A discussion on whether Framework Laptop’s modular design has performance implications; AWS grilled by CloudFlare on overcharging for traffics.
victor9000: I really like the form factor of the XPS 13, but now I'm stuck with the complete inability to upgrade RAM. In what universe is that normal?
tomxor: In a universe where you want high-performance energy efficient RAM.
Premium ultracompact laptops have long moved to LPDDR which is both faster and consumes significantly less energy than regular modular DDR.
And LPDDR does not come in DIMMs.
And then you have Apple who are using custom package technology with custom RAM chips to cut down RAM power consumption to under one watt active
You simply can’t have this with socketed RAM.
Modularity is not a free lunch, as much as some people like to pretend.
Charging for Stocks, Paying for Flows
AWS charges customers based on the amount of data delivered
To visualize that, imagine data is water.
AWS fills a bucket full of water and then charges you based on how much water is in the bucket.
This is known as charging based on “stocks.”
AWS pays for bandwidth based on the capacity of their network.
The base unit of wholesale bandwidth is priced as one Megabit per second per month (1 Mbps).
Typically, a provider like AWS, will pay for bandwidth on a monthly fee based on the number of Mbps that their network uses at its peak capacity.
AWS doesn't pay for the amount of water that ends up in their customers' buckets,
but rather the capacity based on the diameter of the “hose” that is used to fill them.
This is known as paying for “flows.”
Translating Flows to Stocks
A 1 Mbps connection (think of it as the "hose") can transfer 0.3285 TB (328GB)
if utilized to its fullest capacity over the course of a month
you can use this as the base unit of their bandwidth costs,
and compare it against what they charge a customer to deliver 1 Terabyte (1TB),
in order to figure out the AWS bandwidth markup.
Wholesale bandwidth is also billed at the 95th percentile.
That effectively cuts off the peak hour or so of use every day.
That means a 1 Mbps connection running at 100% can actually likely transfer closer to 0.3458 TB (346GB) per month.
AWS can't run all their connections at 100% utilization 24x7 for a month.
Instead, they'll have some average utilization per transit connection in any month.
likely run at between 20% and 40% average utilization.
To be conservative, we’ve assumed that AWS’s average utilization is the bottom of that range (20%)
Based on these assumptions, here's our best estimate of AWS’s effective markup for egress bandwidth on a per-region basis.
Region: Customers pay x times AMZN’s costs
South America (Sao Paulo) 21X
Japan (Tokyo) 17x
Australia (Sydney) 8x
India (Mumbai) 8x
South Korea (Seoul) 3.5x
The older a market is, the more Amazon wrings from its customers in egregious egress markups
AWS Stands Alone In Not Passing On Savings to Customers
For the bandwidth that they exchange with a network like Cloudflare, where they are directly connected (settlement-free peered) over a private network interface (PNI), there are no meaningful incremental costs and their effective margins are nearly infinite.
Add in the effect of rebates Amazon collects from colocation providers who charge cross connect fees to customers, and the effective markup is likely even higher.
Some other cloud providers take into account that their costs are lower when passing over peering connections.
Both Microsoft Azure and Google Cloud will substantially discount egress charges for their mutual Cloudflare customers.
Members of the Bandwidth Alliance waive bandwidth charges for mutual Cloudflare customers.
Alibaba, Automattic, Backblaze, Cherry Servers, Dataspace, DNS Networks, DreamHost, HEFICED, Kingsoft Cloud, Liquid Web, Scaleway, Tencent, Vapor, Vultr, Wasabi, and Zenlayer —
we invited AWS to be a part of the Bandwidth Alliance, and they politely declined.
During the last ten years, industry wholesale transit prices have fallen an average of 23% annually.
93% less expensive than 10 years ago.
AWS’s egress fees over that same period have fallen by only 25%.
AWS’s Hotel California Pricing
AWS charge for data transferred out of their network but not for data transferred into their network
Wholesale bandwidth is symmetrical.
if you purchase a 1 Mbps (1 Megabit per second) connection, then you have the capacity to send 1 Megabit out and receive another 1 Megabit in every second.
If you receive 1 Mbps in and simultaneously 1 Mbps out, you pay the same price as if you receive 1 Mbps in and 0 Mbps out or 0 Mbps in and 1 Mbps out.
And yet, they charge customers more to take data out than put it in.
All the data above is derived from what’s published on AWS’s simple pricing calculator.
There’s no doubt that some large customers are able to negotiate lower prices.
But these are the prices charged to small businesses and startups by default.
mdasen: If it's expensive to move data out of AWS, it's not just about making it hard for customers to leave AWS. It means that any third-party service that wants to sell to AWS customers must also use AWS.
There are other reasons to want a third party to use AWS if you're on AWS. Still, the egress pricing seems to make it very hard for third-party tech providers not to use AWS if their customers are using it.
IceHegel: The post claims an 8000% markup but others have pointed out that AWS is in line with other cloud providers and the same math shows cloudflare with a ~3800% markup.
AWS Egress - $0.09/GB
Azure Egress - $0.0875/GB
GCP Egress - $0.11/GB
Alibaba Egress - $0.123
Cloudflare - $0.09/GB pre-April 2021 $0.045 now