t2a-standard-48 by Google Cloud Platform
(All-cores)
(Single-core)
Specifications
Server Metadata
Vendor ID | gcp |
Server ID | 1220007 |
Name | t2a-standard-48 |
Description | 48 vCPUs, 192 GB RAM |
Family | t2a |
Hw Virt | |
Average Time To Start | 15 |
Status | active |
Observed At | 2026-08-04T19:13:27.260965 |
Availability
| REGION / ID | SPOT | ONDEMAND |
|---|---|---|
Council Bluffs (US) / us-central1 | 0.8655 USD/h | 1.848 USD/h |
Eemshaven (NL) / europe-west4 | 1.1591 USD/h | 1.9413 USD/h |
Jurong West (SG) / asia-southeast1 | 0.6793 USD/h | 2.279 USD/h |
Processor
vCPUs | 48 |
CPU Allocation | Dedicated |
CPU Cores | 48 |
CPU Speed | 3 GHz |
CPU Architecture | arm64 |
CPU Manufacturer | Ampere |
CPU Family | ARMv8 |
CPU Model | Neoverse-N1 |
CPU L1D Cache | 64 KiB |
CPU L1D Cache Total | 3 MiB |
CPU L1I Cache | 64 KiB |
CPU L1I Cache Total | 3 MiB |
CPU L2 Cache | 1 MiB |
CPU L2 Cache Total | 48 MiB |
CPU L3 Cache | 32 MiB |
CPU L3 Cache Total | 32 MiB |
CPU Flags | fp, asimd, evtstrm, aes, pmull, sha1, sha2, crc32, atomics, fphp, asimdhp, cpuid, asimdrdm, lrcpc, dcpop, asimddp, ssbs |
Ecpus | 48 |
Scalability | 100 |
System Resources and Accelerators
| MEMORY | |
|---|---|
Memory Amount | 192 GiB |
Memory Amount Actual | 192 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 arm64 server featuring forty-eight dedicated physical cores and balanced memory configuration for predictable multi-threaded workloads.
Google Cloud Platform t2a-standard-48 is an arm64 general-purpose server instance equipped with 48 dedicated physical cores on an Ampere Altra ARMv8 Neoverse-N1 processor running at 3.0 GHz. It features 192.0 GB of RAM, translating to 4.0 GB of memory per core, with no bundled local storage or GPUs. The instance exhibits strong multi-core CPU performance in stress-ng benchmarks, though single-core processing and text-based operations are relatively weak. Memory bandwidth remains average, with read speeds of 181.23 GB/sec for large uncached blocks. Lacking AVX and AMX extensions, it is not optimized for heavy vector math, which is reflected in its weak performance during large-scale LLM text generation. The t2a-standard-48 is recommended for multi-threaded workloads, static web serving, software compilation, and database operations where physical core isolation and balanced memory-to-CPU ratios are required.
Economics
Performance
Memory Bandwidth
Compression
OpenSSL
Geekbench Single-Core
Geekbench Multi-Core
Passmark CPU Scores
| BENCHMARK | SCORE |
|---|---|
Mark | 27399 |
Compression | 272501 |
Encryption | 7763 |
Extended Instructions | 35712 |
Floating Point Maths | 171834 |
Integer Maths | 226357 |
Physics | 3497 |
Prime Numbers | 214 |
Single Threaded | 1323 |
String Sorting | 156336 |
Passmark Memory Scores
| BENCHMARK | SCORE |
|---|---|
Memory Mark | 1963 |
Database Operations | 20474 |
Memory Latency | 70 |
Memory Read Cached | 11296 |
Memory Read Uncached | 8597 |
Memory Write | 14944 |
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 |
|---|---|---|---|
| t2a-standard-1 | 1 | 4 GiB | 0 |
| t2a-standard-2 | 2 | 8 GiB | 0 |
| t2a-standard-4 | 4 | 16 GiB | 0 |
| t2a-standard-8 | 8 | 32 GiB | 0 |
| t2a-standard-16 | 16 | 64 GiB | 0 |
| t2a-standard-32 | 32 | 128 GiB | 0 |
Similar Servers
| INSTANCE | VENDOR | vCPUs | MEMORY | GPUs |
|---|