What makes Sandy Bridge so good?

yottabit

Golden Member
Jun 5, 2008
1,491
522
146
Rather than focus on the negative reception of the Bulldozer launch, I'd like to take a step back and acknowledge how amazing an architecture Sandy Bridge is. That being said, I wonder what makes it so amazing?

Thinking back to the evolution of the Core CPU line, I see it this way:

Conroe/Penryn:
Pro's: Massive IPC increase and still highly clockable compared to competition
Cons: You have to choose whether you want dual core or quad core performance (both from price, power consumption, and absolute performance in games)
Cons: Need massive amount of cache ($$) as a band-aid to off die memory controller

Nehalem/Lynnfield:
Pro's: Another big IPC jump, similary clocks to Core 2 but more overclocking headroom, moving memory controller on board was a good idea
Pro's: Finally an affordable quad core, and turbo mode that gives it single threaded performance of a dual

Sandybridge: Yet another big IPC jump, higher clocks than anything before, and much higher turbo boost speeds. While Lynnfield took advantage of turbo, Sandybridge really relies on it, esp. in mobile form. Finally practical to have a quad core laptop other than a desktop replacement.

Moving BCLK generator to CPU caused drama, but was likely because of the massive overclocking headroom Intel knew about. Unlocked CPUs make it so easy to overclock everyone on tech forums think they are 1337

-------------------------------
Feel free to let me know any other reasons you think I am missing!

I think if BD had launched back in line with Nehalem, it would've been a much more warm reception. Even if SB had brought it's IPC increase only and not the massive overclocking headroom. But SB is honestly the most fiercely competitive processor I have seen in a very long time.

I wonder things like: how did Intel get clocks to scale so high? Was that their intent or a side effect?

Why did Intel decide to make fatter cores rather than more cores? Was that solely because of hyper threading? Or better margins?

Exactly what changes to the microarchitecture are responsible for the IPC increase AND higher clockrate of SB? I feel like an architecture that brings both of these, on the same process node are very rare. I mean, 32nm Xeon quads and Lynnfield duals did not clock as well as their SB counterparts, and have less IPC. That's pretty amazing.
 

bunnyfubbles

Lifer
Sep 3, 2001
12,248
3
0
Its all perspective. Sandybridge really isn't that amazing, its just evolutionary from Nehalem, not revolutionary.

as far as clockrate scaling its pretty much right where it should be.

First generation Core 2s (ie 65nm) were all the rage because of their high IPC and ability to easily clock into the 3GHz range with the very rare few touching 4GHz. Second generation Core 2 (ie 45nm) made seeing 4GHz a very real possiblity. First generation Core i series (45nm) made 4GHz common place. Now we're on 2nd generation Core i (32nm), and we're now well into the 4GHz range and toying with 5GHz. Its not like its a situation where we're seeing something completely out of the blue as if we were suddenly at 6GHz even with improved IPC or something.
 

Olikan

Platinum Member
Sep 23, 2011
2,023
275
126
it's the ring bus for the l3 cache...

it as fast as a normal l2 cache, but much bigger
 

Meghan54

Lifer
Oct 18, 2009
11,684
5,222
136
What makes SB so good?

All of Intel's prior work and their expertise in fabs.

If AMD had had C2D laid out in front of them, AMD still could have made a dog's breakfast out of it, much like Bulldozer's turned out.
 
Mar 10, 2006
11,715
2,012
126
Sandy Bridge is pretty revolutionary. With only 80W, it gives 4 cores @ 3.4GHz with INSANE IPC. That's a lot more than just a die-shrink will get you -- significantly better performance than 130W i7-960 in nearly half the power.

And that's really the thing -- SNB is a core that is ultimately focused on performance/watt more than anything else, likely due to the mobile/ultra thin focus. And that's why on desktop when it's scaled up, it kicks so much ass -- since even in the mobile form factor it is insanely fast.
 

Maximilian

Lifer
Feb 8, 2004
12,604
15
81
Sandy bridge is good because it does everything nehalem did but better, its got ~10% better IPC and because of the 32nm process it clocks higher than nehalem.
 

StrangerGuy

Diamond Member
May 9, 2004
8,443
124
106
1. $200 for a 2500K is a steal
2. Mobos are cheap
3. Idle power is non-existant - 5W
4. Load power is jaw-dropping low
5. OC headroom is immense
6. Superb all around CPU performance, especially in games
7. HD3000 beats the shit out of AM2/3 IGPs
8. Quicksync
9. No hype surrounding it, no speculation due to official previews, it just simply delivered.
 

Idontcare

Elite Member
Oct 10, 1999
21,110
59
91
Sandy bridge is good because it does everything nehalem did but better, its got ~10% better IPC and because of the 32nm process it clocks higher than nehalem.

I think this can't be understated.

SB is "good" because it doesn't require the consumer to take two-steps back in many of today's apps just for the promise of taking one step forward on tomorrow's apps.

As to how Intel managed to take a step forward on today's apps with SB, while also enabling ISA extensions that take another step forward on tomorrow's apps, look no farther than what AMD did with the stars core in going from 45nm to 32nm. (~2%-7% IPC improvement) Its the cache, and everything that goes into branch prediction and so on.

Not too mention that while Intel was making judicious microarchitecture decisions for SB they were also being extremely aggressive in their process technology development. Were SB 2600K manufactured on GloFo's 32nm it wouldn't be hitting 4.5GHz OC's with 150W power consumption like it does thanks to Intel's 32nm.

Gate-first lacks the intrinsic stress engineering opportunities that gate-last integration has. There is no getting around this. But gate-last integration is more expensive to develop (time and money), so had GloFo pursued gate-last then we probably wouldn't see their 32nm until next year instead of this year. Its all about trade-offs.

But for sure AMD threw too much of the baby out with the bathwater in their bulldozer architecture, and it appears it all comes down to the cache. Which is good because that can be "fixed" in time more easily than say an imbalance between decoders an ALU's.
 

lOl_lol_lOl

Member
Oct 7, 2011
150
0
0
Intel's bipartisan focus on performance and efficiency, coupled with a realistic approach to market requirements. What amazes me is the applicability of the architecture across several market segments, from notebooks to servers. Not a primal fixation on "MOAR COARS".

"Few but Ripe" - Carl Frederich Gauss.

Whats more, Intel attack every possible price point from $60 SB Pentiums to $220 Quad cores to the $999 3960X. Everyone can enjoy SB for whatever they do.
 

Ika

Lifer
Mar 22, 2006
14,264
3
81
2500k hit the sweet spot in price+performance for a lot of enthusiasts. To top it off, almost every 2500k can hit 4.5ghz with a good air cooler and motherboard. That's crazy. Sandy Bridge is a value monster.
 

IntelUser2000

Elite Member
Oct 14, 2003
8,686
3,786
136
Moving BCLK generator to CPU caused drama,

There was an article about how Intel was regretting the decision for integrating the BCLK generator into the desktop SKUs. So on Ivy Bridge, they'll partially lift the restriction, hence the various bus options. Maybe we'll get it fully back on Haswell.

Integrating the clock generator and saving $5 is extremely helpful for Celeron and Pentium laptops costing $400-500, but not much on the desktop.
 

Nemesis 1

Lifer
Dec 30, 2006
11,366
2
0
We just respect winners, like we did the A64 vs. P4, X2 vs. Pentium D, etc. SB is an out and out winner like those were.

AMD64 against P4C AMD wasn't an out and out winner . Some lies never die . This is one of them . Anything multi threaded Intel won against the AMD64. X2 changed all that in a big way.
 

Arkaign

Lifer
Oct 27, 2006
20,736
1,377
126
AMD64 against P4C AMD wasn't an out and out winner . Some lies never die . This is one of them . Anything multi threaded Intel won against the AMD64. X2 changed all that in a big way.

I think you're right when comparing approximate equals in northwood vs. A64, eg; P4 3.2C vs. 3200+, but the thing is that Prescott kind of sucked (worse IPC for most apps, hot, didn't reach good clocks), and that once the A64 was overclocked a little bit, or both max stable air overclock, the A64s ran away with it. The 3400+, 3500+, 3700+, etc rolled out fairly quickly, leaving P4 Prescott a hot boiling mad mess.
 

Makaveli

Diamond Member
Feb 8, 2002
4,806
1,269
136
Sandybridge: Yet another big IPC jump, higher clocks than anything before, and much higher turbo boost speeds. While Lynnfield took advantage of turbo, Sandybridge really relies on it, esp. in mobile form. Finally practical to have a quad core laptop other than a desktop replacement.

I agree with the statement minus the first part do you really consider 10% a big jump?


Sandy Bridge is pretty revolutionary. With only 80W, it gives 4 cores @ 3.4GHz with INSANE IPC. That's a lot more than just a die-shrink will get you -- significantly better performance than 130W i7-960 in nearly half the power.

And that's really the thing -- SNB is a core that is ultimately focused on performance/watt more than anything else, likely due to the mobile/ultra thin focus. And that's why on desktop when it's scaled up, it kicks so much ass -- since even in the mobile form factor it is insanely fast.

Significantly better eh?

http://www.anandtech.com/bench/Product/99?vs=287

Anand didn't have a 960 in the list so that was the closet in clock speed.
 
Last edited:

Makaveli

Diamond Member
Feb 8, 2002
4,806
1,269
136
Its all perspective. Sandybridge really isn't that amazing, its just evolutionary from Nehalem, not revolutionary.

I agree with this most.

Anyone on a highend nehalem system its mostly a side grade. The drop in power and increased overclocking headroom is a huge plus for SB. But the performance improvement wouldn't be enough for me to do a whole new build. if I was on a Phenom 2 or core 2 it most certainly would.

With Lamedozer flopping i'm curious to see what intel plans on doing with the current SB midrange chips they don't really need to release the 2700k now and they can even increase prices do to lack of competition.

I don't think i've ever seen a product that was classified as midrange to offer so much performance SB 2500k/2600K really caught everyone off guard if you asked me.
 
Last edited:

Arkaign

Lifer
Oct 27, 2006
20,736
1,377
126
I agree with this most.

Anyone on a highend nehalem system its mostly a side grade. The drop in power and increased overclocking headroom is a huge plus for SB. But the performance improvement wouldn't be enough for me to do a whole new build. if I was on a Phenom 2 or core 2 it most certainly would.

With Lamedozer flopping i'm curious to see what intel plans on doing with the current SB midrange chips they don't really need to release the 2700k now and they can even increase prices do to lack of competition.

I don't think i've ever seen a product that was classified as midrange to offer so much performance SB 2500k/2600K really caught everyone off guard if you asked me.

Yeah I notice much less heat comparing my 2500k (usually cap it at 45x, but it ramps down when not loaded) to my old PhII that I ran between 3.5 and 3.8 depending on my mood. I bet the power draw is much less.
 

Makaveli

Diamond Member
Feb 8, 2002
4,806
1,269
136
Actually the 975 cost $999, and the 2600K is $315. The replacement was for the 870, and it performed 20-30% better.


Yes but i'm talking Nelahem vs SB not lynnfield since Anand doesn't have a 960 in the list I had to choose the 975 cause it was closet in default clock speed to the 2600k.

in canadian pricing the 2600k is $330 and the 960 is $310! so my point still stands its not a 20-30% improvement across the board at all, if that were true I would have dumped my system for a SB build.
 
Last edited:

Makaveli

Diamond Member
Feb 8, 2002
4,806
1,269
136
Yeah I notice much less heat comparing my 2500k (usually cap it at 45x, but it ramps down when not loaded) to my old PhII that I ran between 3.5 and 3.8 depending on my mood. I bet the power draw is much less.

I think the power draw is probably half of what you were pulling on the PHII.
 
sale-70-410-exam    | Exam-200-125-pdf    | we-sale-70-410-exam    | hot-sale-70-410-exam    | Latest-exam-700-603-Dumps    | Dumps-98-363-exams-date    | Certs-200-125-date    | Dumps-300-075-exams-date    | hot-sale-book-C8010-726-book    | Hot-Sale-200-310-Exam    | Exam-Description-200-310-dumps?    | hot-sale-book-200-125-book    | Latest-Updated-300-209-Exam    | Dumps-210-260-exams-date    | Download-200-125-Exam-PDF    | Exam-Description-300-101-dumps    | Certs-300-101-date    | Hot-Sale-300-075-Exam    | Latest-exam-200-125-Dumps    | Exam-Description-200-125-dumps    | Latest-Updated-300-075-Exam    | hot-sale-book-210-260-book    | Dumps-200-901-exams-date    | Certs-200-901-date    | Latest-exam-1Z0-062-Dumps    | Hot-Sale-1Z0-062-Exam    | Certs-CSSLP-date    | 100%-Pass-70-383-Exams    | Latest-JN0-360-real-exam-questions    | 100%-Pass-4A0-100-Real-Exam-Questions    | Dumps-300-135-exams-date    | Passed-200-105-Tech-Exams    | Latest-Updated-200-310-Exam    | Download-300-070-Exam-PDF    | Hot-Sale-JN0-360-Exam    | 100%-Pass-JN0-360-Exams    | 100%-Pass-JN0-360-Real-Exam-Questions    | Dumps-JN0-360-exams-date    | Exam-Description-1Z0-876-dumps    | Latest-exam-1Z0-876-Dumps    | Dumps-HPE0-Y53-exams-date    | 2017-Latest-HPE0-Y53-Exam    | 100%-Pass-HPE0-Y53-Real-Exam-Questions    | Pass-4A0-100-Exam    | Latest-4A0-100-Questions    | Dumps-98-365-exams-date    | 2017-Latest-98-365-Exam    | 100%-Pass-VCS-254-Exams    | 2017-Latest-VCS-273-Exam    | Dumps-200-355-exams-date    | 2017-Latest-300-320-Exam    | Pass-300-101-Exam    | 100%-Pass-300-115-Exams    |
http://www.portvapes.co.uk/    | http://www.portvapes.co.uk/    |