The list of intel predictions has been so dead right on so many many times, and Amd dead dead wrong even more times, that I am dizzy.
I absolutely hate intel, but also absolutely must admire their accuracy and insight about chip technology.
iNTEL rejected the K3 (or K5?) Amd chip design way back before it was offered to Amd. iNTEL said it would not scale. Amd purchased it and Amd almost went belly up trying to get it to scale.
Amd listened to IBM about SOI wafers when intel rejected the notion based on many flaws. iNTEL was correct and made better chips w/o SOI. Amd bought it ook line and sinker and SOI never lived up to IBM's expections. SOI has cost Amd billions.
iNTEL said that fusion would NOT work like Amd claimed it could speed up cpu speed by as much as "hundreds of time". Here it is years later. No Amd fusion. In fact, intel has the first fusion of a sort that intel said WOULD work - a partial fusion of graphics and cpu. Amd spent $5B to acquire ATI to beat intel to the punch by YEARS. As it turns out again, Amd is late to the party and intel has a full shrink on Amd to boot.
Way back when Amd claimed first to offer 64 bit computing, intel said it would not be needed for 6-10 years. iNTEL was right and this many years later, 64 bits STILL is not needed. Amd plunged in and lost their butt.
Way back when Amd insisted on copper fabbing chips, intel said it was not necessary for several more shrinks. iNTEL was correct again. Amd fall way behind and never caught up.
The list is so long that one could write a book.
Now, Amd has no fabs and their's not enough money or brains to catch up to intel. Amd has resorted to funny mone con games and executive remuneration to keep alive.
Amd has not even ONE glimpse of a chance to actually EARN money again. Con money is infinitely available. Ask Wall Street.
Coming out with 64 bit processors was a who cares moment of technology history it ranks with 16 bit following 8 bit and 32 bit following 16 bit all the followers were inevitable and were not revolutionary tech. What AMD did that was revolutionary was to create a CPU design that allowed a short pipe-lined CPU to kick the behind of a long pipeline design. Intel did not think this was possible or they would not have gone in the netburst direction. So yes AMD was the first X86 company to figure out short pipeline designs are way more efficient.
Next revolutionary advantage AMD had over Intel was that AMD went with a one size fits all mentality when it came to producing chips for the Server and consumer desktop space. At the same time AMD64 was released for the consumer and server space Intel had a seperate product for all the markets it sold chips in. The cost in design and manufacturing for Intel doing so must have been very large. AMD's one size fits all product designed saved AMD a great deal of money and with the onset of C2D Intel copied AMD and it to designed a one size fits all core.
lastly the IMC and HTT bus was revolutionary for its time since it introduced very low latency when compared to Intel's FSB. This very low latency enabled very low frequency core designs coming from AMD to best high frequency Netburst Cores that used the FSB.
So it is a fact that Intel has used ideas gained from AMD in order to get over the competition disaster that was the Netburst era.
It was not the core design per say that has given Intel the competitive lead, it is more the one size fits all mentality that Intel learned from AMD that has Intel prospering today. No more design units of Intel going in seperate directions.
"iNTEL rejected the K3 (or K5?) Amd chip design way back before it was offered to Amd. iNTEL said it would not scale. Amd purchased it and Amd almost went belly up trying to get it to scale."
This shows how much you know. AMD did NOT purchase the K3. There never was a K3. The K - which stood for Kryptonite to kill Superman (Intel) - nomenclature started with the K5. The K5 was not purchased either. The K5 was AMD's first completely home grown design where AMD attempted to create its own X86 microprocessor with it's own microcode so it was truly independent of Intel. I'll bet you can't even explain what scaling is because you're a fraud.
"Amd fall way behind and never caught up."
AMD did NOT fall way behind. AMD has always been a step behind Intel in the fab generations, always.
"The list is so long that one could write a book."
And one would be stupid to read your book because you foam at the mouth and barely have a clue as to what you're talking about.
<<Way back when Amd claimed first to offer 64 bit computing, intel said it would not be needed for 6-10 years. iNTEL was right and this many years later, 64 bits STILL is not needed. Amd plunged in and lost their butt.>>
1. Servers aren't 64bit? Server prices are down because servers can bootstrap off desktop 64bit
2. If 64 bit desktop isn't needed, why did Intel copy AMD's 32/64 bit architecture only a few years after AMD, and not 6-10 years?
3. If 64 bit desktop isn't needed why does Windows XP, Vista, and Windows 7 offer 64 bit desktop operating systems? Microsoft said Vista would be the last 32 bit operating system.
4. Llano true fusion is only 5 months away. Why do you think Nvidia stock is in the tank?
5. Why is Intel stock in the dumps unable to get over 21?
6. Why are so many companies going fabless?
Why, Ace, why?
Intergrated ( irritated ) graphics.. lmfao
Addd a nice ATI card to your sandy triage cpu that supports DX11
Keep blowing cash on your pc to support those intc margins suckers..
All you Atom pumpers gonna learn the hard way.
DX11 on your notenook coming
LOL, you keep bashing integrated graphics (I agree for now) and say add a nice ATI card BUT ALL ATI GPUs are tessellation cripple and are unable push W7D11 into virtual reality. Only Fermi has the transistor muscle that DX11 demands. Most of the new games worth playing push tessellation to the max leaving ATI looking like a slow mo slide show. ONLY the built for W7DX11 Fermi can show DX11 at it's full virtual reality glory turning ATI into a lame old looking sob story.
Integrated graphics have their place but not in a hard core gamers rig. Not unless you have a Fermi GTX480/470 or GTX460/SLI sitting on your mobo.
AMD brought NOTHING to the x86 table that was needed at the time, NOTHING! AMD kept on shooting itself in the foot wasting time and money trying to out run Intel to the next big tech idea. All these so called AMD chip features were nothing but marketing bullet points. They weren't needed and Intel knew that, AMD didn't and ended up wasting years and blowing the war chest on tech blunders like Hector's delusional ATI Fusion fantasy. Again again and again Intel beat AMD like a pinata at Paul Otellini birthday party.
Now it's Intel's mighty 32nm Fusion "Sandy Bridge" that will bury AMD/ATI's fusion delusion once and for all.
Real Men have Fabs, AMD has been castrated by Hector's Fablite. AMD's Fusion is at the mercy of TSMC's 40nm process. HUGE advantage to Intel, PLUS The key feature of Sandy Bridge Intel AVX (Advanced Vector Extensions) which, when used by software programmers, will increase performance in floating point, media, and processor intensive software, according to the Intel. Key features of Intel AVX include wider vectors, increasing from 128 bit to 256 bit wide, resulting in up to 2x peak FLOPs output; enhanced data rearrangement, resulting in allowing data to be pulled more efficiently, and three operand, non-destructive syntax for a range of benefits. Intel AVX can also increase energy efficiency beyond the increases brought by the micro-architectural innovations, and is backwards compatible to existing Intel processors.
It's not looking so good for AMD on the CPU front, and ATI has hit it's high water mark. NVDA will take back what it has lost, while INTC continues it's climb, both are so over sold, sorry I can't say the same for AMD.
May I add this new information about Intel processors at CES 2011
The introduction of the form of Sugar Bay will include a 32 nm Sandy Bridge processors – probably Quad core with integrated GPU – and a motherboard with P67 chipset, which has as usual by the P-chipset to the mainstream market as a destination.
A feature of the new platform will be the most recent Terms of Intel’s Hyper-Threading – or Simultaneous Multithreading – Advanced Turbo Boost technology and how these are changing, however, is still unknown. Later, as used by the generation dual core CPUs and will bring enabled “multiplier on the market.
Besides, Intel will launch the Waimea Bay platform, which will replace the previously used chipset. Waimea Bay consists of a Sandy Bridge processor and the Patsburg chipset, the processor is more precisely an E Sandy Bridge, which has an integrated memory controller and a PCI Express connection and will support DDR3 memory up to 2666 MHz.
For server customers will be developed in the future Bromolow platform, which also relies on the Sandy Bridge processors.
Now that desktop and server market are covered, may not be missing one: the mobile market. And even if this has already been with Clark Field and Arrandale longer served as well, the Huron River platform is already in the wings. Derived from the P67 desktop chipset, H65 and H67, and then follow QM67 and QS67. The series will bring PM6X never on the market because there will be no notebook processors without integrated graphics, this chipset is therefore unnecessary.
Already at week 50 of 2010, these platforms will be delivered, such products should appear mid-January. For 2012 we are pleased, however, already back on the Sandy Bridge Shrink, the 22 nanometer processors Ivy Bridge.