PicardsFlute

So, instead of resolving it privately like you were asked, you come here and complain? Was the bot posts on Reddit not enough for you? Pro tip: There is a right way and a wrong way to approach these things and you are MOST certainly approaching this the wrong way. And before you go accusing me of anything: I am NOT associated with that project, but I DID see what happened. You are acting like a damned child and should be ashamed of yourself.

The TTFT benchmarks don’t look right to me. I don’t use vLLM, but at 16k pre-fill, the M5 Max is 3.6 times faster than the M4 Max. The 5090 is surely faster, but the numbers in the article are not reflecting what I have seen thus far. Perhaps vLLM hasn’t been updated to use the new tensor APIs for metal?

My point is this: The M5 should have reflected this in the charts, but it doesn’t. The situation on pre-fill is not nearly as bad as in the M4 generation.

Thank you for posting that. I was pretty sure the M5 was going to ship with MTE, but the last time I checked the documents, they still hadn't updated them (nor any mention of M5 having Apple10 in the metal feature tables). Some big features there that makes me want to upgrade!

And there it is...like clockwork. Absolutely no one saw this coming /s

Thanks for posting this interesting tidbit! I find this kind of knowledge absolutely fascinating!

M1 Max 64GB 4TB. Honestly surprised about this. I was expecting MUCH more use since I do video editing/coding/remuxing. I got this at launch

=== START OF SMART DATA SECTION === SMART overall-health self-assessment test result: PASSED

SMART/Health Information (NVMe Log 0x02) Critical Warning: 0x00 Temperature: 31 Celsius Available Spare: 100% Available Spare Threshold: 99% Percentage Used: 0% Data Units Read: 48,482,411 [24.8 TB] Data Units Written: 32,693,167 [16.7 TB] Host Read Commands: 570,658,787 Host Write Commands: 280,299,147 Controller Busy Time: 0 Power Cycles: 216 Power On Hours: 265 Unsafe Shutdowns: 12 Media and Data Integrity Errors: 0 Error Information Log Entries: 0

I didn't work at Capital One, but this sounds familiar to what I saw working at another big bank that I will not name. An entire division basically got laid off with a select few people being "promoted" to "Principal Engineer" (more work without pay raise) and they replaced everyone else with contractors. It did not go well. Last I remember, the exec they pulled in that spearheaded this reorg left after only being in that position for a year or so.

I do not miss working there.

That sounds interesting. What is it called?

Its decent. Forget about any modern DX12 games, but I can count on one hand the number of games that have not worked properly outside of that. Considering the type of games I play now-a-days, it's not too terrible of a loss to be honest.

If I am in the mood for something that doesn't work or more modern than say The Witcher 3, I just boot up the desktop in windows. Its like the best of everything: I don't have to use the hell that is Windows 11 for general computing and I can treat the desktop as a compute server when I am crunching numbers.

Doesn't the game use "skill" based matchmaking? If so, then those were bottom of the barrel "tryhards". I had a similar experience in CS:GO YEARS ago when I bothered to play matchmaking except instead of being reported, I got repeatedly team killed for playing the objective or something.

My experiences in overwatch weren't any good either. I remember the match that made me stop playing very well too. We lost, but I had gold in every category with my two friends taking silver and bronze. Essentially, the other two people weren't playing well at all and started spouting the most toxic shit I had heard in a long while. I guess fuck us because we were just playing for fun and still managed to do better than them? Like it wasn't a blowout or anything either.

I just don't understand some people I guess. Your post managed to bring that memory and reminded me why I don't play competitive games anymore lol. There are plenty of toxic players out there, but in my experience, a lot of them are pretty bad and are most likely projecting.

I think you were getting trolled. The auto focus from day one has been excellent and has only gotten better as they have been releasing updates. I own one and don't really care to take part in forum mudslinging, but I would seriously question anyone who thinks that the autofocus is anything but good (for MILC).

I can answer that as I own a Z9 and was I am interested in the tech as well. Essentially, it is using a "Stacked CMOS" sensor that has a readout time of ~1/270 s. For comparison, the Z7ii is 1/30s if I am remembering correctly. Essentially, it just reads out the data on the sensor fast enough to hide any shutter artifacts (with the exception of LEDs, but this happens even with mechanical shutters as well, and isn't that big of a deal)

The only other camera from the big 3 flagships that matches this speed is the Sony A1 with 1/250s. The Canon R3 is 1/180s. It has a mechanical shutter, but I think the Z9 proves that its worth the omission in 99% of the shooting scenarios. Especially when it is coming in $1000 cheaper than the Sony.

Edit: I was beaten to it!