Benchmarking Google Cloud Compute Engine

How Fast IsGoogle Cloud?

Comprehensive performance benchmarks across every Compute Engine VM family. CPU, memory, network, storage, and GPU -- measured and compared.

0
VM Families
0
VM Series
0+
PFLOPS GPU
0
Gbps Network
Explore Benchmarks
Web and app servers

General Purpose

Balanced compute, memory, and networking for a wide range of workloads. From web servers to development environments, these VMs deliver predictable performance at competitive price points.

0pts
Single-Core CPU
N2 series -- Geekbench 6
0pts
Multi-Core CPU
N2D series -- Geekbench 6
0Gbps
Network Bandwidth
N2D series
0GB/s
STREAM Bandwidth
N2D series

Single-Core Performance

Geekbench 6 -- higher is better

Multi-Core Performance

Geekbench 6 -- higher is better

Performance Profile

Relative strengths across categories (normalized)

Available Series

SeriesvCPUsMemoryNetwork
E22-321-128 GBUp to 16 Gbps
N22-1282-864 GBUp to 100 Gbps
N2D2-2242-896 GBUp to 100 Gbps
Tau T2D1-604-240 GBUp to 32 Gbps
High-performance computing (HPC)

Compute Optimized

Highest per-core performance on Google Cloud. Built for compute-intensive workloads that need maximum single-thread speed and sustained clock frequencies.

0pts
Single-Core CPU
C3 series -- Geekbench 6
0pts
Multi-Core CPU
H3 series -- Geekbench 6
0Gbps
Network Bandwidth
H3 series
0GB/s
STREAM Bandwidth
C3D series

Single-Core Performance

Geekbench 6 -- higher is better

Multi-Core Performance

Geekbench 6 -- higher is better

Performance Profile

Relative strengths across categories (normalized)

Available Series

SeriesvCPUsMemoryNetwork
C34-1768-1408 GBUp to 100 Gbps
C3D4-3608-2880 GBUp to 100 Gbps
H388352 GB200 Gbps
In-memory databases (SAP HANA, Redis)

Memory Optimized

Designed for workloads that need large amounts of memory relative to vCPUs. Ideal for in-memory databases, real-time analytics, and large caching layers.

0pts
Single-Core CPU
M3 series -- Geekbench 6
0pts
Multi-Core CPU
M3 series -- Geekbench 6
0Gbps
Network Bandwidth
M3 series
0GB
Max Memory Per VM
M3 series

Performance Profile

Relative strengths across categories (normalized)

Available Series

SeriesvCPUsMemoryNetwork
M332-128960-3904 GBUp to 100 Gbps
ML model training

Accelerator Optimized

Purpose-built for machine learning training, inference, and GPU-accelerated workloads. Powered by NVIDIA GPUs with high-bandwidth interconnects.

0pts
Single-Core CPU
G2 series -- Geekbench 6
0Gbps
Network Bandwidth
A3 series
0Gbps
GPU Interconnect Bandwidth
A3 series

Performance Profile

Relative strengths across categories (normalized)

Available Series

SeriesvCPUsMemoryNetwork
A32081872 GBUp to 1800 Gbps (GPU-to-GPU)
G24-9616-384 GBUp to 100 Gbps
High-performance databases

Storage Optimized

Optimized for workloads that require high storage throughput and IOPS. Local SSDs deliver ultra-low latency for database engines and data-intensive applications.

0pts
Single-Core CPU
Z3 series -- Geekbench 6
0pts
Multi-Core CPU
Z3 series -- Geekbench 6
0Gbps
Network Bandwidth
Z3 series

Performance Profile

Relative strengths across categories (normalized)

Available Series

SeriesvCPUsMemoryNetworkStorage IOPS
Z388-176704-1408 GBUp to 100 GbpsUp to 2.4M read IOPS
The Big Picture

Google Cloud by the Numbers

Peak performance across the entire Compute Engine lineup.

0PFLOPS
Peak FP8 Performance
A3 series with 8x NVIDIA H100 GPUs
0GB
Max Memory Per VM
M3 series -- for the largest in-memory workloads
0IOPS
Peak Storage IOPS
Z3 series -- local SSD read performance
0Gbps
GPU Interconnect
A3 series -- NVLink GPU-to-GPU bandwidth
0vCPUs
Max vCPUs Per VM
C3D series -- for massive parallel workloads
0pts
Best Single-Core
C3 series -- Geekbench 6 single-core score
Head-to-Head

Compare All VM Series

Sort by any metric to find the best VM series for your workload. Click column headers to sort.

Series
Family
Single-Core
Multi-Core
Network (Gbps)
Memory BW (GB/s)
C3Compute Optimized
1,850
14,200
100
45.8
H3Compute Optimized
1,720
22,400
200
--
Z3Storage Optimized
1,700
20,100
100
--
C3DCompute Optimized
1,680
18,500
100
52.3
G2Accelerator Optimized
1,540
--
100
--
N2General Purpose
1,520
8,750
100
32.4
M3Memory Optimized
1,480
11,600
100
68.4
N2DGeneral Purpose
1,410
12,800
100
38.6
Tau T2DGeneral Purpose
1,350
9,200
32
--
E2General Purpose
1,180
4,520
16
18.2
A3Accelerator Optimized----
1,800
--