Pedro Justo

🇵🇹

[Drew Lipps @dlippsYT]

I was wrong about Windows Arm Laptops.

After spending the last few days throwing everything at the @ASUS Zenbook A16, I'm impressed! The hardware's premium & the @Snapdragon X2 Elite Extreme performance is incredible! 🔥😎

Full review link below 👇
#Gifted_By_Snapdragon

[World Rally-Raid Championship 🚗 🏍 @OfficialW2RC]

🔥 Stéphane Peterhansel conquers Argentina again! 🇦🇷🏆
Together with co-driver Michael Metge, he claims the overall Stock victory after a dominant performance throughout the week.

*Subject to the official publication of the results by the FIA
#W2RC #DesafioRuta40 #FIM #FIA

[Pedro Justo @itanium_guy]

@sahajsarup That is a good question. I don't which were the talent migrations into AMD leading up to and out of AMD after the first AMD64 came out. Maybe an Acquired episode on AMD will clear that up ;-)

[Pedro Justo @itanium_guy]

🤯

[Pedro Justo @itanium_guy]

@supersat @__alula With the introduction of the AVX emulation, Arm64EC ABI also changed. It is now legal to use ALL Neon registers (after an AVX feature check). On the flipside, the compiler started using x28 on Classic and, so the GP diff is now 3 and SVE is still not allowed in EC.

[Pedro Justo @itanium_guy]

@supersat @__alula Note that the loss in registers is actually small for GP registers - only 2 (x28 was originally reserved and the compiler never used it even in Classic Arm64 compilation). Different story for Neon, where it is restricted to 1/2 bit it is less common to suffer register pressure.

[Pedro Justo @itanium_guy]

It has been decades since the StrongARM, so obviously much innovation happened since then.
This is a mere gag on the flow of very talented architects which trace back to the DEC days.
I hope they all have a sense of humor 😉

[Pedro Justo @itanium_guy]

@unixterminal @XenoPanther Nested virt is not the best pick for speed. QEMU + WHPX on the other hand...

[Pedro Justo @itanium_guy]

@SebAaltonen Is there such a thing as a cool Blackwell GPU?

[Pedro Justo @itanium_guy]

@HSVSphere @__alula Actually, it is architected in a way that is even more flaxible: DVRTs. DVRTs apply transforms to the PE as it is loaded based on anything environmental. Could be the presence of a hardware feature, errata or... the architecture of the process.

[Pedro Justo @itanium_guy]

@cursedconnector @__alula No. Arm64EC is compiled from source. Produced far more optimal code than translating from x86-64. If you do have the source, this is the way to go. Translate from x86-64 when you don't have a choice (no source access).

[Pedro Justo @itanium_guy]

@__alula If you think about it, it is quite an optimal design. In most cases compiling Arm64EC and Arm64 result in the same code, so storing it twice is a waste of storage (which would have happened with naive fat binaries).

[Pedro Justo @itanium_guy]

@__alula Why is it "cursed" ?

https://t.co/IObLXuQ5h8

[Pedro Justo @itanium_guy]

@pawel_lasek @__alula All that and more in a few minutes:

https://t.co/KpdcWlYYWT

[Pedro Justo @itanium_guy]

@jonmasters BSA is the GOAT.

[Pedro Justo @itanium_guy]

@AmigamagazineGA Isn't this essentially FS-UAE?

[Pedro Justo @itanium_guy]

@ZX48kSpectrum THIS was the game I played the most and longest on my Spectrum. When I think Spectrum Gameplay I think Cyclone. I still play it sometimes.

[Pedro Justo @itanium_guy]

@techknight2 Looks like a Zapa to me...

[Pedro Justo @itanium_guy]

@XenoPanther There are plenty of reasons that may not have anything to do with the OS. For example, are you experiencing thermal throttling? How about losing single-core Turbo? Is your VBox doing blind, overscheduled gang-scheduling?