Google Price:
Geekbench 3 Benchmark:
Geekbench Results:
http://blog.zorangagic.com/search?q=cloudbenchThe below pricing is based on following:
- All pricing in US$ monthly costs for Linux based VM
- AWS pricing for Sydney region
- Google Compute Engine pricing for Asia region based of custom VM:
- $0.03841 / vCPU
- $0.00515 / GB
*Please note that Google pricing would be slightly lower cost if predefined VMs were chosen instead of custom VM
Generation | Instance Size | vCPUs | Memory | Linux Cost (Monthly) | Google Custom On Demand (Asia) | Google Custom Sustained Use (Asia) | AWS/Google Custom On Demand (Asia) | AWS/Google Custom Sustained Discount (Asia) |
ec2.current | t2.micro | 1 cores | 1.0 GB | $14.40 | $30.30 | $21.21 | 47.53% | 67.90% |
ec2.current | t2.small | 1 cores | 2.0 GB | $28.80 | $35.56 | $24.89 | 81.00% | 115.71% |
ec2.current | t2.medium | 2 cores | 4.0 GB | $57.60 | $71.11 | $49.78 | 81.00% | 115.71% |
ec2.current | c4.large | 2 cores | 3.75 GB | $103.68 | $70.17 | $49.12 | 147.75% | 211.07% |
ec2.current | t2.large | 2 cores | 8.0 GB | $115.20 | $86.15 | $60.30 | 133.73% | 191.04% |
ec2.current | m4.large | 2 cores | 8.0 GB | $127.44 | $86.15 | $60.30 | 147.94% | 211.34% |
ec2.current | r3.large | 2 cores | 15.25 GB | $151.20 | $113.39 | $79.37 | 133.34% | 190.49% |
ec2.current | c4.xlarge | 4 cores | 7.5 GB | $208.08 | $140.35 | $98.24 | 148.26% | 211.80% |
ec2.current | m4.xlarge | 4 cores | 16.0 GB | $254.88 | $172.29 | $120.60 | 147.94% | 211.34% |
ec2.current | r3.xlarge | 4 cores | 30.5 GB | $302.40 | $226.78 | $158.75 | 133.34% | 190.49% |
ec2.current | c4.2xlarge | 8 cores | 15.0 GB | $416.16 | $280.70 | $196.49 | 148.26% | 211.80% |
ec2.current | m4.2xlarge | 8 cores | 32.0 GB | $509.76 | $344.58 | $241.21 | 147.94% | 211.34% |
ec2.current | r3.2xlarge | 8 cores | 61.0 GB | $604.80 | $453.57 | $317.50 | 133.34% | 190.49% |
ec2.current | d2.xlarge | 4 cores | 30.5 GB | $626.40 | $226.78 | $158.75 | 276.21% | 394.59% |
ec2.current | g2.2xlarge | 8 cores | 15.0 GB | $646.56 | $280.70 | $196.49 | 230.34% | 329.06% |
ec2.current | i2.xlarge | 4 cores | 30.5 GB | $732.96 | $226.78 | $158.75 | 323.20% | 461.71% |
ec2.current | c4.4xlarge | 16 cores | 30.0 GB | $831.60 | $561.39 | $392.98 | 148.13% | 211.62% |
ec2.current | m4.4xlarge | 16 cores | 64.0 GB | $1,019.52 | $689.17 | $482.42 | 147.94% | 211.34% |
ec2.current | r3.4xlarge | 16 cores | 122.0 GB | $1,209.60 | $907.13 | $634.99 | 133.34% | 190.49% |
ec2.current | d2.2xlarge | 8 cores | 61.0 GB | $1,252.80 | $453.57 | $317.50 | 276.21% | 394.59% |
ec2.current | i2.2xlarge | 8 cores | 61.0 GB | $1,465.20 | $453.57 | $317.50 | 323.04% | 461.49% |
ec2.current | c4.8xlarge | 36 cores | 60.0 GB | $1,663.20 | $1,234.95 | $864.46 | 134.68% | 192.40% |
ec2.current | r3.8xlarge | 32 cores | 244.0 GB | $2,419.20 | $1,814.27 | $1,269.99 | 133.34% | 190.49% |
ec2.current | d2.4xlarge | 16 cores | 122.0 GB | $2,505.60 | $907.13 | $634.99 | 276.21% | 394.59% |
ec2.current | m4.10xlarge | 40 cores | 160.0 GB | $2,548.80 | $1,722.92 | $1,206.04 | 147.94% | 211.34% |
ec2.current | g2.8xlarge | 32 cores | 60.0 GB | $2,586.24 | $1,122.79 | $785.95 | 230.34% | 329.06% |
ec2.current | i2.4xlarge | 16 cores | 122.0 GB | $2,930.40 | $677.89 | $474.52 | 432.28% | 617.54% |
ec2.current | d2.8xlarge | 36 cores | 244.0 GB | $5,011.20 | $1,926.43 | $1,348.50 | 260.13% | 371.61% |
ec2.current | i2.8xlarge | 32 cores | 244.0 GB | $5,860.80 | $1,814.27 | $1,269.99 | 323.04% | 461.49% |
ec2.old | t1.micro | 1 cores | 0.613 GB | $14.40 | $30.30 | $21.21 | 47.53% | 67.90% |
ec2.old | m1.small | 1 cores | 1.7 GB | $41.76 | $34.43 | $24.10 | 121.29% | 173.27% |
ec2.old | m3.medium | 1 cores | 3.75 GB | $66.96 | $42.13 | $29.49 | 158.92% | 227.03% |
ec2.old | m1.medium | 1 cores | 3.75 GB | $84.24 | $42.13 | $29.49 | 199.94% | 285.62% |
ec2.old | c3.large | 2 cores | 3.75 GB | $95.04 | $70.17 | $49.12 | 135.43% | 193.48% |
ec2.old | c1.medium | 2 cores | 1.7 GB | $118.08 | $62.47 | $43.73 | 189.02% | 270.03% |
ec2.old | m3.large | 2 cores | 7.5 GB | $133.92 | $84.27 | $58.99 | 158.92% | 227.03% |
ec2.old | m1.large | 2 cores | 7.5 GB | $167.76 | $84.27 | $58.99 | 199.08% | 284.40% |
ec2.old | c3.xlarge | 4 cores | 7.5 GB | $190.80 | $140.35 | $98.24 | 135.95% | 194.21% |
ec2.old | m2.xlarge | 2 cores | 17.1 GB | $213.12 | $120.34 | $84.24 | 177.09% | 252.99% |
ec2.old | m3.xlarge | 4 cores | 15.0 GB | $267.84 | $168.53 | $117.97 | 158.92% | 227.03% |
ec2.old | m1.xlarge | 4 cores | 15.0 GB | $336.24 | $168.53 | $117.97 | 199.51% | 285.01% |
ec2.old | c3.2xlarge | 8 cores | 15.0 GB | $380.88 | $280.70 | $196.49 | 135.69% | 193.84% |
ec2.old | m2.2xlarge | 4 cores | 34.2 GB | $426.24 | $240.69 | $168.48 | 177.09% | 252.99% |
ec2.old | c1.xlarge | 8 cores | 7.0 GB | $471.60 | $250.63 | $175.44 | 188.16% | 268.81% |
ec2.old | m3.2xlarge | 8 cores | 30.0 GB | $536.40 | $337.07 | $235.95 | 159.14% | 227.34% |
ec2.old | c3.4xlarge | 16 cores | 30.0 GB | $761.76 | $561.39 | $392.98 | 135.69% | 193.84% |
ec2.old | m2.4xlarge | 8 cores | 68.4 GB | $851.76 | $481.38 | $336.96 | 176.94% | 252.78% |
ec2.old | c3.8xlarge | 32 cores | 60.0 GB | $1,524.24 | $1,122.79 | $785.95 | 135.76% | 193.94% |
ec2.old | hs1.8xlarge | 17 cores | 117.0 GB | $4,010.40 | $916.38 | $641.47 | 437.63% | 625.19% |
ec2.old | cc2.8xlarge | 32 cores | 60.5 GB | unavailable | $1,124.67 | $787.27 | ||
ec2.old | cg1.4xlarge | 16 cores | 22.5 GB | unavailable | $533.21 | $373.25 | ||
ec2.old | cr1.8xlarge | 32 cores | 244.0 GB | unavailable | $1,814.27 | $1,269.99 | ||
ec2.old | hi1.4xlarge | 16 cores | 60.5 GB | unavailable | $676.01 | $473.21 |
AWS Storage | Price (per GB / month) | Google Storage | Price (per GB / month) |
AWS standard EBS (standard) | $0.08 | Standard provisioned space | $0.04 |
AWS SSD EBS (gp2) | $0.12 | SSD provisioned space | $0.17 |
EBS snapshot | $0.105 | Snapshot storage | $0.026 |
Standard S3 | $0.0330 | Standard Object storage | $0.026 |
Infrequent Access S3 | $0.019 | Durable RA Object storage | $0.02 |
Glacier | $0.0120 | Nearline Object storage | $0.01 |
It is interesting that Google continue with their aggressive Moore's law cloud pricing policy (discount 20-30% annually) and AWS recently have only had modest modest price decreases since Google announced their Moore's law pricing policy:
"As such, Google introduced a pricing model that will more closely follow the trends in Moore’s Law."
Google currently faces the following challenges:
- Currently not available in Australian region:
- Some applications require local latency, the latency between Australia and Google Taiwan region is high at 136ms hence not suitable for many applications
- Data sovereignty issues
- No clear roadmap when Google Compute Engine will be in Australia
- Even though Google Console is very easy to use once it comes to automation Google APIs are completely different, hence it would take significant time and resources to learn and rewrite/test code to run on Google Compute Engine. This is a huge obstacle.
- Google does not have same broad cloud services as AWS, this further complicates migration to Google Compute Engine as any AWS applications that are heavily integrated with AWS services will need to be rewritten, this again can be very costly.
- Some AWS services are simply are not available on Google Cloud Platform.
- Multi Cloud strategy is complex to implement as IT resources have to learn and support multiple platform that are continually changing. This effort would have significant overhead and carries significant additional cost and complexity. Why not just pick one single cloud for IaaS that offers best aggregate features, stability, performance, security and price. Yes, IaaS clouds are completely proprietary and customers are completely locked in vendors cloud yet we used proprietary products for decades in the past with vendors like Cisco, IBM, HP, EMC, Netapp, Brocade, Oracle, Microsoft, VMware etc...when you purchase expensive, proprietary products you are stuck with the solution for a couple of years. Most enterprise IT customers have been happy to choose proprietary products for decades, hence how is proprietary cloud any different?
No comments:
Post a Comment