phire

36 ms·

571.
by phire 2y ago
It's not great reporting. The article seems to be fixated on how he lied to his finance sources, and acts like that was the only thing he did wrong. If you read the SEC lawsuit [1], the problem is less that he lied to his finance sourc
572.
by phire 2y ago
For the 8086 they had a gate level simulator called LOCIS. The problem is that intel historically designed everything at the transistor level, so the 8086 didn't map onto a gate level simulation very well. The "Coping with the Com
573.
by phire 3y ago
Though the branch predictor can chew though both layers of indirection. It can actually start fetching code from the function (and even executing it) before it even reads the function pointer from the vtable. Though, that assumes a correct
574.
by phire 3y ago
This was true 15 years ago, but not so much today. The branch predictors actually hash the history of the last few branches taken into the branch prediction query. So the exact same branch within a child function will map different branch p
575.
by phire 3y ago
Jumps/calls are actually be pretty cheap with modern branch predictors. Even indirect calls through vtables, which is the opposite of most programmers intuition. And if the devirtualisation leads to inlining, that results in code bloat
576.
by phire 3y ago
Yeah, nearly 30 years of continual incremental improvements to the same P6 microarchitecture. Intel actually went back to the P6 µarch after Netburst (Pentium 4) failed to meet expectations. Fortunately, they had continue to improve P6 as a
577.
by phire 3y ago
Exchanges almost always have large amounts of stable coin, and no limit on withdrawing them. It makes it so easy to get large amounts of cash equivalents out of an exchange and then launder them though mixers (even laundering it though the
578.
by phire 3y ago
Yeah, the amount of energy per coin mined kind of follows the price of the coin (in the longer term, not so much in the short term). So the price going down is great news for anyone concerned about energy conservation, but because the amoun
579.
by phire 3y ago
> I don't understand how it'd work out: too little BTCs rewarded per block The math only works out if value of bitcoin continues to increase in value, at least a constant 25% annual return, year after year (on top of inflatio
580.
by phire 3y ago
That's potentially true if someone is trying to profit off 51% attacks. But the true worst-case scenario is one where a coalition of governments decided to use 51% attacks as part of a plan to kill bitcoin. They don't need to inve
581.
by phire 3y ago
Technically True. But there was an optional registry key that prevents the user from doing anything until after you were logged into the network. So practically, it was a login dialog. But still no security.
582.
Why is PS3 emulation so fast: RPCS3 optimizations explained [video](youtube.com)
6 points by phire 3y ago|1 comments
583.
Backporting .NET Framework 2.0 – 3.5 to Windows 95 [video](youtube.com)
3 points by phire 3y ago|0 comments
584.
by phire 3y ago
RISC-V has slightly more complicated instruction formats, but is otherwise feels a lot like a cleaned up and simplified MIPS. The fact that you can easily see MIPS's 16 bit and (most of) 26 bit immediates in a hex editor is one low-key
585.
by phire 3y ago
The reporter said "it runs off three five inch floppy disks" It's unlikely a system deployed in 1998 would be using three separate 5¼" floppies. I'd believe a single 5¼" floppy, but not three of them. It'
586.
by phire 3y ago
I believe what happened is that Valve said something along the lines of: "We are a little concerned that Nintendo might object, you can only continue if you get an official notice from Nintendo saying they don't object." It&#
587.
by phire 3y ago
Not only do you only have 4KB of instruction cache, but it's direct mapped. So if the linker just so happens to place the function you are calling at some address that's some multiple of 4KB away from your current hot loop, then t
588.
by phire 3y ago
Which is part of the problem. LLMs don't like doing things step by step. You can force an LLM to go step by step with an external tool that continually re-prompts with each step, but the challenge requires a single prompt. When you try
589.
by phire 3y ago
I found this other paper that tests Temperature: https://arxiv.org/abs/2402.05201 It appears that temperature has no impact on problem solving performance. So this paper isn't getting improved performance because
590.
by phire 3y ago
Yes, but picking the most similar output from a bunch of queries with a higher temperature is not the same thing as the output from a single low temperature query.
591.
by phire 3y ago
I'm not sure people in these comments are reading this paper correctly. This seems to essentially disprove the whole idea of multi-agent setups like Chain-of-thought and LLM-Debate. Because this paper introduces their alternative metho
592.
by phire 3y ago
There is no technical limitation, there is even an API for doing so. It's just that apple block any submissions using that API from the app store.
593.
by phire 3y ago
There is a lot of space between interpreters and a full JIT that is under-explored. Because developers typically skip straight to a JIT when they find a interpreter isn't fast enough. Dolphin does have the cached interpreter, which eli
594.
by phire 3y ago
How to you know they don't? When a state actor says "We found this exploit", people will get paranoid and wondering if the fix is actually an exploit. Not saying it happened in this case, but it's really easy for a state
595.
by phire 3y ago
I guess since forwards rendering doesn't need a proper texture cache, you could used the savings to implement a write combining buffer. But I'm only saying that because I know write combining is a thing, and I'm not sure a ha
596.
by phire 3y ago
One of the reasons why the RISC vs CISC debate keeps popping up every few years, is that we kind of stopped naming CPU uarches after the CISC and RISC terminology was introduced in the late 70s. And because there wasn't any new names,
597.
by phire 3y ago
> It’s not random. Not even remotely. Sure, it's technically deterministic. But it's also very chaotic, as it depends on the exact position of the vertices. It's not something a game could predict and take advantage of.
598.
by phire 3y ago
Yeah. It essentially ends up as the Playstation GPU, but with quads and forwards texturing. It's basically inevitable. Any proposal for fixing the Saturn that slaps on a "optimise for BoM" requirement will be pushed towards u
599.
by phire 3y ago
Yeah. You have a point. My proposed changes only really bring the Saturn's 3D graphics into roughly the same technical capabilities as the Playstation. They don't do anything to help with the performance issues or the higher manuf
600.
by phire 3y ago
The only "correct" that matters here is if it produces a result that's useful for 3D graphics. The "some pixels get randomly blended multiple times" behaviour is simply not useful. And if it's not useful, then

More ›