You don't need it, but as someone who has been there: For me making a 3D engine is a lot of fun! But then I never finish the actual game. So if you actually want to ship a game, I recommend using an engine. Personally I prefer Unreal.
For 2D, yeah, making the engine yourself is fast and easy. Can go without a big engine.
122b would be awesome. It is the largest size you can kinda run with a beefy consumer PC. I wondered about gemma stopping in the 30b category, it is already very strong. 122b might have been too close to being really useful.
Not OP, but I ran 122b successfully with normal RAM offloading. You dont need all that much VRAM, which is super expensive. I used 96gb ram + 16gb vram gpu. But it's not very fast in that setup, maybe 15 token per second. Still, you can give it a task and come back later and its done. (Disclaimer: I build that PC before stuff got expensive)
Any good gaming pc can run the 35b-a3 model. Llama cpp with ram offloading. A high end gaming PC can run it at higher speeds.
For your 122b, you need a lot of memory, which is expensive now. And it will be much slower as you need to use mostly system ram.
Seconding this. You can get A3B/A4B models to run with 10+ tok/sec on a modern 6/8GB GPU with 32k context if you optimize things well. The cheapest way to run this model at larger contexts is probably a 12gb RTX 3060.
Here in Japan they started to forward me to the app page when ordering. So you are forced to use the app with a mobile browser. Even though the website could do it perfectly fine in the past.
I do not go often, but if I do I prefer to sit, order in the page and they bring it to your seat. I dont like the Kiosk.
I was thinking to get 2x r9700 for a home workstation (mostly inference). It is much cheaper than a similar nvidia build. But still not sure if good value or more trouble.
I own a single R9700 for the same reason you mentioned, looking into getting a second one. Was a lot of fiddling to get working on arch but RDNA4 and ROCm have come a long way. Every once in a while arch package updates break things but that’s not exclusive to ROCm.
LLM’s run great on it, it’s happily running gemma4 31b at the moment and I’m quite impressed. For the amount of VRAM you get it’s hard to beat, apart from the Intel cards maybe. But the driver support doesn’t seem to be that great there either.
Had some trouble with running comfyui, but it’s not my main use case, so I did not spent a lot of time figuring that out yet
I have a dual R9700 machine, with both cards on PCIe gen4 x8 slots. The 256bit GDDR6 memory bandwidth is the main limiting factor and makes dense models above 9b fairly slow.
The model that is currently loaded full time for all workloads on this machine is Unsloth's Q3_K_M quant of Qwen 3.5 122b, which has 10b active parameters. With almost no context usage it will generate 59 tok/sec. At 10,000 input tokens it will prefill at about 1500 tok/sec and generate at 51 tok/sec. At 110,000 input tokens it will prefill at about 950 tok/sec and generate at 30 tok/sec.
Smaller MoE models with 3b active will push 70 tok/sec at 10,000 context. Dense models like Qwen 3.5 27b and Devstral Small 2 at 24b will only generate at around 13 - 15 tok/sec with 10,000 context.
This is all on llama.cpp with the Vulkan backend. I didn't get to far in testing / using anything that requires ROCm because there is an outstanding ROCm bug where the GPU clock stays at 100% (and drawing like 60 watts) even when the model is not processing anything. The issue is now closed but multiple commenters indicate it is still a problem. Using the Vulkan backend my per-card idle draw is between 1 and 2 watts with the display outputs shut down and no kernel frame buffer.
Talking to friends who have fought more homelab battles than I ever will, my sense is that (1) AMD has done a better job with RDNA4 than the past generations, and (2) it seems very workload-dependent whether AMD consumer gear is "good value", "more trouble", or both at the same time.
Edit: I misread the "2x r9700" as "2 rx9700" which differs from the topic of this comment (about RNDA4 consumer SKUs). I'll keep my comment up, but anyone looking to get Radeon PRO cards can (should?) disregard.
I have 2 of them. I would advise against if you want to run things like vllm. I have had the cards for months and I still have not been able to create a uv env with trl and vllm. For vllm, it’s works fine in docker for some models. With one gpu, gpt-oss 20b decoding at a cumulative 600-800tps with 32 concurrent requests depending on context length but I was getting trash performance out of qwen3.5 and Gemma4
If I were to do it again, I’d probably just get a dgx spark. I don’t think it’s been worth the hassle.
The Keyboard Speciality shop you linked is great. It is always surprisingly crowded for such a niche topic. The have good test boards with different key switches.
Note that in the area are quite a few other good shops. Tsukumo has various hardware across many floors and also keyboards, in b1 they also have a razer store. Then there is a shop across the street Galleria or something, more targeting esports, if you are into that.
I wonder if the winning game becomes your own boss and tiny companies.
I want to do the jump, but lack of courage, good ideas, sales skills and a very good salary still holding me back (open for suggestions).
But if the very good salary would go away, the scales tip instantly.
Probably not the answer you want to here but I'll share my perspective. Three years ago my wife and I sat down and optimized our finances so I could soft-retire and focus on a few of my life goals while simultaneously working on ways to generate income without the stress of being in the employ of others. It was tough work which mainly involved paying down a lot of debt so we can live more lean. We did a lot of optimization and of course some compromise and lifestyle changes. Fortunately, my wife earns enough for us to still live comfortably on a single income.
Now I am her part-time personal assistant which has taken a big load off her plate and reduced her stress significantly. A lot of this work is clerical: writing emails, grants, curriculum/lessons (she's a teacher), ordering supplies, working with spreadsheets, doing misc. graphic design and other office work. I also take care of the household, finances (mostly) and pets. In my spare time I pursue my lifelong passions (writing, game design, and programming), but with each of these my focus has been channeling those passions into generating income. This is not a requirement of my soft-retirement, but rather a choice I made to create balance between us.
Overall, we are much happier and fulfilled and have managed to carve out a life where we work meaner and leaner without huge sacrifices. In reality, it feels like we are financially better off than we were before.
Whenever I read something like this I have to ask if kids are in the picture? Or maybe they've already moved away.
I'd like to do something like this but everything that has to do with kids is both too expensive and too unpredictable for lean living to be an achievable goal.
- how old are they? If the poster is ~60, likely has savings and may even have Social Security income. If they worked as (say) a police officer for 20 years, they may have pension income. A 47-year-old former military officer could reasonably have kids at home and also pension income from the military.
- Many people inherit houses (most houses are eventually inherited). Most sell them, but it can be a viable choice to just move into an inherited house to zero out housing expense. OR one could inherit a house that is >> valuable than one's own, such that selling the inherited house allows one to pay off one's own house.
- Location. The Discourse typically divides between HCOL and LCOL, but ignores that in both there are also people who spend much less than the average. In NYC the average home price is ~$850k, but there are today listings for 3BR homes in the low $200s (<$1,500/mo).
And of course these are stackable. One could have a military pension and buy a cheaper place and have a buffer from an inheritance. (None of this is uncommon.)
they said "teacher" but also mention writing grants. A high school teacher isn't writing grants, their wife could be bringing in a lot more than the typical teacher.
A tenured position in a reasonably good university can give you quite a good standard of living, and depending on your area, there are even opportunities for occasional consulting work.
Not to mention that the professional prestige itself in an academic profession gives your family a lot of status that other people usually try to attain by buying expensive stuff.
Even in the fanciest neighborhoods, nobody cares if a Princeton Professor drives a 20 years old Volvo.
I think the heart of what they're getting at is that while on paper they are bringing in less income, they have gotten off the hedonistic treadmill, and as a result, quality of life per dollar has increased dramatically. They are less stressed about finances than they were prior, even though their income is lower.
We live in a world where someone has to clean the sewers, unblock toilets, maintain electricity lines in snow storms, weld deep underwater, clean, wipe the butts of old people, and 10,000 other thankless, tiring, and dangerous jobs which no one in their right mind would ever do because they found it fun and interesting. Until we have very highly capable robots to do these jobs, we need some way to incentivise doing work which few others want to do, or are capable of doing. Right now we use money as the incentive. On top of that, there are things people do which bring a lot of value to others. They invent new things, for example, and sell them. Others buy them. We also want to incentivise that, even though it's not easy, and not everyone is capable of doing that.
I do think AI and robotics will usher in a much more abundant world in the future. It's unclear how we navigate that - economically, politically, socially.
Alternatively, you live in a society that has conditioned you to devalue manual labour and erronously assume that no one exists who actually enjoys physical interaction with the world.
As you're likely to be in the US, you could always watch the Mike Rowe Dirty Jobs back catalog.
> Alternatively, you live in a society that has conditioned you to devalue manual labour and erronously assume that no one exists who actually enjoys physical interaction with the world.
I think the burden of proof is on you to prove that the same number of people would be interested in wiping butts and fixing mains lines in snow storms if they weren't paid for it. To me this seems unbelievable and naive on its face, but I'm willing to keep an open mind if you have some evidence.
I generally think people who believe these jobs would be done without some kind of incentive have never worked them before. They suck. Sometimes jobs really do suck, but they still need doing.
It can be enjoyable in the context of failure analysis: troubleshooting, finding root causes, documenting other people's fuckups then tracing through the assignment logs on who interacted with the server last.
Leaving aside the scene from Life of Brian, I have no issue cleaning shit - I've raised children, they poop, I have livestock, they shit, kids will happily frisbee cow pats, raking out sheep shit from under shearing sheds is a job that I've done, as have many .. you end up with a couple of tonne stacked high on a double axle trailer that's great for the garden.
For what it's worth, I don't mind a bit of higher dimensional data reduction when processing raw multi channel data, or geophysical world modelling (magnetic fields, gravity, radiometrics, etc).
I'm heading to the Graeberian world of bullshit jobs which ironically tends to head towards the direction of meaning.
I'm pro "everyone cleans their own shit" but the meaning of a garbage truck driver could immense compared to a honest hedge fund manager or a VC Patagonia vest.
Cleaning time of our own shit hopefully won't be a full time job. We'll just figure out the ones creating too much shit and educate them as a society :D
>> We live in a world where someone has to clean the sewers, unblock toilets, maintain electricity lines in snow storms, weld deep underwater, clean, wipe the butts of old people, and 10,000 other thankless, tiring, and dangerous jobs which no one in their right mind would ever do because they found it fun and interesting.
>> I do think AI and robotics will usher in a much more abundant world in the future. It's unclear how we navigate that - economically, politically, socially.
Delusional optimism. If AI and robotics take over, the only effect will be another wave of layoffs and unemployed, not even the willingness to unblock toilets or wipe butts will save you from homelessness and destitution. We're already on the way to Victorian era poverty, if robots take the shit jobs too, we're back to Oliver Twist: please sir, can I have some more ... tokens?
Given how we handled the industrial revolution and more recently, the destruction of Midwest industry in Chinese offshoring, you may very well be correct. People will cheer cheaper products and services while watching unemployment rise around them.
However if it happens so fast, and so many of us are impacted, I have to believe that will impact how we vote.
How many acres are you personally willing to farm to let others eat without payment “in a just world”?
How many days per month are you willing to pick up trash, sit in a fire station, or teach elementary school?
It’s not slavery (if you) that other people won’t give you their output without payment. In fact, it’s closer to being slavery in the other direction if they have to work and you get the benefits of their output without payment…
> In fact, it’s closer to being slavery in the other direction if they have to work and you get the benefits of their output without payment…
This sounds a lot like you've been conditioned to think there can't be an alternative to the current system. Even if I don't know what a better system would be, I can absolutely imagine that there are better options than what we've got. We should all want that and push for that and ask ourselves what it might be until we find it.
I can tell you this much about what I think would be part of that better system: we wouldn't leave people to sleep on the streets and we wouldn't have for-profit healthcare.
> I feel this fucking form of slavery as well hard
I think you'd do well to learn more about how slaves were treated before making these comparisons. Have you been whipped until your flesh opened and had salt, lime juice, and peppers rubbed in the wounds because you messed up at work, where you are also forced to lived?
That was always the "winning game". Only problem is that's a lot of work. The more things change, the more they stay the same; if you want more money, work harder. People who don't want to work harder complain that other people make more money because they either don't understand or are in denial about the amount of work the people they envy put in.
Yes there are exceptions. No pointing out exceptions won't help you, though it might make you temporarily feel better about yourself.
The elephant in the room is health insurance. We have a system where even if you have a fairly good income, buying insurance as an individual (or as a small company that isn't buying in a volume high enough for insurance companies to want to give you a discount) means that you'll in all likelihood be paying a lot more for a lot less coverage. The ACA attempted to solve this by having insurance companies offer plans on "public exchanges" by state and then subsidizing the costs, but because most people making good money get insurance through a job with benefits rather than buying it directly, in practice there aren't really any options on the public exchanges calibrated for people with high incomes. (Plus, if you live in a red state, they've likely refused to take the subsidies, which means either the prices are higher or the plans are even more meager based on what the insurance companies expect people without benefits through employment to be able to afford, or both).
> Plus, if you live in a red state, they've likely refused to take the subsidies
No. Individual states can refuse Medicaid expansion, but that does not have any bearing on the health insurance marketplace / premium tax credit ("subsidies"), which states cannot opt out of.
That's no longer true. You can have ICRA etc plans, for the tax benefits.
The truth of the matter is that employers pay a humongous share of the health insurance bill, and if you shop directly, you will pay that 100% on your own.
You do have to put in a little more effort, but as an employer you can build a hybrid plan and contract with certain networks, and lower your bill tremendously.
They're still maybe right. In CA, it's pretty common for the best plan you can buy as an individual to be half as good as whatever your employer offers and to cost twice as much as the combined employee+employer contribution.
How does ICHRA fix that? What's this "contracting with certain networks" you're referring to?
Write down 10 of your TODOs that generate income. As per natural law, it's likely that 1-2 of those TODOs will have 80% of the impact while the other 8-9 will have almost no impact. Now here comes the interesting part. You propably already know the 8-9 tasks that have almost no impact, but as per another law, those are also the easiest tasks (checking mails and such). On that list the TODO that feels like the biggest hassle and you least want to do will likely have the biggest impact. Sit down and just do it. Now, without delay. That already makes you more productive than 98% of your colleagues.
> People who don't want to work harder complain that other people make more money because they either don't understand or are in denial about the amount of work the people they envy put in.
I assure you, I have never in my life worked 20 times harder than someone making minimum wage.
No executive has ever worked as hard as the girl pushing carts at your local supermarket or the "Illegal" hand picking the fruit you eat for 12 hours a day for less than minimum wage or the teen mechanic dealing with a "2 hour" warranty job on a vehicle rusted to shit.
There is no such thing as knowledge work that takes that much out of you. Sure, thinking hard and making choices all day will exhaust you, but you won't stop moving at age 55 because your body was literally used up for pennies to make someone else wealthy.
If you fly business class, you are the elite making your wealth by skimming from people doing the real Labor. Your wealth is enabled by a paper and some writing. You contribute nothing.
A burger flipper cannot flip 20x the burgers. There isn't really any way to produce more output flipping burgers. Even if you could, if there isn't a queue of people waiting to collect their orders, there isn't any point in producing more blindly.
The person responsible for designing the process that thousands of franchises use probably does make a lot of money.
I bet if you could make it interesting, YouTube/TikTok/Instagram/Whatever could make it possible to get paid to dig holes in your backyard.
You could argue that the value is in the entertaining filming/acting/story telling etc, but if the videos are about digging holes then I think it's valid to say someone is paying you to dig holes.
Yeah i hate what you are suggesting, because soon there are uninteresting people chasing every subject trying to convert it into a career. Just leave some things alone ok and quit strangling my hobby with both hands
I get this, but also genuinely interested to know how to measure outputs. For me it's almost impossible to get it objectively right.
Maybe this doesn't apply to your case, but how would you measure outputs of say product development, or any data related project. Lot's of things don't have a good measure of output before the thing is done. Maybe your product / analysis improves profitability by 10x or maybe it was a flop and lost money.
Tangential, but I'm also seeing the quality of measures going down, with AI it seems that the number of [emails|code|analysis] produced is again a good measure.
> I get this, but also genuinely interested to know how to measure outputs.
Measuring outputs or inputs (hard work) is always hard. Did someone get the thing that was asked done both quickly and correctly? Do they do this consistently?
I also find inputs harder to measure because someone could be in the office 12 hours/day, but on Facebook the whole time. They could also just spin their wheels doing 'fake' work.
I spend some time going through what programmers wrote over the past years and many of them were rewarded for getting things done quickly with no complaints.. The more diligent ones probably didn't last since they got things done correctly which takes a lot more time and thought.
It's why I said quickly and correctly. I think it's a cop out to say someone was slow because they were building it correctly. Famously, the old space shuttle software was developed very slowly because it had to be 100% correct at all times. Most software does not need that level of correctness. Part of a SE's job is to understand that.
I pay a lot of attention when someone claims to have solved a problem I suspect to be NP-hard. There are a lot of possible explanations, for example they may have an incorrect measurement function or they may have chosen a simpler related problem that isn't really NP-hard, or both.
It is. I am a fractional CTO running my own consulting business and I make 3-5x as much as I ever did working for one company. And all my clients are very happy.
I've considered it myself; I don't want to make a business doing contract work again, because I did not enjoy that.
If I were to start my own business it would have to be a product. I have plenty of interesting projects that I work on in my free time, but I'm not sure any of them are monetizable, or at least not monetizable enough for a venture capitalist to throw money at me (especially since most of them do not involve AI). I could probably think of something that could be monetizable if I really tried but if I don't actually enjoy the work I'm doing on the side for fun then I'm probably not going to do a particularly good job on it.
Though even if I did have some brilliant project that I could sell, I have no idea how to go about finding VC investors. And even if I knew how to find these investors, I think I would ultimately be too afraid to actually commit to it.
Increasingly it's seeming that I will probably not be worth billions of dollars in my lifetime, for no other reasons than I'm too much of a coward and I'm too discriminating with what I actually work on. Sometimes it depresses me to think about it, but hard to feel too sad for myself when I still have a high salary job that involves me staring at a computer screen all day.
> Increasingly it's seeming that I will probably not be worth billions of dollars in my lifetime, for no other reasons than I'm too much of a coward and I'm too discriminating with what I actually work on
I can't tell if you're wildly underestimating how many circumstances are outside of your control or just have an extremely high opinion of how much of an outlier you are (or maybe this isn't meant seriously and just went over my head), but I think that there are vanishingly few people (if any) in the world whose only impediments to a high likelihood of becoming a billionaire are self-imposed. I don't think that even extremely smart and charismatic people are particularly likely to do that. For every one that reaches that level of wealth, there are far more who try and fail, and it's not always because they weren't willing to work on shady things or weren't smart enough; some factors are just beyond the ability of an individual human to overcome, and you might just be lucky or unlucky.
> Clearly poor communication on my end; I was joking about the billion dollars. My bad.
No worries, I'm pretty bad at being able to recognize humor over text even when it's done well, so I legitimately can't tell the difference between whether it's on my end or yours! And if it is on your end, it's certainly still a more palatable character trait than the amount of ego needed to say it seriously.
It is genuinely hard for me to say anything completely seriously and without an abundance of sarcasm. My therapist says it might be an avoidance mechanism that I devised as a kid in order to avoid confronting serious topics.
There probably is some truth to that, and it can understandably come off as me not listening or being mindful to things people say to me because they think I'm blowing it off, even when it's more of just how I deal with things. My wife, being pretty awesome in many ways, realized this pretty early on and thus more or less always understood that when I make a smartass comment even for serious topics, I'm not really trying to make light of it as much as its just how I cope with things.
Anyway, yeah, I'm sarcastic a lot of the time and I realize that that doesn't always come through with text.
I can definitely relate to that to a certain extent. My (mistaken) inference that you were serious probably was because so much of the comment before that did seem genuinely thoughtful to me, so I assumed that maybe all of it was intended to be. If anything, it sounds like maybe holding off on the sarcasm for so long forced you to have to make a quip near the end (or I just missed more of the sarcasm earlier).
Forgive me for saying this, but I think you may be drinking too much of the kool-aid.
If some of your projects are monetizable, couldn't you move forward without VC help?
Perhaps related, why do you need to be worth billions of dollars? I feel your visions for what you want your future self to be are highly unrealistic and you're probably setting yourself up for a lot of disappointment and unhappiness.
Sorry for the bluntness, but I think one could be happy on a lot less.
I was being a bit joking and hyperbolic about the billion dollars, though obviously that wasn’t communicated clearly. I don’t really need a billion dollars.
I don’t think anything I have right now is very monetizable; most of the fun stuff I work on now ends up being formal methods stuff, which is cool but hard to make any money with.
I guess what I was saying is that I think I am ultimately think I am too cowardly to just go for it and make my own company. I don’t think I am capable of purposely avoiding income for N months for a project to pick up.
Given that two separate people didn't understand the joke, clearly bad communication on my end. I do not accept your "my bad!", and instead insist on it being my bad.
> I wonder if the winning game becomes your own boss and tiny companies.
Building a successful set of tiny companies is very hard. Unless you get lucky with the exact right idea, execution, and market timing it’s really hard to build a single business that pays as well as our tech jobs do. Building multiple companies is even harder.
I think everyone sees the survivorship bias examples like the levels.io guy or a few of the app developers who got rich and thinks it must be easy because their businesses were simple. The indie hacker communities are filled with people trying to follow in their footsteps and not getting anywhere despite years of hard work. The levels.io success story is not something that is easily replicated because his signups depend so heavily on his huge Twitter presence, where he pushes his sites under the guise of friendly information sharing. People without Twitter audiences try all the time to replicate his success and then wonder why they’re not getting signups like he does.
Most countries used to be hereditary dictatorships ("kingdoms") just a few hundred years ago, then people picked up rifles (and guillotines) and changed that. Now, since we already have some semblance of democracy at the top layer of power, maybe we could revolutionize the lower layers without the risks inherent to picking up rifles.
I don't see why we should be controlled by sociopaths[0] 8 hours a day, quite often against our interests.
---
The other part of the solution is automatically and continuously redistributing ownership of the company according to hours worked and skill level. This of course has to be required by law, otherwise those sociopaths at the top have so much leverage that if you ask for anything other than money during negotiation, you'll get laughed out of the room.
reply