25 chips "that shook the world"
- Thread starter Simon F
- Start date
Nice to see the original Alpha and Opteron chips on the runner-up list, and Cell certainly deserves to be there.
No Pentium or PPro mention though?
No Pentium or PPro mention though?
Silent_Buddha
Legend
Well, the PPro never gained much traction and the Pentium was just an evolution of the x86 architechture. Had it not been for Intel wanting to differentiate from the competition (by patenting/trademarking Pentium) it would have been called the 586.
Regards,
SB
Regards,
SB
The PPro however was the start of the P6 architecture, on which the PII and PIII were also based.
It then evolved into the Pentium M, which moved on to Core, and then Core2 and Core i7.
In that sense PPro is the beginning of the bloodline of the current superscalar out-of-order x86 CPUs. I think it would have had more right to be mentioned than the Opteron, to be honest.
It then evolved into the Pentium M, which moved on to Core, and then Core2 and Core i7.
In that sense PPro is the beginning of the bloodline of the current superscalar out-of-order x86 CPUs. I think it would have had more right to be mentioned than the Opteron, to be honest.
Silent_Buddha
Legend
Yes but even the PPro was just an evolution of the same basic 32-bit architecture. The 286 and 386 should deserve more recognition IMO than the PPro. But even then those were still just logical evolutions of an existing chip.
I think the main reason that the Opteron received a runner-up award was the move from 32 bit to 64 bit. And more importantly forcing Intel to adopt a non-intel solution from it's main competion. That last bit probably earns more points towards something that "shook the world" than any Intel evolution of the 8086/8088. It basically shook the foundations of computing putting AMD as a superior choice to Intel for a few years. And in the process shook Intel out of it's complacency and back into a hungry competitor.
Regards,
SB
I think the main reason that the Opteron received a runner-up award was the move from 32 bit to 64 bit. And more importantly forcing Intel to adopt a non-intel solution from it's main competion. That last bit probably earns more points towards something that "shook the world" than any Intel evolution of the 8086/8088. It basically shook the foundations of computing putting AMD as a superior choice to Intel for a few years. And in the process shook Intel out of it's complacency and back into a hungry competitor.
Regards,
SB
Blazkowicz
Legend
Opteron is the chip that killed sparc, itanium, alpha etc. ; special hardware was replaced by PCs.
one major chip should be the Intel i810. Even though we hated it. That slow shit is the grandmother of netbooks and €40 mobos with everything integrated, I don't believe you can even find a µATX mobo without an IGP those days.
one major chip should be the Intel i810. Even though we hated it. That slow shit is the grandmother of netbooks and €40 mobos with everything integrated, I don't believe you can even find a µATX mobo without an IGP those days.
Well my point was that it wasn't.
It was a completely new way of handling x86 code, by first decoding it into micro-ops, then reordering it, and then executing it by a RISC-like backend.
I don't see it that way.
Alpha was already dead long before the Opteron arrived. In fact, AMD bought Alpha technology to use in its K7/K8 designs.
And Itanium isn't dead just yet. Intel is actually planning to go up a notch in development pace for the Itanium.
Sun also still has its own hardware, although Sun itself is now owned by Oracle, which may not be interested.
But CPUs like the Niagara are still very popular for webservers and databases. Just a year ago I worked at a university that used these Sun systems to run their massive website, and it performed great with the massive multithreading.
IBM's POWER is also still going strong.
Heck, it's Opteron that's about to be replaced by Intel's Nehalem. Ironically Nehalem is pretty much 'Opteron done right'.
Silent_Buddha
Legend
True, but none of that is anything that "shook the world." It was still an evolution of the 32-bit architecture. Perhaps a slightly more radical evolution, but still nothing that "shook the world."
The Itanium was far more revolutionary for comparison, but again nothing that "shook the world."
Regards,
SB
If you want to go down that line, then no x86 could ever possibly have shook the world at all.
Yet there are a few in these lists.
Hardly an accolade considering it's Opteron + 5 years
Actually I see that the other way.
AMD had 5+ years to get it right, and still failed in the end (obviously during these 5 years they have constantly been working on improving the design).
Intel got it right on the first try.
Pentium Pro was the first chip to bring SMP to x86 however, and the beginning of x86 challenging RISC in the high-end arena. The chip itself may not have achieved huge sales but it began a legacy of processors that did and opened up new markets, so I'd say it was a product that 'shook the world'.
IIRC Penitum is able to do SMP.
In terms of "shook the world" I think Pentium Pro deserves more than Opteron. Pentium Pro is the first x86 chip which was considered as a serious contender for server market. It broke the "common sense" back then when most people think RISC is the future and it's impossible for x86 CPU to catch up because they are too complex. Yet now server market is dominated by x86 CPUs. It's all started from Pentium Pro.
Extending x86 from 32 bits to 64 bits, on the other hand, IMHO is relatively "trivial." It's a direction everyone think it will happen eventually, the question is just when it'll happen. The specific implementation is just details, and x86-64 is not exactly a very elegant way for extending x86 to 64 bits.
In terms of "shook the world" I think Pentium Pro deserves more than Opteron. Pentium Pro is the first x86 chip which was considered as a serious contender for server market. It broke the "common sense" back then when most people think RISC is the future and it's impossible for x86 CPU to catch up because they are too complex. Yet now server market is dominated by x86 CPUs. It's all started from Pentium Pro.
Extending x86 from 32 bits to 64 bits, on the other hand, IMHO is relatively "trivial." It's a direction everyone think it will happen eventually, the question is just when it'll happen. The specific implementation is just details, and x86-64 is not exactly a very elegant way for extending x86 to 64 bits.
aaronspink
Veteran
no, not really. P6 really was the death bell for all the boutique architectures out there. It has performance parity with alpha and crushed everything else. It is the chip that put x86 into the server space.
I agree, especially since Intel already pulled the same trick from 16 to 32-bit with the move from 286 to 386. AMD simply repeated Intel's old trick. Intel instead didn't think it would be a very good trick, so they tried to come up with a new architecture together with HP.
Otherwise yes, it would have happened eventually, and Intel would probably have beaten AMD to it (their Itanium predates AMD's first 64-bit chip by quite a few years, so Intel was clearly commited to going 64-bit before AMD did, they just chose a different path).
Quick Google indicates you are indeed correct, Pentium was SMP capable, Pentium Pro allowed quad or octal processor arrangements.
I guess it would be better to say Pentium Pro was the first to popularise SMP on x86 by offering (initially) superior integer performance than competing RISC processors combined with lower pricing.
Silent_Buddha
Legend
Actually AMD got it right on the first try. Scaling well past 4+ processor configurations. Intel however improved upon the design moreso in the intervening 5 years than AMD did.
Regards,
SB
Silent_Buddha
Legend
Not really because the PPro which it allowed Intel greater penetration into the Server market didn't exactly shake things up significantly at the time.
Opteron as an evolution of x86 happening in a void may not be all that earth shaking. AMD basically setting Intel on its heels and forcing Intel to follow AMDs lead (even if only for a couple years) most certainly shook things up. In many ways the Intel of todays owes much to that very backstep that the Opteron forced Intel to take.
Had that not happened, it's quite possible we'd still see Intel working on refining the netburst architechture and quite possibly there would be Core 2 duo or Nehalem as we know them today.
However, due to the impact and shake up that Opteron was able to do. Intel regained it's focus and hunger to remain on top. Thus seeing them fast track pentium M -> Core -> Core Duo -> Core 2 Duo. All the while accelerating the demise of netburst.
IMO - it was a far larger shake-up of the world of computing in general than the PPro.
Regards,
SB
It looked right because they were the only ones doing it.
But they couldn't really get an advantage for 1-2 processor systems. And even 5 years down the line they didn't manage to get it off the ground.
