- cross-posted to:
- programmer_humor@programming.dev
- cross-posted to:
- programmer_humor@programming.dev
You must log in or # to comment.
Outdated. They stopped labeling the axes or for GPUs they label it some bullshit like “relative performance” so they can claim 16x performance with upscaling and slop-generated frames.
A couple of years ago I saw a thing about how Apple had announced the M3 or M4 chips were something like “16x faster!”
And yeah. They were.
Than the last-gen Intel chips they used. Three or four years before.
It’s bullshit all the way down.
And those chips were badly throttled. There were videos showing you could get 20% more performance from the Intel MacBook by putting it in a puddle for better cooling.
The devil’s advocate version is that nobody on Apple Silicon needs to upgrade yet so they compare to CPUs people might actually be considering upgrading from.
I still hate how deceptive it is… And it’s completely unnecessary because they’ve been achieving stellar gains almost every generation.
That is such a pet peeve of mine. Charts without labeled axes might as well just be art projects.
never trust a chart you haven’t manipulated yourself
Don’t trust the ones you’ve manipulated yourself either: that’s just your own bias…
I remember when people got upset by things like this. I wonder why it’s “ok” now.
People started replying with “nerd-face-emojo acthually” to make fun of anyone that points this shit out.
The people who pointed it out got tired, and most people will come to accept anything that doesn’t directly harm them if they’re not repeatedly informed why it’s bad.
I remember all the anti-establishment, anti-sellout, anti-consumerist comics from the 1980’s.
Pancreatic cancer did Bill Hicks a huge fucking favor.
The rest of us are poorer for it. Here’s Tom with the weather.
It can in some cases be beneficial to have a windowed graph like that. Most of the time I see one it is just misleading though.
I’d argue it’s only useful when you have more than two data points. With only two, it’s almost always going to be misleading in some way at a glance.
I get this. I think the average PC builder only goes into the grit when they think it’s upgrade time. So you start your research to catch up on the last few years you have been out of the loop on. You just know shit’s better now because it’s everywhere. The FOMO…
It starts with checking parts, pricing, etc. Then start diving into the research as you narrow down to three respectable options. Then you start checking the “versus” stuff for bang for buck. Then you notice your current shit way down in the graphs. But then you notice it’s really not that bad on the numbers when you realise where the graph starts.
Four hours later the conclusion is, “I don’t need to spend $800 on something that’s 15% better. Come to think of it, my shit’s really not doing all that bad. Hell, people are still seeking it out as a fantastic budget option. I’ve got at least another 2-3 years. I don’t need this.”
The only reason I got a new laptop recently was I got it for decent price for its specs.
A 20% better system for 15% cheaper than my 6 year old one… It penciled out for me.
I spent the past year watching builds and prices. There is so much overpriced garbage on the market today.
It makes a tiny bit more sense with laptops imo since hinges and batteries and stuff wear down and it just won’t be nice to use after enough years.
For desktop PC parts it better be a 50% uplift in some benchmark that’s relevant to me without also having to jump up in price to reach it
There has to be a killer app that I really want to play. For the current iteration, it was Cyberpunk. I’m really glad I went top-tier right before all this chip price-fixing nonsense started.
My most recent computer purchases were more about seeing that the world might not stay in a place where it’s as easy to get computer parts as it is now and there’s a decent chance I’d regret not buying what I can now. My previous system isn’t that bad but it’s now my backup in case my good one fails and parts are impossible to get or stupid expensive.
And since I got it before memory and storage prices took off, I’m already not regretting it.
I had the same thought but didn’t pull the trigger. My system’s still absolutely fine, though. The next upgrade round will be a biggy, though. I’m guessing new socket type so new mobo. And since I keep my CPU and GPU harmonious to prevent bottlenecking, that means new video too. Will likely need a beefier PSU, and at that point may have to look at all-in on DDR5 as fine as DDR4 is for now. But I got a few years…
I made a chart like this once, it was to encourage my wife durong the start of her weight loss journey
How did 100Mhz increase help her with this?
Moore’s Law is struggling against the limits of what’s physically possible and the days of massive jumps in performance each generation are long over. It’s hard to advertise a new CPU when it’s running at nearly the same clock speed models ran at ten years ago and all you’ve done is thrown more L3 cache at an otherwise 99% unchanged previous model.
Even the number of cores has stagnated, and that was their Hail Mary to keep the performance gains coming. And I bet the Spectre/Meltdown mitigations eroded a large part of those gains too - IIRC existing chips lost ~12% performance after the firmware patches, and new ones had to be redesigned to handle speculative execution in a less efficient (but more secure) way.
Though there has recently been another step forward with transistors being able to get under a nanometer, 0.7 nm from 2 nm to be precise, and the last time a measurement needed to go another level smaller was in 1984 when it went from 1 micrometer to 0.8 micrometers.
That 0.7 nm one marks another jump in power while being significantly more efficient, which is I guess Moore’s law living out of spite of not actually being a scientific law.
Also it’s funny this is about the 2nd or third time I’ve seen someone say “Moore’s law is dead” since that 0.7 nm process was figured out not that long ago.
Interesting! I’d always heard that the nanometer wars were mostly marketing fluff these past several years and the quoted numbers divorced from reality, but they’re claiming genuinely sub-nanometer chips. And 3d chip designs, which is equally impressive. I haven’t followed chip tech in a few years, but last I knew heat dissipation was still a huge bottleneck preventing stacked designs.
Though it’s a shame they’re focusing all that innovation towards the AI sector. Their consumer lines have been stagnant for a while now.
Intel’s nanometer-ish numbers are BS. They call their 14nm process “Intel 10”, their 10nm process “Intel 7”, their 7nm “Intel 4”, etc. As far as I know, other companies are more truthful.
