Prasad Sankar

So - one of my close friends has had the most heartbreaking experience of his lifetime with someone who he thought was his girlfriend. Man got played like a PS4 in Covid y'all, he was talking to this girl for about 8 months, and he really thought that he's got her undivided attention, and secretly, he fell in love with her. He really thought that he has all of her heart, soul and spirit, but then - doomsday arrived for him. It turns out, this girl had him in genjutsu of the HIGHEST order. She was talking to not one, not two, but THREE lads at the same damn time. And all those thought that they were the only ones who had her - JUST like multiple processes believe they have some control and access over the entire memory space, while in reality, it's all just virtual memory space that is mapped to physical memory that's managed by something called to MEMORY MANAGEMENT UNIT Now that you are here, let's see how this plays out in the first place - Whenever you execute a program, the OS creates a process with its own virtual address space. Instead of directly accessing physical RAM, the process operates using virtual addresses. So, Process A can access virtual address 0x1000, and Process B can access that exact same address, yet both can refer to completely different physical locations. How? The MMU translates virtual addresses into physical addresses using information maintained in page tables. Memory is divided into fixed-size blocks called pages (virtual memory) and frames (physical memory). For example, with 4 KiB pages: Virtual address: 0x1234 Virtual page: 0x1 Page offset: 0x234 If virtual page 0x1 maps to physical frame 0x8, the resulting physical address is: 0x8000 + 0x234 = 0x8234 Simple enough, right? But looking up page tables for every memory access would be expensive. Here, we et help from something called as the Translation Lookaside Buffer (TLB) - a small, fast hardware cache that stores recently used address translations. A TLB hit means the translation is readily available. A TLB miss means the system needs to retrieve it through the applicable page-table-walking mechanism. And no, a TLB miss is NOT the same as a page fault. A page fault occurs when a memory access requires OS intervention - perhaps to allocate a page, retrieve it from storage, establish a mapping, or reject an invalid access. But virtual memory does more than translate addresses. It provides process isolation and memory protection. Page-table permissions control whether a process can read, write, or execute a particular memory region. This prevents one process from casually accessing or corrupting another process's private memory. And here's another interesting bit: virtual memory doesn't mean every process gets its own dedicated chunk of physical RAM. Processes can share physical pages containing read-only executable instructions, use memory-mapped files, or employ mechanisms such as copy-on-write to avoid unnecessary duplication. So, bringing it all together: ->Virtual address spaces give each process its own view of memory. ->Page tables define how virtual pages map to physical frames. ->The MMU performs address translation and enforces access permissions. ->The TLB caches translations to speed things up. Page faults allow the OS to handle memory accesses that need additional intervention. The beauty of virtual memory is that a process doesn't need to know where its data actually resides in physical RAM. It simply works with its own virtual address space. And that's abstraction in computer architecture for you. A relatively simple interface hiding an extraordinary amount of complexity underneath.

Introducing Jev @typesafeai powered social-api Try it at treg.to/social - One API to access any social data - Track every mention of you & your competitors - Find trending posts in your niche, filtered by Jev - Under $0.05 to pull 1,000 competitor ads Open source, 0% markup 👇 Repo link below

Introducing Soniox STT v6, the world’s first speech recognition model that knows what your voice agent is saying and automatically filters out the agent’s own voice. Two connected innovations make this possible: → Agent text input gives recognition context from both sides of the conversation, improving its understanding of user responses. → Agent voice echo cancellation filters out the agent’s playback while continuing to listen to the user, even when they interrupt. The result: fewer false interruptions, fewer clarification loops, and more reliable conversations. V6 also delivers significantly higher accuracy across 60+ languages for voice typing, meeting transcription, multilingual recognition, and real-time translation, with stronger speaker diarization, language identification, endpoint detection, and context customization. One unified speech model. All 60+ languages. The full feature set, with languages freely mixed. Available now for real-time streaming and asynchronous transcription. Explore what’s new in Soniox STT v6: soniox.com/blog/soniox-stt-v…