Essentially, RAM comes in small sizes (8, 16, 32 GB) but storage is in TB? Why is
this? Would more RAM be better?
In: 0
The way modern computers are designed, they require both RAM (working memory to facilitate executing programs) and storage (long term memory, to persist data between power cycles so it can be used later). So we cannot make generalised statements about one being better than the other.
Today, both types of storage retain data by trapping electric charge in tiny electronic components.
Due to the nature of physics (subatomic behaviour of electricity, limitations of the materials used), there is a trade off between being fast and being persistent. Random Access Memory and Solid State Drives take opposite extremes of these trade-offs, with RAM being orders of magnitude faster than SSD, but needs constant power to retain data; while SSDs can retain data for years without power, but the electronics physically take longer to change state (so they are slower) and also degrade after a certain number of changes.
More importantly however, RAM is much more expensive per-GB compared to SSDs, that’s why you’re seeing such size differences. Luckily (or perhaps, intentionally), computers are designed to work efficiently with smaller amounts of faster (and thus more expensive) components and larger amounts of slower components.
I suspect if we were able to build a data storage technology that was both as fast and reliable as RAM but as cheap and persistent as SSDs, computers would be designed differently (though you probably still want a section of memory that can reset program state in the event of errors, which is the role RAM serves now). But that hasn’t been the case since we first invented computers, and as mentioned the physics involved makes it unlikely.
A single 64GB stick of RAM costs around $400 US (depending on specs it can be less or more, but not by a lot in most cases).
For that same $400 you can get a 2TB solid state hard drive or an 18TB spinning disk hard drive.
Price per GB:
RAM: $6.25
SSD: $0.20
HDD: $0.02
Basically, each tier is an order of magnitude more expensive, but its also much faster.
This is because there’s no technology that would provide speed, capacity and permanent storage. You can choose at best only two of this characteristics. If we would only use RAM, then that’d be very inconvenient, as you’d have to provide power to it to avoid loosing data, and such storage would be very costly per gigabyte. But very fast, yeah. If you’d use SSD instead of RAM, then data reading would take quite a long delays and your programs would perform much worse. There’s also multiple cache memory levels inside CPU – they’re much faster than RAM, but they’re even more limiting in size and cost.
But by combining HDD, SSD, RAM and cache you can achieve best of them all – HDD for long term large storage, for things that you don’t access very often, SSD for system files or things that you read frequently, RAM for applications data that they use in process, and cache for data that processor is processing right now. This way, programs or users could decide where to store what data depending on how important to access it fast and how much data there’s to store.
If at one day there would appear technology that could access large quantities of data with miniscule delay and is cheap – that technology could as well replace both RAM and storage. But that won’t happen any time soon.
32GB of DDR4 is ~100$. 1000GB of SSD is ~100$. 4000GB of HDD is ~100$. It’s simply a matter of RAM being expensive as fuck. There are server computers that can load multiple terabytes of RAM sticks, but they cost five-six figures kitted out for obvious reasons.
Latest Answers