Originally Posted by DaveLT
C'mon man. Nehalem is a whole new architecture. First of all it differs radically. No FSB which is what the Core architecture is based on (PenM + NetBurst FSB)
However IT'S WRONG TO SAY IT'S RECYCLED.
The Athlon64 was still a souped up Athlon XP despite having 64bit, no EV6 bus, an IMC, etc.
SB is still based off Nehalem despite actually being an APU. (The Nehalem chips with an iGPU had the IMC and iGPU on a separate, 45nm die to the actual 32nm CPU)
Llano is still Stars/K10 based despite being an APU while none of the other CPUs are, and having PCIe lanes on board.
Trinity/Richland are still Bulldozer based despite being APUs with onboard PCIe lanes.
If you read Nehalem architecture analysis it's pretty clear it's Intel further incorporating the good ideas from Netburst into the more efficient architecture and just extending it elsewhere at the same time, iirc even the Front-End and execution engines were unchanged from Penryn
. I suggest you read this.
Originally Posted by almighty15
I'm just going to ignore him, he clearly lacks basic understanding skills.....
8080: Improved into 8085 and Zilog Z80 while retaining same basic architecture, 8085 gets improved into the 16bit 8086/8088 (8088 using 8bit support chips on the motherboard) which are then improved into the 80186/80188. (They put more of the support chips on the CPU)
80386: Intel designed this, then the 286 as the 386 was too difficult to manufacture at the time. The base i386 architecture, also the first real 32bit CPU.
80486: Drastically improved, put the FPU (Used to be a separate *87 chip) on the CPU, added 8Kb cache (iirc) and also added the multipliers we know and love on our CPUs today. Also known as the i486 in gcc.
Pentium: Drastically improved 486, now it talks to the rest of the motherboard via a 64bit GTL+ bus, added pipelines and was the first super-scalar x86 CPU. Known as the i586 in gcc.
Pentium Pro: Intels first CPU to use decoding, in that it used another microarchitecture internally and just translated x86 to and forth to keep compatibility. Also known as P6 or the i686 architecture in gcc. This is what the Pentium II, (A Pentium Pro with external cache and a new process node) Pentium III (Pentium II with SSE..Then some massive improvements) and the Pentium M were based off of.
Netburst: Long pipelines, abandoned after the Pentium D.
Then as you can see in the link I posted before as well as your quote, Conroe was based off of the Pentium M's architecture (At least 20% of it was kept) which was then improved and shrunk into Penryn, Intel then improved that into Nehalem, shrunk it to Westmere, improved that to Sandy Bridge, shrunk that to Ivy Bridge, etc.
Exactly as I said, you can clearly see the evolution from Pentium Pro to Core i7...So whose lacking the basic understanding skills? I've provided proof of the Dothan to Conroe change and now of Penryn to Nehalem, (Look up) you've provided one quote that proves what I said. (That Conroe was based off of Dothan)
Let me provide that quote again: "As a result, it would be absolutely incorrect to claim that the prospective processors will be none other but adapted (for the new applications field) Pentium M."
It's an adapted Pentium M, which is an adapted Pentium III, which is an improved Pentium II, which is an improved/adapted Pentium Pro.