It was purely hypothetical, but why wouldn't it be feasible to spend your cash and liquidate all of your short-term investments? Obviously, a company needs some amount of working capital to operate, but then revenues wouldn't ever go all the way to zero, either. As I said, it was hypothetical.
The main point was they have a cushion that should be more than adequate.
Sure, we all know they can survive, they survived for many years on almost no money during their bulldozer time, I just don't think that anybody wants a repeat of that. They need to hang on to that cushion so that they can spend it on research and wafers.
Can you or @jkflipflop98 shed any light onto what Intel means by "tape-in"? I'm seeing this phrase a lot more, lately, and I don't know if there's even a standard definition.
FWIW, my general understanding of "tape-out" is when the design phase of a chip is complete and the design is sent out to manufacturing, for production the first engineering samples.
If Habana doesn't start getting market traction now, I fear they never will. Or, at least that's what I'll bet some at Intel are thinking. It would be like almost nobody buying your GPUs, during the crypto boom.
As I understand it - and I'm sure it's probably changed by now - tape-in means that a section of the overall design is complete. So if you have a multi-chip package and your compute cores are design complete they're "taped-in". Then you tape-in the GPU then the I/O package. Once the whole thing is design complete it becomes the normal old "taped-out" and it's ready to start production.
Intel retained earnings (accumulated deficit) from 2010 to 2024. Retained earnings (accumulated deficit) can be defined as profits reinvested in the corporation after dividends have been paid out. <ul style='margin-top:10px;'> <li>Intel retained earnings (accumulated deficit)...
www.macrotrends.net
Intel Annual Retained Earnings (Accumulated Deficit)
(Millions of US $)
2022
$70,405
2021
$68,265
2020
$56,233
2019
$53,523
2018
$50,172
2017
$42,083
2016
$40,747
2015
$37,614
2014
$33,418
2013
$35,477
2012
$32,138
2011
$29,656
2010
$32,919
Yes my local brick and mortar intel corner store is shacking in their boots...oh wait, they are a completely different type of company, one that doesn't get killed by online stores being much cheaper and more convenient.
Also intel is making the CPUs that people need, they aren't trying to upsell you on a server CPU.
"It's not that Intel puts out bad processors but having an attitude that 'he's going to have to spend that money sometime' and treating me like I'm a criminal for wanting good value for my money is a great way to get punted to the curb."
Nobody is threatening anybody, or treating you like a criminal.
But the response to "they have to reduce prices every time the market goes soft" is still that people have to update sometime (let alone all the new people and new companies) so you just have to wait it out.
A&P's decline started in the 1960's. During that decade:
-Amazon was still considered a river in South America
-E-Tailing was considered Texas slang for detailing
-The business plan for Federal Express was submitted as a college term paper (fun fact: the professors told future founder Fred Smith it would never fly 🤣)
Sears, Roebuck and Company went into decline in the 1980's. During that decade:
-ARPANET (the internet) was first developed
-Amazon was still a river in South America
-Walmart actually became the largest retailer, surpassing Sears
In fact Amazon didn't even show up on any radar until the 1990's (I remember when they were still a bookstore lol). And all that was left by then was to finish them off.
So if you don't think Intel can be beaten, just think about what I just said.
Any business can be beaten, if they forget to take care of their customers.
I'm not saying Intel has to lower prices. They just need to remember customers have needs, and those needs change, and you need to change to meet them!
I'm not saying Intel has to lower prices. They just need to remember customers have needs, and those needs change, and you need to change to meet them!
One of the main problems Intel currently has is too much inventory. They over-produced Alder Lake CPUs by a lot. They have to clear out that inventory somehow and they have a limited time period to do that, before demand shifts completely over to the next gen.
A core part of the deflation phenomenon is simply the expectation of it, which suppresses demand and makes price-cuts a self-fulfilling prophesy. Terry is trying to lower expectations of further price cuts, so that anyone in the market for an Intel CPU won't have cause to remain on the sidelines.
One of the main problems Intel currently has is too much inventory. They over-produced Alder Lake CPUs by a lot. They have to clear out that inventory somehow and they have a limited time period to do that, before demand shifts completely over to the next gen.
A core part of the deflation phenomenon is simply the expectation of it, which suppresses demand and makes price-cuts a self-fulfilling prophesy. Terry is trying to lower expectations of further price cuts, so that anyone in the market for an Intel CPU won't have cause to remain on the sidelines.
Those prices are not going to come down much until the next gen of chips arrives. In the meantime, they will simply reduce production so they can clear out inventory.
But the bigger question is why aren't they selling? My guess is Intel failed to meet the needs of the customer, so they won't buy them.
Let give you my line of thinking: I'm concerned about sky high utility bills. I have a choice between two chips: Ryzen 9 (the x3d model) and the I9.
Both perform really well, but the I9 doubles as a space heater, where the Ryzen 9 uses half the power to get the job done.
Same reason as most other tech-related stuff that is recording some of its lowest sales in 10+ years: post-COVID sugar rush crash. Everyone who wanted or needed an upgrade already did so and won't be looking to upgrade again for another 3-7 years.
When most of your business is recurring corporate and government refresh cycle on a 3-5 years schedule where a different subset of those come due every year for a steady overall demand but now all sync'd up within 1-2 years of each other, the next 3-4 years until demand evens back out again will be gnarly.
At the same power running a bunch of server type apps ryzen wins by an average of ~13% , so intel will run for 13% more time using up 13% more power to do the same work that would be around 18W per more hour running.
At idle the ryzen runs at 250% the power the i9 runs at if it is matched with the slowest ram.
That's 10W more per hour you are at idle or 15W more per hour if you have fast ram.
AT single it's also around 10W.
Which would you choose if you didn't need to run server workloads for many hours per day?
And don't start with "out-of-the-box" or "default" because ryzen CPUs blow up at default settings and you have to reduce voltage by hand to avoid that.
And don't start with "out-of-the-box" or "default" because ryzen CPUs blow up at default settings and you have to reduce voltage by hand to avoid that.
And not every intel mobo is set up to run at 300W + ... for it to use twice the power on an i9 than an r9.
There are mobos that limit the CPU to 65W even if it is an k cpu.
As others have posted, this is just the market correcting itself after the surge of COVID-19 related purchasing. Intel exacerbated the issue by not realistically meeting it's customers expectations and giving an opening for AMD to take market share from them.
As for x86 itself, it was IBM and not the US Government that required Intel to provide a secondary license to another manufacturer. IBM made this requirement to every component vender for it's IBM PC platform. They had seen several other PC platforms fail due to vender related issues and wanted to protect itself from vender supply blackmail. The entire PC industry as we know it today was a direct result of IBM requiring all it's venders to have compatible standards, and is why we see the words "IBM PC compatible" printed in small letters on every consumer part made. If one CGA card manufacturer was having "issues", IBM could just get a CGA part from a different manufacture, same with memory, monitors, keyboards, storage and even CPUs. With every single component working off a standard IBM could mix and match suppliers and have them compete with each other. This model was a resounding success and lead to IBM becoming as big as they were. Of course this also meant that once someone reverse engineered the IBM PC BIOS software, the OEMs didn't need IBM anymore to sell compatible clones of the platform that would run all the same software.
Without IBM's insistence on multiple venders with open hardware designs, we get platforms like the Apple II and Amiga, which while technically superior, end up being expensive and have difficulties maintaining an expansive software library from one generation to the next.
Facebook
Twitter
Instagram