Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

SoftRAM 95 (1996)

drdobbs.com

Post reply on HN
Print view View original

41–50 of 64 posts

Re: SoftRAM 95 (1996)

#41

Post by ygra »

mrob wrote: post #9

>While this seems an easy relationship to satisfy, more-complex models based on real-world performance parameters and access patterns indicate that it is, in reality, very difficult to obtain a performance boost with this method. Now that improvements in CPU speed have outpaced improvements in storage speed, it's easier to benefit from compressed swap in ram. Linux includes "zswap"[0], which in my experience works we…

Windows also compresses pages to alleviate memory pressure before resorting to swap nowadays. Probably the same reason as with Android, as on lower-end hardware, especially with hard drives, compressing/decompressing isn't really noticeable, while hitting the disk is.

Re: SoftRAM 95 (1996)

#43

Post by m0zg »

forgotmypw wrote: post #18
m0zg wrote: post #11

I wish someone would do a study like this on e.g. cosmetics which claim to do miraculous things, but in reality rarely do anything at all. In fact I strongly suspect it all comes from the same giant vat of petroleum byproducts and the only difference is the fragrance, color, and emulsifiers.

If you just want to know the truth, not "prove" it "globally" in a "study", which essentially amounts to proving to the wolf that he is dangerous to the sheep, all you need to do is look at the ingredients list. For extra credit, look the ingredients up and see how many of them have above-zero blue hazmat ratings.

That's not what I want though. I want an ex-industry insider with an axe to grind to tear them a new one, without mincing words, and with abundant evidence presented, similarly to how it's done in this article. It just blows my mind when some women who wouldn't eat non-organic food if their life depended on it turn around and rub petroleum byproducts into their faces every evening. Food colorings are "poison" but lipstick and hair dye are A-ok. :-)

Re: SoftRAM 95 (1996)

#44

Post by pmjordan »

ygra wrote: post #41
mrob wrote: post #9

>While this seems an easy relationship to satisfy, more-complex models based on real-world performance parameters and access patterns indicate that it is, in reality, very difficult to obtain a performance boost with this method. Now that improvements in CPU speed have outpaced improvements in storage speed, it's easier to benefit from compressed swap in ram. Linux includes "zswap"[0], which in my experience works we…

Windows also compresses pages to alleviate memory pressure before resorting to swap nowadays. Probably the same reason as with Android, as on lower-end hardware, especially with hard drives, compressing/decompressing isn't really noticeable, while hitting the disk is.

macOS and iOS do this too. iOS doesn't even use any traditional swap and gets away with it by putting the onus on developers to make sure their apps never go OOM when in the foreground and to deal seamlessly with apps being OOM killed when in the background.

Re: SoftRAM 95 (1996)

#45

Post by cookiecaper »

mrob wrote: post #9

>While this seems an easy relationship to satisfy, more-complex models based on real-world performance parameters and access patterns indicate that it is, in reality, very difficult to obtain a performance boost with this method. Now that improvements in CPU speed have outpaced improvements in storage speed, it's easier to benefit from compressed swap in ram. Linux includes "zswap"[0], which in my experience works we…

zswap and zram are actually different systems. zram is a compressed ramdisk that acts like a block device and can host any filesystem (including swapfs).

zswap (nee compcache), on the other hand, is a region of memory that tries to store pages compressed instead of swapping them out -- basically a last resort before swapping. zswapped pages are not considered swap, so you have to check sysfs to see how much it's getting used. Many default configurations are pretty aggressive, allowing zswap to use up to 20% of physical memory before finally resorting to the system's configured swap space.

Re: SoftRAM 95 (1996)

#47

Post by unionpivo »

newnewpdro wrote: post #23
zamadatix wrote: post #15

Earlier quoted context omitted.

Storage speed has really outpaced CPU speed for the decade zswap has been out. It still has its place but if you have modern flash storage it may well be a throughput drop and power usage increase to enable zswap (particularly with the default algorithms). Latency for pulling single pages has never really been limited by CPU performance, that's always been faster and will probably never change due to physics.

It's not like you really want to use SSD for frequently used swap though, it's just going cause unnecessary wear.

Stick of cheapest 1tb nvme is around 115 EUR (qlc) where I live, so even if you have to change it every 2 years it still worth it.

And i measured disk usage on my desktops (home, job) and laptop, and my main dev "server" (xen, with bunch of vms).

Surprisingly my desktop had most writes to disk, but even then it would take me 6 years on the cheapest (qlc, kingston 1tb) before wearing it out

So i don't think that is a big issue, especially with how much you can improve your life, with faster swap.

Re: SoftRAM 95 (1996)

#48

Post by mrtksn »

Teknoman117 wrote: post #37
PhasmaFelis wrote: post #34

Earlier quoted context omitted.

I know several attorneys, and man, I wouldn't wish that on anybody.

That's become my feeling as well. I've always looked at them making a ton of money without needing to start a company and become somewhat envious (none of what I actually enjoy working on seems to be startup material). Then I see two of my friends pursue careers in law after becoming attorneys, and wow, I now think my 9-5 job for a tech giant is a better gig. They're miserable.

Can you elaborate? What's so bad about being an attorney?

Re: SoftRAM 95 (1996)

#49

Post by irth »

seriesf wrote: post #27
mrob wrote: post #9

>While this seems an easy relationship to satisfy, more-complex models based on real-world performance parameters and access patterns indicate that it is, in reality, very difficult to obtain a performance boost with this method. Now that improvements in CPU speed have outpaced improvements in storage speed, it's easier to benefit from compressed swap in ram. Linux includes "zswap"[0], which in my experience works we…

squashfs is another linux feature that boils down to compressed swap.

it's a read only compressed file system so I am not so sure about that

Re: SoftRAM 95 (1996)

#50

Post by jacobush »

angry-sw-dev wrote: post #32
scarface74 wrote: post #17

Earlier quoted context omitted.

RAM Doubler for the Mac wasn’t a scam. Because of the piss poor way that Classic MacOS handled memory where you had to manually tell it how much memory an application could use and applications needed that much contiguous memory, you could easily find yourself with plenty of RAM but it being fragmented and unusable. Also, you needed to have as much disk space as you had RAM + any “virtual memory”. RAM Doubler used th…

It's amazing how fast capacity and pricing changed back then, and it still happens today. I remember my first PC clone in 1992 it was a white box with a 40MB hard disk and 2MB of RAM, within 6 months or so I spent $200 on two more 1MB 70ns SIMMs. Our high school had a small lab running Mac IIci's with 8MB of RAM and those things were probably $8K each with display, it was a big deal to get time on them. We had more p…

Onion?!
Post reply on HN