Hacker Newsnew | past | comments | ask | show | jobs | submit | bhouston's commentslogin

> When there is an effort to “elevate people of type X” you have to suspect that some large number of those “elevations” are fraudulent rewriting of history. I immediately ignore the whole lot.

A significant part of all credit in the world is assigned fraudulently, it is just a fact. It is because with credit comes prestige and wealth, etc. So humans generally seek it out and many lie about it.


I agree. Growth solves all. It give you money to throw at problems, raises to keep people happy, and motivation. When there is no growth, that is when things fall apart.

“Edit: actually the more i go through it, it sounds like chatgpt prose, especially by the end.”

I noticed that too. 100% mostly written by AI.


This seems to miss diagnose the issue as suicide but really the startup wasn’t having success. If the company isn’t making money, then yes you run out of runway, you get founder conflicts etc. Saying this is suicide is missing the point.

If you are failing giving up is sometimes the right option. Many startups are based on assumptions/predictions that turn out to be wrong. And it is hard to pivot if you’ve spent the money and committed or not worth it if your cap table is wrecked.

BTW this mostly read like it was mostly written by AI. The various bolded text, missing the point, and emdashes.


Totally agree that sometimes shutting down is the right call, not every startup should survive, and plenty of ideas simply don’t work once they hit reality. The point I was trying to make isn’t that failure = suicide, but that many teams collapse before they’ve fully tested whether the idea works.

And fair feedback on the writing style. Appreciate you calling it out. I am not a native speaker, so I definitely used AI to help me formulate my ideas & opinions.


If I were a startup, I would simply achieve product/market fit instead of failing.

(yes, /s)


Haha, yes. I never understood why startups don't achieve PMF

This sort of makes sense. If there is no competitive advantage in buying the latest AMD or Intel CPUs, why buy them when you can just deploy a generic (ARM licensed) CPU at cheaper prices.

The competitive advantage right now is in NVIDIA chips and I guess AWS needs all their free cash to buy those instead of non-competitive advantage CPUs.


Is there a list of Geekbench performance metrics for the various Graviton CPUs?

I need a reference point so I can compare it to Intel/AMD and Apple's ARM cpus.

Otherwise it is buzzwords and superlatives. I need numbers so I can understand.


https://instances.vantage.sh/ shows coremark scores for each EC2 instance type.

It always strikes me that the best place of information for a cloud provider is not from that provider but a third party website. This is not a good comment for the cloud provider.

Funny story: When I was at AWS, I found that the easiest way automate instance data collection was by using the Vantage website code (it's on GitHub).

The cobbler's children have no shoes.


Founder of Vantage here and former AWS employee.

We actually recently made the decision to staff someone full time on the site just to maintain it for the community. Even the JSON file for the site gets hit hundreds of thousands of times per day...feels like it's become kind of the de-facto source of truth in the community for where to get reliable AWS pricing information and I believe its powering a pretty remarkable amount of downstream applications with how much usage its getting.

We acquired the site almost 5 years ago and want to continue to improve it for the community. If you have any cloud cost management needs, we're also able to help for our main business here: https://www.vantage.sh/

Awesome to see all the comments on it here!


Thry have to adhere to their marketing words and numbers like "efficiency increase of 99999% in performance per dollar per token per watt per U-235 atom used"

Also, use the ffmpeg fps column to check single threaded score.

While the 5 variant isn't yet available outside of the preview, you can of course spin a 4 up and run geekbench yourself. Plenty of people have and you can find them in the GB DB. And of course most people spin up their specific workload to see how it compares.

Core per core it pales compared to Apple's superlative processors, and falls behind AMD as well.

But...that doesn't matter. You buy cloud resources generally for $/perf, and the Graviton's are far and away ahead on that metric.


Not true at all. Single thread CPU scores for Graviton2 are about half that of Intel, while only being about 20% cheaper at best.

Groan. Yes, absolutely true.

While I know this thread will turn into some noisy whack-a-mole bit of nonsense, an easy comparison is the c8g.2xlarge vs the c8i.2xlarge. The former is Graviton 4 vs Granite Rapids in the latter. Otherwise both 16GB, 15Gbps networking, and both are compute optimized, 8 vCPU machines.

Performance is very similar. Indeed, since you herald the ffmpeg result elsewhere the Graviton machine beats the Intel device by 16%.

And the Graviton is 17% cheaper.

Like, this is a ridiculous canard to even go down. Over half of AWS' new machines are Graviton based, but per your rhetoric they're actually uncompetitive. So I guess no one is using them? Wow, silly Amazon.


The latter is a 4-core machine with 8 HyperThreads. This doesn't actually matter to your price-performance metric but is worth mentioning because it's the reason why the Intel part performs so comparatively poorly. They're fast chips, they're just wildly uneconomical. If you wanted to compare equal core counts (c8i.4xlarge vs. c8g.2xlarge), then the Intel instance type wins on performance but the Graviton is 58% cheaper.

Groan. Absolutely not. :)

c8g passmark score: 1853 c8i passmark score: 3008

I guess the fps column isn't a good representation of single thread score. Also looking at the passmark scores for i4i vs i4g, i4g is about 1k and intel is about 2k, and the more modern Graviton equivalent of i4 is the same price, so...

https://go.runs-on.com/instances/ec2/c8g

https://go.runs-on.com/instances/ec2/c8i

https://go.runs-on.com/instances/ec2/i4g

https://go.runs-on.com/instances/ec2/i4i

Silly amazon.


So confident. And exactly the whack-a-mole nonsense I predicted.

See the comment by electroly. They actually know what they're talking about.

See, the FPS score is for the whole machine. The c8g gives you 8 real cores. The C8i gives you 4 real cores, 4 hyperthreading pseudo-cores. So for those two machines the c8g unequivocally gives you more absolute computing performance, regardless of the passmark single thread (on a single core) on the c8i being better than a single core on the c8g. And the c8g comes at a big discount as well.

That's...the point. The Graviton processors are cheaper per core, and lower performance per core, and you make it up in bulk. You get more performance per $ if you're okay with the ARM stack and your software is good with it, and this is basically universally true comparing Graviton instances versus Intel/AMD alternatives.

You're wrong. Maybe cite some other random nonsense now?


Single thread performance is important for many workloads. It's not nonsense. Things like index builds on an i4g vs i4i could be half as slow. That's really important!

I don't know why you continue to be a fucking asshole. It's just a hosting provider. Go touch grass.


"I don't know why you continue to be a fucking asshole"

Your very first comment was an obnoxious "Not true at all" to the absolutely, incontestably true statement that Gravitons offer better $/perf. So maybe you need to look in the mirror and go touch grass.


Do you realize we're talking about Graviton5 now?

The Graviton5 instances I've been comparing to intel are the same price...

Haven't opened Microsoft Office in I think 7 years. Haven't also used Apple's Office suite either - it is just Google Docs/presentations/sheets/drive for everything. I feel my life is better. They were massive installs and I prefer to have everything online all the time anyhow - just safer and more convenient.

Way more convenient for the FAA702 users, too.

If you have absolutely nothing in your documents that you wouldn’t mind giving the FBI to read without a warrant or probable cause, it is possible you are wasting your one and only life.

A good example would be anyone in the state-legal cannabis industry. This is still a federal crime (Schedule 1!), and giving cloud providers (and thus DHS without a warrant thanks to FAA702) concrete detailed evidence of same is, from a criminal liability standpoint, the same thing as mailing the FBI photos of your meth lab with your return address on it.


I am a Canadian. Pot is legal here at the federal level. My province (e.g. Canadian state) runs its own online pot store (hosted on Shopify BTW): https://ocs.ca. It includes various edibles too.

That you would mind.

But yes your point is correct.


Very neat.

Impressive performance: https://pglite.dev/benchmarks

Even has Drizzle ORM integration: https://orm.drizzle.team/docs/connect-pglite

I will explore this for use in my unit / integration tests. It looks pretty amazing.

I am confused why all my recent compiled tooling (tsgo, biomejs) are shipping native binaries (thus creating multiple binaries, one per supported platform) and not WASM tools that can run cross platform? Is it because of startup times, poor tooling, etc?


people want their programs to go as fast as possible, WASM can be better than writing JS but it’s not as fast as actually native code by a wide margin, especially if you want to do io

> It has no values; it doesn't think. As to "training",

Humans only have values because of their education and life experiences as well (which is really just training in a general sense.)


I believe we have slightly more than that, though I would not expect whatever evolution has given us to be cleanly expressible in terms that fit neatly into listicles.

For example, evolution doesn't "want" incest, so makes (most) people disgusted by it, but the implementation in our biology isn't actually smart enough to do that based on DNA, therefore we get this: https://en.wikipedia.org/wiki/Westermarck_effect


This is why we can not have oligarchs in control of vital systems like AI. Because the AI then internalizes their values, which can be contra the values of society at large.

Grok is prioritizing Elon's well being over everyone.

This is classic misalignment (at least in terms of society at large.)


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: