |
Integer Maths | 199432 |
Physics | 6503 |
Prime Numbers | 349 |
Single Threaded | 2720 |
String Sorting | 97370 |
Passmark Memory Scores
| BENCHMARK | SCORE |
|---|---|
Memory Mark | 2230 |
Database Operations | 23460 |
Memory Latency | 74 |
Memory Read Cached | 25230 |
Memory Read Uncached | 9444 |
Memory Write | 8322 |
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
| INSTANCE | vCPUs | MEMORY | GPUs |
|---|---|---|---|
| a3-highgpu-1g | 26 | 234 GiB | 1 |
| a3-highgpu-4g | 104 | 936 GiB | 4 |
| a3-highgpu-8g | 208 | 1872 GiB | 8 |
| a3-megagpu-8g | 208 | 1872 GiB | 8 |
| a3-edgegpu-8g | 208 | 1872 GiB | 8 |
| a3-edgegpu-8g-nolssd | 208 | 1872 GiB | 8 |
| a3-ultragpu-8g | 224 | 2952 GiB | 8 |
Similar Servers
| INSTANCE | VENDOR | vCPUs | MEMORY | GPUs |
|---|