Donate!
Welcome, Guest. Please Login or Register :: View Members
Pages: 1 2 3 4 5 6 ... 31
Send Topic Print
2020 -- new Intel failures & successes (Read 12299 times)
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #45 - 08/16/18 at 07:37:07
 
 
...... the pitiful squealing sound you hear is Intel, bent over the workbench out in the exercise yard ......



ARM takes aim at laptops with upcoming Deimos, Hercules processors

https://liliputing.com/2018/08/arm-takes-aim-at-laptops-with-upcoming-deimos-...







So what’s next from ARM? The company says its Cortex-A76 architecture is already competitive with an Intel Core i5-7300U processor at some (very specific) tasks, while consuming far less power.

Next year the company plans to release a new “Deimos” design for 7nm processors that will offer a 15 percent performance improvement. Later in 2019 the company plans to launch its new “Hercules” architecture with another performance boost as well as a 10 percent increase in efficiency and power consumption.

Deimos chips will be manufactured using a 7nm processor, while Hercules chips can be 7nm or 5nm processors.

Generally speaking ARM-based chips have a well-earned reputation for consuming less power than their Intel and AMD counterparts, which means you can expect laptops powered by the upcoming chips to offer long battery life and feature thin, light, and probably fanless designs. The chips also often have integrated graphics and built-in support for WiFi, Bluetooth, and cellular connectivity, which reduces the need for multi-chip solutions.

That said, it’s not like Intel and AMD are sitting still waiting for ARM to catch up. Intel’s Gemini Lake Celeron and Pentium processors are low-power chips for low-end computers that tend to use around 5 watts of power, and the company’s Core Y-series chips are higher-performance (and higher-priced) 4.5 watt processors based on architecture more closely related to the company’s 15 watt U-series chips.

All of which is to say, if battery life is your sole concern, you can already find plenty of thin and light Intel-powered computers that offer 8-10 hours of battery life. And AMD’s new(ish) Ryzen Mobile processors for laptops use between 12W and 25W while offering performance that’s competitive with Intel’s U-series chips.

Still, as Intel and AMD duke it out in the top-of-the-line, multi-core desktop CPU race, it’s interesting to see the companies are facing increased competition at the low-power end of the spectrum from ARM.


These are ARM certified SoC designs that are not dependent upon Qualcomm and anybody can license, modify, improve and use them.  This opens up the bottom end of Wintel's world to any of the Hockey Stick boys to go in and take what they want to take.  

These designs can (and will) support laptops, tablets, power phones, if you can name it they can do it.

If you run these designs hard (with proper cooling) up at the same watt/consumption numbers as Intel uses for the i5-7300 they will punch far far far higher up in the Intel line up than just Core i5-7300 ......
Back to top
 
« Last Edit: 10/03/18 at 02:13:59 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #46 - 08/16/18 at 09:42:03
 

https://www.engadget.com/2018/08/16/arm-says-chips-will-outperform-intel-lapt...

Engadget is a PC based publication, not an ARM fanboy publication by anybody's guess.   So, this is opposition press saying Intel is getting ready to go under water in the toilet bowl swirl contest ......

ARM says its next processors will outperform Intel laptop chips     ...... the lines officially cross around December of this year ......




While ARM already believes that its recently unveiled Cortex-A76 is competitive with Intel's 2.6GHz Core i5-7300U, it expects its 2019 "Deimos" and 2020 "Hercules" designs to clearly outperform that CPU. You would get "laptop-class" speed from a more efficient mobile chip, according to the company.

Of course, it's worth taking ARM's braggadocio with a grain of salt. The figures don't include Intel's comparable 8th-generation Core chips that pack twice as many cores and could easily shrink the performance gap. This is also based on one synthetic, integer-oriented benchmark (SPEC CINT2006), not a broader suite of tests that would measure floating point math and other performance traits. ARM is putting its best foot forward rather than offering definitive proof.

Even so, it's telling that ARM might be in the ballpark. There's already talk of Qualcomm making a truly laptop-worthy Snapdragon in the near future -- Deimos and Hercules could be fast enough that you might choose an ARM-based PC for a speed advantage, not just longer battery life. And that could give Intel a good reason to panic.  Intel is still struggling to make 10-nanometer chips at the same time as ARM is talking about 7nm and 5nm parts. If Intel can't find a way to stay ahead, it may see big PC makers switching to ARM as they look for more powerful options.



Look up at the chart, it shows the current per core performance trends from Intel and from ARM.

These curves partially explain why Intel has doubled down and doubled down again on the number of big energy wasting CISC cores that they are using to stay "competitive".

The bogus blog narrative Intel is promoting on the Win sites right now says that Core i5 is a low end chipset, not a mid-range chipset.   We expect Intel to re-name and re-rank all their stuff yet again in a frantic attempt to muddy up the water about them losing out all their low range AND THEIR MID RANGE  to ARM products by the end of the 2018 year.

This contest does not mention AMD who has been following Intel's "many many more big cores" ideas and is simply doing it marginally better than Intel is doing it, by actively using 10nm and 7nm TSMC lithography which is BTW working out just fine for AMD.

Also please remember, by 2020 ARM will be shipping real products using 5nm TSMC SoCs and Intel will just barely by then be shipping Intel built 10nm CPUs (if they actually do make it work this time around).   ARM is currently shipping real 7nm TSMC SoCs for real and TSMC is doubling their production building space for 5nm as we speak as they have already proven out the 5nm EUV process pretty well now on advanced memory products at this point in time.   It works just fine.    AMD and Zhongshan Subor (Chinese domestic producer of AMD designs) have pinched off about half of Intel's current oriental chipset business already, with more to be taken by year's end.

Since TSMC is now planning on using their entire new building for 5nm lines, a new building has broken ground this past month which will be the TSMC 3nm production building which is going to roll in very quickly now that 7nm is a roaring success and 5nm has proven out well making memory.  

Yes, TSMC 3nm starting in 2020-2021.

And the guy writing in blue above doesn't get the point that Intel is using HUGE massive CISC cores in 6-8-12 counts to try to claim "efficiency" when the small ARM A-76 cores have started to individually outperform the massive Intel CISC cores and ARM can easily match whatever core count Intel uses for whatever market they care to contest.  

In short, Intel can't win this race, either in whole or in particulars.    ARM's new designs are built to be Intel laptop killers, that was their intended purpose from design start point that began a year ago.


Roll Eyes          Undecided          Tongue


Yep, Intel is indeed fairly deep down in the whirlpool throat at this point in time, spinning madly and Intel is still headed further down, down, down into the shitter.    

You need to check  your retirement plan, as many plans are deep into holding lots and lots of Intel common stock that just got down rated again by the pundits.   When all the retirement plans quit auto buying Intel, then the house of cards begins to collapse.



Get yourself out of Intel before you go lose a chunk of your retirement money ......


Back to top
 
« Last Edit: 08/24/18 at 07:21:48 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #47 - 08/17/18 at 08:18:59
 

https://arstechnica.com/gadgets/2018/08/intels-10nm-cannon-lake-chip-gets-ano...



Intel, desperate again to show stockholders some sort of moves towards the future, has again run the 10nm line for yet another broken trial production lot, making yet another pile of ONLY HALF FUNCTIONAL CHIP SETS that have lost well over half of the production quantity already on the very first sort after production.      And that sort pre-assumed the GPU wasn't workable, they were just looking for a functional CPU side of the chipset.

So far only a single low volume proposed Intel NUC finished unit has been proposed, no unit production dates or unit costs have been announced.   No units means no performance tests have been done to date.   By doing it all in house Intel gets to hide as much of the ugly stuff in as much as they can ---- HOWEVER the finished units will not be able to be hidden however and will get into the hands of the press and the pundits and the benchmark boys.
 
Next is a pair of NUC Mini PCs named Crimson Canyon. Unlike the kits, these are complete systems: processor, memory, storage, and an operating system (Windows 10 Home). They also have very unusual internals: they use Intel's weird 10nm Cannon Lake i3-8121U (two-core, four-thread, 2.2-3.2GHz), previously only available in a single Lenovo laptop. Intel's 10nm process is fraught with difficulties—the company isn't expecting volume 10nm production until the second half of next year. That Intel is using this chip in new systems suggests that its yields may slowly be improving and that it has more of the processors available than it once did.

One of the things that makes the 8121U unusual (aside from it being the sole 10nm Intel chip) is that it doesn't have a functional GPU, with suspicions that the 10nm yields are so low that Intel can't get the GPU to work reliably. To handle this deficit, Intel is pairing the processor with a discrete Radeon 540 GPU with its own 2GB of GDDR5.


So, you will be paying for a SLOW as molasses Intel SoC that half of it works poorly and the other half simply doesn't work at all and you will then have to pay even more $$$ for an AMD Radeon 540 GPU (complete with video memory) to make up for the Intel inside graphics and on-board memory that Intel can't seem to get to work period.    And the resulting kluged together mess is simply sub-par performance-wise and does not work as well as current 14nm Intel chipsets can do for less money.

Plus, this Crimson Canyon is quite expensive with all the replacement GPU and memory costs that it includes.

And Intel proudly touts this to their stockholders as "great progress in 10nm" as all their competitors are busily leaving the 10nm arena for 7nm and 5nm for CPUs and 3nm for memory.

Intel called the original failed trial lot sorted components "Crimson Trace" perhaps because of the blood smear it left on the tile floor in passing.  Now, this time around Intel calls it "Crimson Canyon" ...... perhaps this is that a code name for "a larger bloody mess" or is it simply commentary on what it does for Intel's production profits for the quarter after running it yet again?

Make no mistake, manning and cranking up a failed production plant isn't a cheap thing to do if you can't readily sell all the product that it makes.    

And designing and producing a special NUC motherboard that allows installation of a make up AMD GPU chip (complete with separate replacement video memory) is an insanely expensive thing to do just for a stockholder PR stunt.
Back to top
 
« Last Edit: 08/24/18 at 07:20:28 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #48 - 08/21/18 at 17:06:56
 

https://liliputing.com/2018/08/steam-for-linux-now-lets-you-play-some-windows...

We mentioned Gabe at Steam was  going to support a Wine derived methodology to run Windows games on Linux over at Steam.

This just went Beta and is available now right now for use on Steam Client games and on Steam Play Service games.

These are a subset of all games that are run off of Steam loaded and Steam controlled software and drivers, not using the raw AMD or NVIDIA drivers/software.

Two dozen AAA titles came out with the announcement, with other titles coming out as the software gets proven out by Steam users worldwide.

Gabe is also an early supporter of using Flatpacks to install Steam itself on Linux machines.   These too are available to use right now.

So, I have used Gabe's Flatpack tech to put Steam on my newest version of my same old Linux box and I have about half my Windows games are now available to run on my Linux install.   Not too shabby for something that only became known a few months ago.

The cracking sound you hear is Microsoft's gaming monopoly cracking like an over-stressed ice sheet ......

Windows Games on Linux ---- yes indeed, it smells like a real thing it does.


Another New Thing ........ Win 95 is now freely available now as an app.   Many older Windows games can run off of Win 95 as that was legacy supported by the Windows gaming industry until just 3-4 years ago.    Flatpacks jugging both are showing up as we speak.


NEWS as of August 28th --- Steam has converted over 1,000 Windows games now to run on Linux flatpacks.   Since Gabe knows what you have in your library, some of your Windows only games instantly rolled over to the "supported on Linux" category.




==================================================



Next real thing, 7nm - 5nm 3rd generation SoC samples are now shipping from both Huawei and Qualcomm as of Tuesday of this week.   These new production lines much much quicker than before using a direct to silicon EUV process.   No multiple multiple  multiple tank dunkings and ever finer masking steps are required any more.    

Production orders are being taken by both players to be bulk produced by TSMC as soon as Apple finishes up their production runs for this year or an additional one of the brand new (very newest type) ASML lines finishes installation at TSMC.  

https://www.youtube.com/watch?v=AKxJ3SQB0og      Watch this one, shorter and better

https://www.youtube.com/watch?v=jH6Urfqt_d4

Rumor has it that the very newest ASML lines can do either 5nm or 7nm, the program focusing and aiming the EUV electron laser beam being the only real difference as the EUV electron laser beam can be tightened up still further, potentially down to 3nm out in the future.

So the same ASLM EUV set up can put close access memory down right on the same substrate as the SoC proper ...... or it can actually blend the systems memory lay down into the SoC'o design itself for the very closest fastest access memory location and to make for the very best throughput speeds.  

So far Apple and Qualcomm are the only players in this new "memory inside SoC" tech wrinkle.    Because of the functional die shrink there is LOTS of free room on the SoC silicon to hold this new close access memory (or a bunch more cores, or a big AI subsection --- you name it, there is LOTS of room for it and the newest ASLM machines can certainly put it there).

Why Intel can't just go buy a couple of these new ASLM set ups is far far far beyond my limited understanding .......   excessive pride and 'not invented here" simply isn't enough justification to my limited "comprehension ability" to allow your company to go down the toilet swirl as Intel is currently doing.



===================================================



https://seekingalpha.com/article/4201847-intels-10nm-problems-implications-fa...

https://www.tomshardware.com/news/intel-cpu-10nm-earnings-amd,36967.html

https://www.fool.com/investing/2018/05/27/intel-could-pay-a-costly-price-for-...

Intel has screwed the pooch and these folks predicted what would follow.   At the time they did these predictions, Intel had only lost 1% of their market share.   These folks predicted 20-25% share loss inside the first year and they were correct so far as Intel has dipped 10% of that 25% already, with the pace of loss increasing each month which is sorta scary.

Intel's attempts at mitigation have SO FAR not been effective, except in jacking up the cost of Intel products with all those extra cores (ditto for the costly "other people's GPUs" that have been tagged on to Intel products recently).    Indeed, large security concerns are now popping up all over Intel's attempts at mitigating their market share losses,  BRAND NEW security concern items which will eventually cost poor old Intel far more than the Intel speed ups ever saved.
Back to top
 
« Last Edit: 10/03/18 at 02:21:14 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #49 - 08/26/18 at 18:15:00
 
 
https://pocketnow.com/arm-5nm-hercules-2020



Semiconductor design firm ARM has done something extremely out of character: it has announced a roadmap for its product timeline into 2020.

The chip designer ARM is claiming that it is confident it can keep up with the pressure of always-connected 5G network demands and can beat Moore’s Law by breaking above the double-performance barrier in two year’s time. Furthermore, at a time when Intel’s position in the chip market appears to be unstable with Apple and traditional PC makers experimenting with different silicon for different form factors, ARM may be able to cement itself as the IP source for chipmakers into the next lucrative phase of computing — Windows on ARM through Qualcomm is a big example of that.

It will begin this year with plans to bring the Cortex-A76 design, announced in May, down from its current second-generation 10nm fabrication to 7nm with commercial products launching by the end of this year. ARM then moves into codename mode: “Deimos” will take over for 2019 based on the 7nm process, then “Hercules” will take a refined 7nm and then bring on down to 5nm in 2020. The company told us that Hercules will bring a 250 percent performance jump from 2016’s Cortex-A73.

All designs will run on a thermal design of 5 watts or less versus the 15W on standard Intel laptop products — realistically, that means phones will have to run with big.LITTLE arrangements of a few powerful cores and several efficiency cores. And unlike Intel’s offerings, workloads will still be on single threads. That said, the cores will be marketed to work with phone and tablet form factors. The watermarks that ARM expects us to be talking about in the next couple of years are streaming 8K video and high-quality mixed reality content.

Samsung has also announced its roadmap for silicon this year and it seems to be taking chips down to 4nm with new lithography technology and new transistor designs. Media were given an opportunity to ask questions about whether ARM’s fabricator partners were ready to take on challenges and costs associated with smaller dies. The company gave us assurances that it was lock step with said partners, but has yet to provide specifics.

Update: We asked a couple of questions relating to how ARM’s fabrication partners may be able to scale to new transistor technologies as die sizes shrink. Cost is a big issue in making these transitions and those costs eventually get passed along to consumers.
Here’s the response from ARM:

Arm IP is designed to be used on multiple generations of process nodes, and the IP itself is configurable to stay within area and power budgets.

For example, the Cortex-A76 is targeted at 10nm/7nm, and ‘Hercules’ at 7nm/5nm. Cortex-A76 also supports DynamIQ technology, which allows partners to scale the use from 1+3, 2+6 or 4+4 configurations. These configurations are examples used to serve entry, mid-range, and upper premium markets where process nodes vary.

This flexibility of process, and configuration options of the IP at implementation time, and DynamIQ technology, allow Arm partners to balance their needs across all the tiers in economically viable process nodes.


What does it all mean?

The new ASLM process cells are working well and Intel is ripe and has already been primed by ARM to lose additional market share (increases beyond the current 21%) before Christmas time this year (2018).

AMD and their two Chinese co-contractors have already gotten their carving knives out and have already started slicing on some tasty slabs of ham off the Intel Christmas ham.   Ditto for Huawei and the other Chinese phone makers --- all of them are now making laptops and Chromebooks now-a-days and all of them are putting forth AI empowered designs with larger and larger AI blocks that are based on phone type technology.  

Intel is rapidly losing market share in China and is trying to obfuscate how bad the losses really already are by any trick they can come up with.

Please remember, those new ASLM production lines installed at TSMC can run 7nm, 5nm, (and by Samsung's latest real world sample releases all the way down to 4nm) RIGHT NOW as we speak.

Shocked


Back to top
 
« Last Edit: 08/30/18 at 06:49:55 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Eegore
Serious Thumper
*****
Offline

SuzukiSavage.com
Rocks!

Posts: 7990

Re: 2018 -- new Intel failures & issues
Reply #50 - 08/27/18 at 05:56:49
 

 Where did you get the 20-25% market share losses numbers?
Back to top
 
 
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #51 - 08/27/18 at 15:57:58
 
 
"Where did you get the 20-25% market share losses numbers?"

read up in page #2 of this thread for the first mention of these compiled figures

The first predictors assembled these predicted 20-25% figures based on what AMD and ARM were doing separately.   So far only 15-20% market share loss has reliably impartially hit "measurably really" actually occurred in USA consumer PC space, to this could be add an additional "up to 10% additional lost market share" by year's end if you take in China and the AMD/Hygon servers and what Hygon is doing to Intel rack space market share over in China's server space.    

However, the rate of shift between "theoretical prediction" and reality is increasing at an amazing rate month on month.

AMD is also partnering with Zhongshan Subor and other Chinese console and all-in-one makers, literally cooking them up specially designed "just for them" AMD SoC versions with state of the art on chip graphics that have everything they want on the same inexpensive piece of silicon wafer.

Actually, the main unexpected "official support" for this 20-25% figure comes from the mouth of the ex-CEO of Intel, Brian Matthew Krzanich. who punted the 20-25% number himself publicly a week before Intel ran him off from his CEO job.

Issue with all of these numbers is that Intel isn't going to tell anybody officially about their market share loss numbers --- as a matter of fact "according to Intel" they are still growing instead of shrinking.

https://www.cnbc.com/2018/07/27/reuters-america-intel-shares-slip-on-worries-...

https://wccftech.com/intel-coffee-lake-amd-ryzen-cpu-market-share-july-2018/

Here is the hard facts, Intel started out this year with a virtual monopoly in PC space and in gaming space combined with a virtual monopoly in server rack space world-wide.  Since then AMD has risen up to be selling in just as many newly produced AMD chipsets as Intel is selling in newly produced Intel chipsets, with both of these numbers only counting only current production sell thru of newly produced units.   These are worldwide numbers, not local American marketplace numbers.  Intel is still relatively strong in the USA market, but Intel is not as strong as it was last year by any one's guesstimate.

(Intel is known to have bulging warehouses full of old stuff that isn't moving at all because Intel hasn't discounted it enough to make it appealing to end users and PC builders).   But Intel sure has replaced all the old part numbers with brand new part numbers (generally with a pair of extra CPU cores mounted on the topmost chips as an extra cost bonus).

https://wccftech.com/intel-coffee-lake-amd-ryzen-cpu-market-share-july-2018/

Another way to look at it (a quick and dirty way) is that any reliable new AMD sales figures are by definition actually also Intel market share loss figures for the last 2 year period.   Start out with a solid monopoly and when your competitor actually picks up xx% then you just lost that exact same percentage in worldwide market share.

https://www.amd.com/en/press-releases/press-release-2018-07-25

“We had an outstanding second quarter with strong revenue growth, margin expansion and our highest quarterly net income in seven years,” said Dr. Lisa Su, AMD president and CEO. “Most importantly, we believe our long-term technology bets position us very well for the future. We are confident that with the continued execution of our product roadmaps, we are on an excellent trajectory to drive market share gains and profitable growth.”

Q2 2018 Results

Revenue was $1.76 billion, up 53 percent year-over-year and 7 percent quarter-over-quarter. The year-over-year increase was driven by higher revenue in both the Computing and Graphics and Enterprise, Embedded and Semi-Custom business segments. The sequential increase was driven by higher revenue in the Enterprise, Embedded and Semi-Custom segment.
Gross margin grew to 37 percent, up 3 percentage points year-over-year, driven by the ramp of new products. On a sequential basis, gross margin was up 1 percentage point primarily driven by a richer mix of revenue in the Enterprise, Embedded and Semi-Custom segment.
On a GAAP basis, operating income was $153 million compared to an operating loss of $1 million a year ago and operating income of $120 million in the prior quarter.
Net income was $116 million compared to a net loss of $42 million a year ago and net income of $81 million in the prior quarter. Diluted earnings per share was $0.11, compared to a loss per share of $0.04 a year ago and diluted earnings per share of $0.08 in the prior quarter.
On a non-GAAP1 basis, operating income was $186 million compared to operating income of $23 million a year ago and $152 million in the prior quarter.
Non-GAAP1 net income was $156 million compared to a net loss of $7 million a year ago and net income of $121 million in the prior quarter. Non-GAAP diluted earnings per share was $0.14, compared to a loss per share of $0.01 a year ago and diluted earnings per share of $0.11 in the prior quarter.
Cash, cash equivalents and marketable securities were $983 million at the end of the quarter.











===================================================



Remember, Intel can open up with their old warehouse stocks of the last 2 generations across all of Intel's processor ranges and Intel could generate a massive amount of short term sales once they get really serious about discounting all their old stuff enough to simply get it to move on out of those warehouses.

The market is expecting Intel to do exactly that about the same time as the ARM new product wave hits, with Intel suddenly willing to take these big losses to move a huge mass of old inventory while intentionally paying this discount financial bookkeeping loss adjustment simply to stifle ARM's sell in wave.    The trick here is that the better half of the old Intel stuff will compete favorably with the new ARM stuff once Intel cuts their selling price in half or less ,,,,, then discounts it again for volume to builders for a fast sale.

These will be the months to go buy you a new PC, as you will get some real deals then, some serious serious good deals.

Roll Eyes


Back to top
 
« Last Edit: 10/03/18 at 02:23:44 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #52 - 08/29/18 at 15:59:39
 

Passmark has posted their 2nd quarter market share numbers for Intel and AMD.   Intel has indeed lost 21% market share to AMD, with the numbers cycling around the 80% / 20% split for the first half of the year.






LATEST NEWS FROM AMD

AMD and Global Foundries have parted efforts on trying to use Global Foundries to do anything smaller than 12nm.   Global has always followed Intel's technical lead and following that Intel lead has taken Global deep into a technical disaster that they cannot get out of (just as Intel is stuck and cannot get out their disastrous mess either).

Just like Intel, Global is now stuck in a huge huge repeated time after time series of 10nm failures in their repeated attempts to get to 10nm and below using multiple multiple masks and old school immersion lithography.    With the AMD abandonment, the Saudi princes funding Global have pulled the money plug on any future 10nm attempts from Global Foundries.

This translates into a serious effort to replace the Saudi dollars by immediate cost cutting efforts -- Global Foundry just cold fired 450 employees to help balance their internal costs better.

AMD is rolling all of AMD's development effort and $$$ over to the new ASLM / TSMC 7nm to 5nm processes that are currently making 7nm processors at rate at TSMC, with good evidence already existing that 5nm and 3nm are already able to work pretty well and are going to be able to move into place within  the next two years.

As AMD rolls completely over to TSMC, they will reap additional cost advantages and throughput advantages over Intel that should increase their market share further past the 20% point.

Also note there may soon be a third ARM PROCESSOR line showing up on these charts as ARM/Qualcomm just posted the first test run results on their new Snapdragon 1000 laptop chipset --- first results are up in the mid Core i5 range.  Samples of this SoC are now being provided to Chromebook and Laptop builders with Christmas unit production still theoretically possible if Qualcomm kept the same socket and pin connection pattern so that the new chip could drop right into existing ARM based Win 10 laptop motherboards.  

ARM has followed this up with the announcement of TWO BRAND NEW Big and Little  core generations and new "stack-able" AI modules and new graphics booster modules to go along with it.   These new items are STATED directly from ARM to be Intel Core i5 and Core i7 competitors while running at less than half the Intel required wattage.

First glimmers from the Snapdragon 1000 test units show 27 hour battery life at mid to lower Core i5 performance levels.   Cost is unknown, but a product built off this at a reasonable price could sell well enough to get Microsoft to tune an OS variant specifically for it.

These are new market share losses that are pending for Intel.     The larger ones that will really count the most first off will be the new Chinese foundries that the China Gov. just finished, making AMD derived Chinese rack server chipsets (especially once they kick into making external commercial production -- all that capacity is currently taken up right now making lots & lots of chips for some really big top secret military implementations).

Note that the Chinese Military was a very large Intel customer that was certainly counted in Intel's market share numbers in the chart up above but has now rolled over to using domestic produced AMD Epic style server chipsets.

Roll Eyes


Back to top
 
« Last Edit: 09/01/18 at 07:55:03 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Dibble
YaBB Newbies
*
Offline

SuzukiSavage.com
Rocks!

Posts: 1

Re: 2018 -- new Intel failures & issues
Reply #53 - 08/31/18 at 07:02:33
 
Eegore wrote on 08/27/18 at 05:56:49:
 Where did you discover Skinny Views and get the 20-25% market share losses numbers?


Future looks bright for AMD if that's correct. And for us as end users too. You never know with these things though. Things can change fast.
Back to top
 
« Last Edit: 06/17/22 at 05:19:05 by Dibble »  
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #54 - 08/31/18 at 21:04:00
 

https://liliputing.com/2018/08/huawei-introduces-kirin-980-7nm-octa-core-chip...



READ THE POWERPOINT SLIDE CAREFULLY -- it is announcing a chunk of the technological leadership changeover stuff that we have been tracking for quite a bit now.

If you totally believe all of Huawei's PR stuff then Apple is NOT the phone tech leader any more,  Qualcomm is not the phone processor nor the cell tower radio tech leader any more,   Intel isn't squat for nothing and Google is over there to the sidelines now instead of out on the center of the playing field.

Now here is what is really shocking --- all of these items are actually partially to mostly true at this particular moment in time over in the China / India / Indonesia zone.

And yes, until Apple releases the next wave of two new iPhones Apple has indeed lost the numerical #1 crown to Huawei, but that happens each year anyway (Samsung generally picks up the #1 volume crown for 3-4 months out of each year).

What is the real truth?   Huawei is a solid #3 contender who has bumped off whomever is  #2 several times in the last 2 years, taking #2 (or now somewhat steadily -- sometimes as high as the #1 spot) based on volume of sales.  Huawei sells a lot of phones in China and in India where their quirky cell tower broadband system is the "regional standard" as Huawei is the largest cell tower tech provider over there in Asia.

Huawei has a very solid tech base and is now COMPLETELY CURRENT with ARM's A76 generation and 7nm TSMC production design.   They have their own odd form of 4G LTE and their own odd form of 5G that are formidable if backed up with their own home grown cellular tech installed inside the cell towers (they are the tower standard in the India/China marketplace, btw).  This new tech supposedly does not violate Qualcomm base patents or else it actually holds a license for the very few old Qualcomm tech items that still have to be used ......

What Huawei has accomplished is to take out Samsung in raw volume.   This is really all Huawei has accomplished volume-wise right now.   This is also what Huawei needed to do, and they planned well and they did it.  

You can ditto Qualcomm also if you want to, as Qualcomm isn't as important now at all to the Chinese phone makers as they are buying their radio components from each other now instead of from Qualcomm.   Neither Qualcomm CPU nor Qualcomm radio patents are "base natural" to the Chinese cell tower systems any longer.   Improved alternate standard have been put forward and accepted, so there is an entire chunk of the world that doesn't pay royalties to Qualcomm any longer.

And this is all really a simple reflection of the China/India/Indonesia market simply growing so so much faster than USA/Europe/Russia.
Back to top
 
« Last Edit: 09/05/18 at 02:48:21 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #55 - 08/31/18 at 21:53:47
 

Prediction Time:

Just as Global Foundry has moved their 10nm multi-mask immersion lithography process lines out to the curb as "unfixable" so should Intel.   Or else simply continue to use those multi-mask immersion lithography lines at 14nm-12nm like you have been doing and stop wasting time and money trying for 10nm and below on equipment that jest flat cannot do it ......

Give it up, the technology you bought back then and cannot get to work right now STILL isn't going to be cost competitive at all right now when thrown up against the lower production costs and far higher yields of the brand new ASML 7nm-5nm-3nm "direct burn" lithography processes, new stuff that can now be bought simply for lots & lots of money and a 2 year waiting list wait.    Everybody in Asia took a number already, so you are now delayed at least another generation level even if you had the down payment coins to take a number.

You should both go take a number to get a year 2020 brand new 5nm-3nm equipment installation appointment with ASML ....   or else simply admit you are planning to exit the industry when your old stuff loses all its 14nm to 12nm demand.

Or do like IBM before them, you can just stick with only providing the older tech chipsets until the demand drops away to zero, then simply fade away to "out of business" status.



===================================================



There are a lot of players that by default are actually now planning to enter the ASML direct burn lithography levels at 5nm-3nm, simply because they didn't act soon enough (ot lacked the ready coins to buy in quickly) to take a "make my line" down payment ticket in time to get a 7nm production cell -- they didn't act early enough to actually be there at the 7nm time span.    

The 7nm direct burn generation may wind up being a short duration lithography stage anyway since the both the 5nm and 4nm direct burn levels have already proven out at Samsung already.    Sammy will only sink money on a long term lithography level, not a short term level that has already passed everybody else by.

TSMC and Qualcomm have both done SoCs at 7nm and 5nm already.   TSMC is selling 5nm memory in large production volumes already.   Of course memory is a lot easier than CPUs and GPUs, which is why they debug lithography changeovers on memory production first --- you get to run the piss out of it at high yields and make a good profit while doing it.

Apple has already run its A-12 generation of iPhone chips at 7nm already and is gonna want to drop down a lithography level in 2019 to take advantage of the very newest ARM designs.    

In 2019-2020 Apple will fund TSMC to go work out all the rest of the ugly out of a 5nm "could maybe even cover a laptop" 5nm direct burn A-13 production level for their new Apple A-13 SoCs.  

Samsung has already bought a dozen of their ASML 5nm capable production equipment and has already run samples down as low as 4nm at this point in time.
  This is direct burn lithography in action -- faster levels of progress, faster production times, less cost in the finished product.

On the flip side, folks see Global Foundries paying the death price for betting on the wrong technology, following right along behind Intel who showed them the pathway to failure in the first place.
Back to top
 
« Last Edit: 09/09/18 at 09:06:08 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #56 - 09/07/18 at 17:01:46
 

This one is odd ...... requires a little thought on your part as AMD is right now selling in new stuff in as fast as Intel is, so that half the current supply amount is not just coming from Intel right now.  Intel is failing to supply their less than 50% of total PC demand at this point in time.

So whatever is happening at Intel it is effectively at a 50+% level .....

https://www.fool.com/investing/2018/06/.../intel-cant-crank-chips-out-fast-en...

https://www.pcgamesn.com/intel-14nm-supply-shortfall

https://www.pcgamesn.com/intel-14nm-h310-production


AMD is now getting even more new orders because Intel cannot supply what the customers want when they want it.

Throughput and yield issues at Intel 14+++ are simply killing Intel, and they can't move down to 10nm because it doesn't work for them AT ALL.

Meanwhile, Samsung has just rebuilt and tuned up their old immersion & mask equipment to do 8nm ARM designs with "standard multi-mask" right now (taking orders for it right now).

TSMC has picked up all of Intel's ASML build slots that Intel dropped the ball on, converted them to 5nm orders and now has at least 20 new ASML machines in process to be delivered inside 1 year.

Samsung just took delivery on 12 new 5nm TSMC machines and converted a bunch of their first gen 10nm immersion lines to 8nm.

Intel, ZERO new EUV machines in process beyond the one (1) that went to the R&D building.


Roll Eyes


Intel can BS their way out of many things (and they do, ongoing) but they can't avoid the BS tax when they miss shipping real orders that they agreed to accept enough times in a row and FLAT LOSE their customers for lack of performance.

Intel's latest strategy (raise the price significantly to lower the demand to what they can currently supply) is really a stone stinkin' loser of a strategy.    Short term it makes you an extra dollar while your customer struggles to find another supplier that can supply the goods that he really needs at a price that can be afforded.

In other words, it tells affected customers it is time to move to somebody else and it just encourages Intel's newest competition to come on out even harder and faster.  The price gouging just tells Intel's customer base it really really is time to move on over to  a second line of product built using an AMD chipset.

One questions if Intel actually does intend to depart the foundry business inside the next 2 years and what we are seeing now are the Intel exit strategies rolling out into public view?    

Brian Krzanich touted this "exit the Consumer Foundry business strategy" as a plan several times in the past two years and actually tried to shut down new consumer chip development once for almost a full year before being forced by his board to restart things when his Automotive and IoT "take over the world" initiatives tanked badly.



===================================================



(three days later)

Confirmation of Intel's reported inability to make the chips that they had already committed to make comes from Hewlett Packard Enterprise.
just read it ..... sorta blunt and easy to understand

https://www.semiaccurate.com/2018/09/07/intel-cant-supply-14nm-xeons-hpe-dire...

In this report Intel is sluffing off on the higher complexity chipsets and using what 14nm build capability they still have to do their simpler consumer CPUs.



===================================================



(four days later)

AMD is busy kicking Intel's arse all over the place cost-wise using chips that are built to AMD's specs at Global Foundry and at TSMC.    TSMC is taking more and more of this volume, especially at 10nm and below.

Intel has finally paid attention to this AMD/TSMC trend and Intel has "reportedly" just placed a set of chip orders with TSMC for some of their standard bread and butter chipsets to be built at 14nm, Intel intends just to cover their production losses due to low process yields on the newer Intel processors (the last ones with the extra two cores tacked on to existing Intel designs).

Run this up the flagpole and see if the troops salute it ---- Intel intends to use more and more TSMC contract production runs, especially using TSMC for the ever lower lithography levels that Intel can't get to work right at their own factory.


Shocked      ...... makes you sorta wonder, doesn't it ......
Back to top
 
« Last Edit: 10/03/18 at 02:33:22 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #57 - 09/15/18 at 22:17:18
 

https://www.semiaccurate.com/2018/08/16/deimos-and-hercules-appear-on-the-arm...

                                                                                             
                                                                                               A-76      Deimos      Hercules



The future according to ARM

This slide comes with a lot of caveats. First the workload is strictly single threaded but the power is at the SoC level. Intel is at 15W, ARM is at 5W. The 2 core Intel machine could hyperthread four of those threads at 15W, albeit with some loss of peak 1T performance while the ARM SoC could probably do the same at a much lower 5W as it is an unnamed compute form factor SoC which is unlikely to have less than four big cores. That means it can likely sustain the 5W TDP with ease for extended periods of time and the same holds true for the Intel CPU at 15 TDP.  Whatever those losses may be on either side the 3x energy use delta pretty much hands the performance and battery life win to ARM.

*(Note: ARM was asked why they didn’t include the newer 8xxx Intel CPUs in the graph. The answer was pretty simple, all of the testing was done with purchased devices and ARM couldn’t find an 8xxx Intel CPU on the market to test with when they ran the numbers.)

The intent of this reveal is to show that ARM is capable of not just beating Intel at their strong suit, single threaded performance, but doing so at significantly lower power levels. This may seem like a stretch but if you look at current ARM architectural license cores like the Apple A11 and the Cavium/Marvell Thunder X2/Vulcan, it is pretty clear that ARM cores can hit pretty high performance levels.

More importantly the vanilla ARM cores are designed for a wide frequency range so they leave some performance on the table to gain that flexibility. If you take an ARM core and design it for a small number of workloads and frequency ranges, you can get a lot more performance out of the SoC, and we mean a lot. The classic example of this is the Apple A11 which currently trounces the best Intel has at performance per Watt and in many cases raw performance. The numbers SemiAccurate has seen for the A12 show it gains about 50% more single threaded performance without changing energy use. Some of this is process related but most of it is architectural. You will see in a few weeks and you will be impressed.

So in the end Deimos and Hercules are the next two from ARM and the gains are pretty impressive. The current A76 is a big step over the still kind of current A75. Hercules will be just as big a step over A76 but Deimos 7nm+++ slots itself in-between.

In any case ARM should be right in the mix for thin and light laptops but with substantial feature improvements over Intel like 5G, always on, instant on, and more. With the added performance off the new architectures, raw CPU performance will be right up there with the best.



OK, when this came out originally everyone was all amazed and everyone said "Hey, ARM NEVER releases information on anything until it is acturally being built by a vendor and they say something publicaly."
I said that myself too, and I surely wondered what was up.    

What's up is becoming a lot clearer now.

Apple is what's up.   Samsung is what's up.   Qualcomm is what's up.   TSMC and Samsung 7nm+++ is what's up.  Samsung/Qualcomm/TSMC 5nm is what's up.   All this stuff got announced by the players at the last German PC show.  

Apple is now working on doing trial production runs at TSMC's EUV 7nm+++ process at this time on a Deimos based (very very tweeked) SoC called the A13.


Look at the big chart with the explanations added to it.   It isn't just some proposed BS vaporware like Intel would do -- these are trial run SoCs already really exist and there are proposed 2019 and 2020 products are being designed and built around these sample SoCs from these sample runs at this time.  

So ARM actually did what it always does, only after their customer/vendor/builders make their initial first real moves will ARM announce their next generation publicly.  

The last two dots on the chart are Deimos and Hercules from ARM.

Deimos is a 7nm+++ half generation EUV process move on a 5nm intended Hercules full generation move, with the SoC design and the tech being used is essentially the same.  

So the ARM full secrecy goes back into place now.  So you can now expect a new ARM generation will be announced to replace Hercules right about the same time as 5nm Hercules is being built out in large runs of products and these first trial lots of the new not named yet ARM stuff at 3nm gate all around are built.

Apple's A13 will be the first tweeked member of this current fully announced Deimos generation.  

Apple's A14 will likely be the first 5nm ARM Hercules design coming out inside the next two years.

Huawei and Apple are neck & neck right now on their 7nm++ "firsts" but they are both being leapfrogged right now by the others going down to 5nm in the background right now as we speak.


Roll Eyes            7nm+++  (alias Deimos) is firmly in the pipeline for next year for full production in phones.   Current Apple production efforts says it is real.


So it is all becoming real now ......  Intel actually went underwater on per core performance at that first A-76 dot on the chart and the lines will officially cross there at that first A-76 dot when the next version of this chart comes out.



===================================================



*    A point to be made here about Intel.   Intel makes outlandish claims for planned new stuff and touts the the new stuff like crazy at least a year before you can buy any of it.   This does not mean it is real or is ever going to be real.   Intel marketing lies a lot, in other words.

ARM won't even talk about their stuff until a vendor is actually producing it.

This means the Intel boys are posting bogus performance claims like a year in advance of Intel's "reality only you can buy it right now stuff".    Yes, baffling their consumer customer base with verbal trickery is certainly a part of Intel's little black bag of tricks.

If ARM is firmly ahead in this sort of race, it means ARM is really WAY WAY out in front by the time Intel finally gets itself into the starting blocks for real ..... if they ever get there that  is.  

ARM is staying another whole 1-2 generations ahead in general now .......

It takes ARM less time to roll an entire processor generation than it takes Intel to get around to actually making the stuff they are bragging about right now.
Back to top
 
« Last Edit: 09/22/18 at 03:19:02 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #58 - 09/17/18 at 11:09:46
 

Prediction Time


This is where I make predictions and we all see how many come true.

Intel is going to fail noticeably on production, shortfalling their own already pre-sold chip production numbers by varying increasing amounts for the next year, increasing the shortfall amounts to over 20% by early next year.  

This is with using the "stopgap" TSMC contract production of Intel's CPU designs, with Intel trying to bridge this gap with non-Intel production of Intel designs.

Intel, in direct contradiction of all their PR from the last 3 years will bring out a line of 22nm new products.  Why?  Intel had very good yields at 22nm, something they do not have at 14nm as we are learning more and more.   Plus, Intel has under utilized 22nm lines and has finished proven designs to run on them.

Intel will also re-stencil some old warehouse stocks of 22nm products that have been slowly moldering on the shelves ......  buyer beware !!!

AMD will make hay during this period of time, with all machine building vendors coming out with AMD processor based units to meet their own unit production demand figures.  

Intel's agreements with the larger builders to not use anyone else's chips will falter and break because Intel cannot meet the builders production requirements.
Intel's ongoing production failures will break the agreements, not the builders.

ARM based Deimos and Hercules processors will be created during this period of time and any of these units that get built and sold will be a net market share loss for Intel and for AMD.

Get used to seeing TSMC make Intel's most modern chip designs for them.

Watch out very carefully to see if what you are thinking about buying actually has an old tech 22nm Intel processor in it.
 

Before long Intel will want TSMC to build some 7nm and 5nm and 3nm Intel designs in ever larger batches because Intel simply can't do it themselves.

Roll Eyes    

Watch Intel begin walking away from the cutting edge of the foundry business just like Brian Krzanich told us Intel was planning to do.   Krzanich said this the in the weeks right before Intel fired his arse.



===================================================



https://www.macrumors.com/2018/07/05/apple-informs-intel-5g-modems-not-needed...

So, why is Intel "phasing down" on their smaller chip building technology?    Intel is currently 2 generations behind the industry lithography level and the voltage requirements of their various cellular modem technologies (22nm at 5 volts) is getting out of step with the current 7nm products generation that only provide 3.6 volts natural inside the device any more.

Because of this, Intel is losing customers.  They just lost Apple as a cellular modem customer for 2019-2020 due to Apple designing and building their own modem / base band set that fully integrates into the Apple SoC design.

Products that make the jump to 7nm and below can't readily use old higher voltage style components.   As screens go down to lower and lower voltages this divide will become bigger and harder to bridge.

So, both Qualcomm and Intel are losing cellular modem customers due to self designed SoC level "integrated implementations for lowered costs".   Both companies are seeing more of their larger customers roll their own products in preference to a "built for everybody" Qualcomm or Intel solution.  

Qualcomm is seeking to move over into Intel's PC processor markets while Intel is just specializing in shortfalling their supply levels to their existing customer base and in jacking up their pricing too too much right now.  

Intel is just encouraging the new interlopers with their big synthetically jacked up profit margins, in other words.
Back to top
 
« Last Edit: 09/22/18 at 03:22:37 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Oldfeller--FSO
Serious Thumper
ModSquad
*****
Offline

Hobby is now
"concentrated
neuropany"

Posts: 12636
Fayetteville, NC
Gender: male
Re: 2018 -- new Intel failures & issues
Reply #59 - 09/21/18 at 05:07:23
 

https://www.tomshardware.com/news/intel-14nm-shortage-h310c,37819.html



Please note the signs that old identification stenciling has been removed from this chipset

We've confirmed through multiple sources that Intel is fabbing its new H310C chipset on its 22nm process. That means the chip-making giant has taken a step back to an older process for the H310 chipset as it struggles with its ongoing shortage of 14nm processors. Contrary to recent reports, our sources confirmed Intel manufactures these chips and not TSMC (which has been reported in recent weeks), though that could be subject to change in the future.

The shift in Intel's strategy comes as the company struggles with the fallout from its chronically delayed 10nm process. Now the company is dealing with an increasingly loud chorus of reports that Intel's 14nm shortage is now impacting its server, desktop and mobile chips.

The worrying lack of motherboards with the H310 chipset, which began back in March, served as the first sign of an impending shortage of Intel's 14nm silicon. In May, reports surfaced that Intel had suspended production of the chipset, and in July, the company finally acknowledged a much larger issue with 14nm production.

Intel typically produces chipsets on a larger node than its current-gen processors, but the delayed 10nm production has found both chipsets and chips on the same 14nm node, creating a manufacturing bottleneck as the company experiences record demand for 14nm processors.

Word of a new H310C chipset surfaced last month. Leaked images of the new H310C on mydrivers.com revealed that the new H310C, which measures 10 x 7mm, is much larger than the 14nm H310, which measures 8.5 x 6.5mm.



Intel is struggling badly to supply customer orders and is spraying black squid ink into the water to confuse their customers about what Intel is actually doing to them right now.

H310 is the "proper" Intel 14nm processor that actually got benchmarked at 14nm quite a while back.   This is what is being advertised for performance numbers ......

Intel has put out LOTS of rumors about a 10nm processor to confuse consumer purchasing customers and to get them to go ahead and buy "their newest stuff" that quite bluntly DOES NOT EXIST YET.    and won't exist until 2019-2020 at the earliest

There is also a variant supposedly on order from TSMC as a 14nm stopgap offload measure that is due to the Intel production shortfalls.   This is also called "H310" and it too is still very much not real at this point in time.

In reality, Intel has either recently cranked up their 22nm machinery again or else has found a bunch of similar 22nm chipsets sitting around in a warehouse somewhere because these unknown unidentified things are being soldered on to motherboards that were specially built (with a somewhat different solder connection pattern) to hold a much larger 22nm chipset.    

Although looking at the picture there are clear signs that the original chip identification was removed intentionally by solvent wiping.     This mystery chipset is now being sold as a H310C.

Intel apparently now misleads and lies to their consumer customer base rather freely of late due to production shortfalls ......

YOU DO NOT KNOW WHAT YOU ARE GETTING and Intel isn't going to tell you either.

Roll Eyes               Tongue               Undecided              Embarrassed             Angry
Back to top
 
« Last Edit: 09/25/18 at 02:29:16 by Oldfeller--FSO »  

Former Savage Owner
  IP Logged
Pages: 1 2 3 4 5 6 ... 31
Send Topic Print


« Home

 
« Home
SuzukiSavage.com
04/19/24 at 15:47:50



General CategoryThe Cafe › 2020 -- new Intel failures & successes


SuzukiSavage.com » Powered by YaBB 2.2!
YaBB © 2000-2007. All Rights Reserved.