my123

63 ms·

811.
by my123 6y ago
Cinebench is a rendering load that isn’t that much optimised. (Doesn’t even use the newer AVX levels when available, and isn’t properly optimised for Arm either). Cinema 4D, the program that Cinebench benchmarks, normally does the renders o
812.
by my123 6y ago
It's far from being as definitive as you say. Trying to find with one of the best 5800U scores: https://browser.geekbench.com/v5/cpu/compare/7449909?baselin... It's for 5800H which is at 45W where R
813.
by my123 6y ago
Then just turn SIP off via csrutil disable. You can then debug anything, load unsigned kernel modules...
814.
by my123 6y ago
Nope, about Zen 3. 5800U (Zen 3 mobile) and M1 are actually very close multithread perf wise despite one being a 8 core part and the other a 4+4 part. And with more power use for Zen 3 too. What hurts Zen 3 there is while the M1 maintains
815.
by my123 6y ago
A dual-core Cortex-A9 with 1GB of RAM chokes at almost everything today even when updated. Source: Surface RT today is still getting security updates. And will continue to get them until 2023. However, it doesn’t matter much in practice. An
816.
by my123 6y ago
And are still funding that CPU design team, which is designing some upcoming products.
817.
by my123 6y ago
A VSCode extension can just run code in a host context anyway. It doesn't really matter there.
818.
by my123 6y ago
It’s the macOS/iOS ABI solely. It isn’t related in any way to the hardware itself.
819.
by my123 6y ago
It’s not the M1 chip by itself, but rather is macOS/iOS.
820.
by my123 6y ago
They use LLVM, but without the Clang part. (even available to 3rd-party apps as NVVM, which currently takes in LLVM 7 byte code) Upstream Clang supports compiling CUDA code however.
821.
AWS and Nvidia to bring Arm-based instances with GPUs to the cloud(aws.amazon.com)
2 points by my123 6y ago|0 comments
822.
by my123 6y ago
Yes that’s what you do, you might want to look at options other than Qemu though for more performance. (I assumed FSE from your 40x worse perf figure, which is a very bad case)
823.
by my123 6y ago
You should try running an Arm64 OS and then using ExaGear (Linux VM) or just run your x86_64 app (Windows). Forget about full system emulation, especially with Qemu’s approach which doesn’t leverage the host MMU. Emulating an MMU alone is v
824.
by my123 6y ago
It’s strictly x86 only, and wasn’t written with other platforms in mind at all. A port to Arm wouldn’t be really a port but a rewrite.
825.
by my123 6y ago
Qemu isn’t representative of anything else perf-wise. You may want to use user-mode emulation too.
826.
by my123 6y ago
https://news.ycombinator.com/item?id=26367726 Not anymore.
827.
by my123 6y ago
At least Windows has EGL incl streams for some scenarios (related to WSL’s graphics stack mainly). On more embedded scenarios that I’ve worked on (Qt w/ EGLFS), it’s quite popular too.
828.
by my123 6y ago
GVT-g is gone on 10th-gen GPUs (Ice Lake) and later. Not supported on Intel dGPUs either.
829.
by my123 6y ago
The RTX A6000 is at USD 4650, with 48GB of VRAM and the full chip enabled (+ECC, vGPU, pro drivers of course) The RTX 3090, with 24GB of VRAM is at USD 1499. Customer dGPUs from other HW providers do not have virtualisation capabilities eit
830.
by my123 6y ago
On Linux. EGLStreams is the actual Khronos standard ( https://www.khronos.org/egl ) as such existing on Windows, QNX and others.
831.
by my123 6y ago
Newer doc: https://download.nvidia.com/XFree86/Linux-x86_64/465.19.01/R... And https://download.nvidia.com/XFree86/Linux-x86_64/465.19.01/R... for power management, including t
832.
by my123 6y ago
Don’t use Bumblebee anymore. Use this instead: https://download.nvidia.com/XFree86/Linux-x86_64/440.31/READ...
833.
by my123 6y ago
For a cloud infrastructure, you have a significant part at idle, for when customers want to instantly spawn a VM.
834.
by my123 6y ago
It’s more like changing the BIOS/UEFI chip, as that’s stored there. You can always plug in a second SSD via a PCIe slot too.
835.
by my123 6y ago
https://github.com/libimobiledevice/idevicerestore (Works for restoring T2 Macs too)
836.
by my123 6y ago
Those are protected by patents, not copyrights.
837.
by my123 6y ago
With TURN, all the traffic to the clients is routed through the TURN server indeed. That makes hosting a discussions server more traffic-heavy than otherwise... (and it turns out that the server software that I use implements TURN and STUN
838.
by my123 6y ago
I use https://github.com/coturn/coturn , provided as the coturn package on Ubuntu 20.04.
839.
by my123 6y ago
As I said, "without going through a relay". And TURN is one of those relays. (I host a STUN and TURN relay myself, because I had to for my personal VoIP server for enough people to be able to connect on it. Downside is more use of
840.
by my123 6y ago
Far from only a feature for pro users. Notably, it is a must for VoIP (without going through a relay) and BitTorrent when you don't want to manually configure a firewall. (allows to create holes in a controlled way for a NATted network

More ›