Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Intel Honesty

stratechery.com

Post reply on HN
Print view View original

351–360 of 426 posts

Re: Intel Honesty

#351

Post by seec »

keepamovin wrote: post #314
svnt wrote: post #279

Earlier quoted context omitted.

I worked there on a contract while they were building a phone and had as many as six managers over six months. One of them would come in at 11, talk to me about nothing until lunch, then come back after lunch, talk more, and go home. I had to come in early to get anything done. When I was toward the end of my project and sure I wasn’t going to try to get a blue badge I laid into someone in a meeting. It wasn’t my bes…

This is a great indicator of how it's often presented as collaborative or even slightly-"woke" on the outside, when in reality, advancement is cutthroat. But conducted in an cliquey/politically repressive atmosphere where everyone is afraid to act like that, but when someone has the courage to start speaking their mind, the rest look to them as the one who can save the malaise from the stagnation. I wonder how it bec…

I believe that any successful company is like that. There are always people that have a vested interest in pretending otherwise but that's just to avoid other people succeeding or to present a good front.

As far as I'm concerned there is no other way to get real progress; there are always a lot of pretenses on better ways to "manage" stuff but they never work because of group dynamics and basic human behavior...

Re: Intel Honesty

#352

Post by seec »

Panzer04 wrote: post #338
bryanlarsen wrote: post #46

Earlier quoted context omitted.

By what metric do you consider INTC to be low? I took a look because I'd be happy picking up some cheap INTC, but it still looks expensive to me.

Their revenue is >twice AMD's. If IFS disappeared they'd probably have a better valuation than they do right now solely based on processor performance, but their success in the foundry business is probably contributing negative value to their market cap (presumably because it's soaking up so much cash for nil returns) Intel upside right now is possibly very high, if they can fix the foundry. Their current stock is pr…

And it seems like they don't have a whole lot to fix to be back on top or at least extremely competitive. They know how to run foundries, they just need to catch up on the tech side, it's going to take a bit of time and cash but it's not like if they don't have experience doing it previously.

On the chip design side of things, they seemed to have almost caught up (at least to AMD/Qualcomm, Apple is another story) because their upcoming low power laptop chips generation looks like it will be very competitive in terms of power per watts. And they benefit from years of refinement in the software side as well as quite competitive GPU stack, people seem to forget too easily but AMD products still have many annoyances that Intel ones just dont (recently WIFI issues on their latest laptops) and Qualcomm is a no go unless you do browser-based stuff for the most part.

I don't understand people giving up Intel for dead, because even at their worse in the past few years, they were still competitive, both in performance and price; once they figure out the process and design (seems to be on good way) I would worry a lot more about the other side of competition, but who knows...

Re: Intel Honesty

#353

Post by seec »

mjevans wrote: post #220
fluoridation wrote: post #119

Earlier quoted context omitted.

Roughly speaking, that running Linux as a desktop OS would be easy enough that even non-technical users would be able to do it with difficulties comparable to, say, Windows XP.

The goal posts always seem to move. Windows is an ugly pain in the butt to create install media for (when you're not on the native OS), and it's just as bad to Install on random hardware. Maybe more so since it's less likely to ship network drivers for random wifi gear and lan connections. Even Debian has non-free blobs for drivers with the (at least 'Live' versions) these days, finally. Repair shops are the next big…

Driver issues on Windows only happen very early in the hardware cycle. If you buy cutting edge hardware and try to install a regular/old version of Windows it might not work automagically. But it's really no big deal, not only can you load the drivers from USB just fine but you also can probably just ignore until you get to the desktop because the OS will use a generic replacement or do without in the meantime.

I know for sure because my PC was a cutting-edge HEDT at some point and indeed the network drivers weren't present at installation and GPU driver wouldn't auto-install either. But it just meant I had to right click install the network after desktop load and just download the GPU drivers from the internet.

Now that the PC is old, any re-installation is extremely straightforward with everything automatically loaded and basically zero intervention to get fully functional hardware on Windows.

The thing about Linux is that not only it wouldn't have worked at all when the hardware was cutting edge but even today, some patching and manual intervention is required for the network/motherboard that is not usual. It's a process way more involved and annoying than I ever had to do on the Windows side.

Any fool can go on the manufacturer website and download a driver bundle to execute sequentially, just figuring out what you'll need to do on Linux is a problem in itself...

Re: Intel Honesty

#354

Post by arder »

jhallenworld wrote: post #90

The Altera merger pain is maybe enlightening. Altera 10-series FPGAs were massively delayed due to Intel's 10 nm fab problems. But I speculate that there was also a big difference in toolchains, does anyone know for sure? I mean Altera was using TSMC previously, and I presume were using industry standard tools: Cadance / Synopsis. But I would guess Intel was running their fab on home-grown tools... what is the status…

The Altera merger was a perfect storm. I could write for pages about what went wrong. But here's a few key things: First FPGA companies have been massively hit and miss with their product - they were perfectly capable of screwing up their next chip all on their own (it's a mutually beneficial duolopy of crapness with Xilinx). Second, when Altera agreed to fab with Intel the deal was written such that Intel could never buy any FPGA company other than Altera, so from day 1 this wasn't "Let's produce this together" it was "Let's line up Intel to buy Altera" and so Altera never had to deliver anything, they just had to wait for the acquisition. Third, Altera were one of Intel's first customers and it turns out that Intel's fab process couldn't make some of the gates that Altera needed, which is why in the end they partitioned off the entirety of the transcievers to a separate tile fab'd by TSMC and sellotaped to the edge of the Stratix 10.

Also just to check your facts on the 10nm fab problems. Intel's first FPGA for Altera was always planned to be 14nm. It was totally trash for reasons entirely within Altera's control (don't rock the boat, just wait for the acquisition to close). And the synergistic products were a dead end. I'm sure the 10nm catastrophe didn't help but really that time would've been well spent unpicking the disaster of Stratix 10.

Re: Intel Honesty

#355

Post by yvdriess »

bluecalm wrote: post #343
Panzer04 wrote: post #337

Earlier quoted context omitted.

Intel's revenue is 54B to AMD 22B. If Intel's CPU division was spun off, I see little reason to believe AMD deserves a 3x premium to that - and you get IFS with Intel as well right now! Granted, IFS is the millstone hanging around their neck right now, unless they can fix it.

Intel still has 75% of data center market share (for CPUs). I think this is mainly momentum and they will lose most of it in coming years.

To add to that, AMD's has the more lucrative higher-margin part of that market.

Re: Intel Honesty

#356

Post by qwytw »

roenxi wrote: post #287
qwytw wrote: post #241

Earlier quoted context omitted.

> Is a market where housing supply meets demand, housing is a depreciating asset. How so? Supposedly house prices increased by 50% in 5 years there. So obviously it does not meet demand... > Why would you want to buy if renting is cheaper? Because it's usually considerably cheaper long-term? Rent prices go up all the time, so you'd lose a very significant amount of money over 10-20+ years.

If it is cheaper to buy than to rent, then presumably they can buy a house. The difference in prices just has to be bigger than interest rates and it becomes cost-effective to get a loan. Say rents are >6% of the equivalent cost of ownership, then a mortgage at 6% will be cheaper than renting. So renting can't get that out of control relative to house prices in the long term, they have to go crazy together in a way t…

I mean in normal markets you'd expect price-to-rent ratio to up to 20x. If it's higher buying probably doesn't make a lot of sense unless you expect prices to continue growing at a very fast rate (which of course makes it too risky if it's your primary home).

In Taipei it seems to be about 50x.

So if you live there and want to have children it's certainly not cheaper (you can't afford it anyway) and you end up paying already very high and continuously increasing rent payments all your life. By the time you retire you have no equity and very little savings.

Situation is the the same in a lot of major cities. It's just that rents are naturally caped at some proportion of median income (that leaves you with very little disposable income) while real estate prices have become almost completely detached from it.

Re: Intel Honesty

#357

Post by eigenspace »

umvi wrote: post #180

> today AMD has both better designs and, thanks to the fact they fab their chips at TSMC, better processes On paper AMD is better, but in the scientific community, it seems that Intel has much better performance for things like NumPy and SciPy. The reason seems to be "Intel® MKL" ( https://en.wikipedia.org/wiki/Math_Kernel_Library ). I'm using a lot of weasel words like "seems to" because I haven't rigorously proved…

> But anecdotally, my company's AI pipelines which are NumPy/SciPy heavy run an order of magnitude faster (2 seconds vs. 20 seconds) on my laptop's Intel i7 than the do on my Ryzen 7, despite the Ryzen 7 being a newer gen than the i7.

I very very seriously doubt this is a like-for-like comparison. I suspect one case is multithreaded and the other not, or that one is using generic non-SIMD fallbacks or something like that.

Re: Intel Honesty

#358

Post by starspangled »

jack1243star wrote: post #335
starspangled wrote: post #318

Earlier quoted context omitted.

They weren't a startup for 20 years, they were a huge company with billion dollar fabs, thousands of workers, and they were exchanging technology lead with other silicon manufacturing companies many times (back before 2020s, mid/late 2010s, TSMC was not undisputed leader). UMC was a Taiwan foundry that started even before TSMC and it eventually folded.

>UMC was a Taiwan foundry that started even before TSMC and it eventually folded. What are you even talking about...

Same thing as everybody else in this thread, leading edge logic market.

Re: Intel Honesty

#359

Post by heresie-dabord »

y-c-o-m-b wrote: post #270
AnotherGoodName wrote: post #11

Intel is trapped. Its debt repayments alone are massive. The poor performance in stock has encouraged 20years of wage stagnation to the point where you can literally earn over 50% more at amd or nvidia or even an startup for equivalent roles so they aren’t hiring the best. Their pay in the Bay Area is double tsmc in taiwan for equivalent roles in raw dollar terms but the ppp differences means your better off working…

I worked there for a few years a little over a decade ago in Oregon. The wages were abysmal, so you're not wrong, but they still drew a lot of good talent because Intel was more accessible to the suburbs than the city of Portland (commuting to downtown Portland was awful) and remote options were limited. Krzanich era layoffs and the availability of more remote work options resulted in a rapid drain of talent though.…

Your second paragraph, with a slight reworking, is a succinct diagnostic for all bloated and dying corporate entities:

"Diagnostic for immense failure. Does the organisation...

= enable a toxic working environment that is openly discussed,

= encourage internal conflict act every level, business unit and staff,

= cultivate top-heavy management, promote low-talent that makes poor choices and burns money on worthless projects that never deliver,

= maintain the threat of layoffs to destroy morale and engagement,

= send critical functions off-shore and/or to third-party contracts,

= fail to do strategic research, make poor choices in acquisitions and partnerships, and build unnecessary facilities?

If so, the patient manifests late-stage corporate morbidity."

Re: Intel Honesty

#360

Post by 4ggr0 »

kridsdale3 wrote: post #129
FredPret wrote: post #98

Earlier quoted context omitted.

It's grim at INTC but don't forget the CHIPS act and whatever else will follow that. The US government spends 0.5T per month - more than the market cap of all but the top 15 American companies [0] [1] Not saying it's a good or bad idea, or that it will or won't happen - but if the US government decides to reinvent Intel, they can easily write a cheque that (if spent wisely) might do the trick. [0] https://companiesma…

Why doesn't the USA just buy out all of Intel's debt, let them start fresh? Call it a Silicon Era War Bond, in reverse.

Socialize losses, privatize profits. How cute.
Post reply on HN