Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
hajile
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
38 ms
·
511.
▲
by
hajile
2y ago
Yeah. If they need all those certifications, I’d add another zero to the price. Things are different today. The ISA helps reduce lock-in and a lot of otherwise proprietary costs get amortized across companies and industries. There’s more co
512.
▲
by
hajile
2y ago
The price will be very high, but the vector unit isn't the reason at all. These target a very specialized market and they won't be selling billions of them like they do with other PIC designs. If a specialized chip costs $250M to
513.
▲
by
hajile
2y ago
HALF of X1's compute is F16 only which is absolutely wasted silicon for most desktop games. Their entire tiling setup is great for simple mobile games, but (as shown in the article) is also an inefficient approach for desktop games. 64
514.
▲
by
hajile
2y ago
Paying big prices for old AAA games on a substandard UI isn't a winning proposition. Most people already have those old games on Steam or their console if they want them. While we're on that topic, using the App store generally su
515.
▲
by
hajile
2y ago
I don't think SIMD and SMT is an either/or proposition. SMT-4 or SMT-8 with a bunch of SIMD has the potential to get better perf/area due to the threads hiding the latency.
516.
▲
by
hajile
2y ago
In the US, the Right to Freedom of the Press has NOTHING to do with "journalists" and everything to do with the freedom for ANYONE to write, publish, and distribute whatever they'd like. If I as a US citizen didn't sign
517.
▲
by
hajile
2y ago
Most reasonable people would denounce BOTH. You seem to be pushing toward the idea that "if they do something evil, my evil is no longer evil".
518.
▲
by
hajile
2y ago
I wonder if they were inspired by Meego SwipeUI on the Nokia N9? It was visually quite different from everyone else at the time.
519.
▲
by
hajile
2y ago
Do you have any evidence for this statement? Their actual core wasn’t a Pentium. It added in most/all the x86 bloat from the last 3 decades too. Cutting all that stuff and saving power means you can squeeze in a couple more cores and i
520.
▲
by
hajile
2y ago
RVV was only ratified Nov 2021 (not too far from the release of SVE2). It shouldn’t be surprising that adoption will take some time. The first cores using the final extension are hitting the market this year, so addition is happening really
521.
▲
by
hajile
2y ago
Larabee was launched as the first generation Knight-series chip (it even had the GPU-specific stuff still in there disabled IIRC). I believe the Knight-series chips were killed because they ran into a bunch of issues and politics with Knigh
522.
▲
by
hajile
2y ago
The x86 tax ALREADY matters. Intel was only able to increase the number of decoders by adding entirely independent decoder complexes while reducing the number of decoders per branch. In practice, this means that decode increases for branchy
523.
▲
by
hajile
2y ago
Market cap for tech companies is a bubble. It has more in common with the lottery total than with a representation of real value.
524.
▲
by
hajile
2y ago
You are missing the point. We're discussing if Apple can use their own chips more cheaply than buying Nvidia's chips. The Max TDP is not actual peak power consumption. Gamer's nexus recorded 500w peak and almost 670w overcloc
525.
▲
by
hajile
2y ago
If they can get 5 4070s for the price and power of one 4090, that's a win for them as they'll get more performance per dollar and per watt.
526.
▲
by
hajile
2y ago
They don't need the entire mac. Their cost per Max chip is probably $200-300 which beats the 4090 by a massive margin and each chip can do more than a 4090 because it also has a CPU onboard. 4090 peaks out at around 550w which means th
527.
▲
by
hajile
2y ago
If automation didn't reduce labor costs and overall costs, nobody would do it. The jobs provided to the guys making the machines are paid less overall than the jobs they replace. Most modern jobs are so-called "bullshit jobs"
528.
▲
by
hajile
2y ago
And massive minimum wage hikes driving up costs until they automate those jobs away.
529.
▲
by
hajile
2y ago
You can add anything you want to a contract, but that doesn't mean the courts will uphold it.
530.
▲
by
hajile
2y ago
In the US, the police have NO legal obligation to help or protect you. "Rights to healthcare" ultimately means "rights to enslave healthcare workers". If healthcare workers refuse to serve you, you have no healthcare unl
531.
▲
by
hajile
2y ago
The pro private gun ownership argument.
532.
▲
by
hajile
2y ago
That kind of clause won't stand up in court and they know it.
533.
▲
by
hajile
2y ago
Darktable is decent (in my experience). https://www.darktable.org/
534.
▲
by
hajile
2y ago
In this case, you tell them to either settle for $0 or you sue for breach of contract.
535.
▲
by
hajile
2y ago
Avoid class action and take them to small claims court or simply backcharge them and let your credit card handle it. Either way, they lose a lot more than a class action that will take years and get you nothing.
536.
▲
by
hajile
2y ago
I hear this argument, but I don't really believe it. If you're talking apps before say 2015 (10 years ago), they can be emulated on ARM faster than they ran natively. That rules out 95% of the backward compatibility argument. Most
537.
▲
by
hajile
2y ago
There are definitely workloads where turning off SMT improves performance. SMT is a crutch. If your frontend is advanced enough to take advantage of the majority of your execution ports, SMT adds no value. SMT only adds value when your fron
538.
▲
by
hajile
2y ago
There are a couple methods that could be used. Static analysis would probably work in this case because the in-order core would be very GPU-like while the other core would not. In cases where performance characteristics are closer, the OS
539.
▲
by
hajile
2y ago
L0 latency is 4 cycles while old L1 was 5 cycles, so it's definitely a new cache even though the size is the same. Apple does a lot better. I'm not sure about newer chips, but M1 has the same 192kb of L1, but with the 4 cycle late
540.
▲
by
hajile
2y ago
I think the answer here is dedicated cores of different types on the same die. Some cores will be high-performance, OoO CPU cores. Now you make another core with the same ISA, but built for a different workload. It should be in-order. It sh
More ›