AMD's 65nm Preview Part 2 - The Plot Thickens (Updated with Information from AMD)
by Anand Lal Shimpi on December 21, 2006 12:12 AM EST- Posted in
- CPUs
Of Die Sizes, Voltages and Power
When we published our first Brisbane article, an astute reader pointed out that AMD appeared to have rather poor die scaling with the 90nm to 65nm transition. Given perfect scaling, you'd expect a 65nm shrink of a 90nm core to be approximately 52% the size of the larger core. Looking at Brisbane, AMD went from 183 mm^2 with its 90nm Windsor core down to 126 mm^2 at 65nm, making the newer core almost 69% the size of the older one. If we look at Intel, most of its die shrinks are coupled with new architectural functionality or larger caches, so it's not unusual to see scaling in the 70 - 80% range. However, with Brisbane, transistor counts remained the same according to AMD (approximately 154M) yet we still saw relatively poor scaling with die size. The table below provides some reference points for die sizes and transistor counts:
| CPU | Manufacturing Process | Die Size | Transistor Count |
| AMD Windsor | 90nm | 183 mm^2 | 154M |
| AMD Brisbane | 65nm | 126 mm^2 | 154M |
| Intel Smithfield | 90nm | 206 mm^2 | 230M |
| Intel Presler | 65nm | 162 mm^2 | 376M |
| Intel Prescott-2M | 90nm | 135 mm^2 | 169M |
| Intel Cedar Mill | 65nm | 81 mm^2 | 188M |
Note that when Intel moved from 90nm to 65nm with its Smithfield to Presler transition, the 65nm core ended up being almost 79% the size of the older core. However when you take into account that transistor count went from 230M to 376M, all of the sudden the scaling looks a lot better. The bulk of that increase in transistor count was Presler's extra L2 cache, which happens to shrink quite well, so it's unfortunately not the best comparison. Looking at Prescott-2M to Cedar Mill, Intel saw very good scaling with a 40% smaller chip at 65nm (60% the size of the 90nm core).
Obviously some structures within a core will shrink better than others in terms of surface area, so perfect scaling isn't necessarily a target reality, but one of the questions we asked AMD was why the new core is seemingly so big. We couldn't get an official answer from AMD as many of the folks that would be able to get us such a thing were on vacation and unreachable, but the gist is that coupled with the fact that not everything scales well with manufacturing process, this is AMD's first 65nm chip, and AMD tends to make many improvements to its manufacturing process over time. The chip we're comparing Brisbane to was made at the pinnacle of AMD's 90nm manufacturing cycle, so it's quite possible that, with time, AMD will improve its 65nm process to the point where a smaller Brisbane would be possible. Until we can get a more technical explanation from AMD, that's the best we can report on this issue. On to number two...
We weren't impressed with the power consumption of Brisbane at all in our first review; while it was lower than its 90nm counterpart, in many cases it wasn't all that much lower. Once again this is an issue of comparing a very mature 90nm process with AMD's first 65nm chips. You see, the voltages that Brisbane will be manufactured at range from 1.250V to 1.350V, with the coolest running, highest overclocking, least power consuming chips running at 1.250V and the worst examples running at 1.350V. Both of our Brisbane samples, the 5000+ and 4800+, ran at 1.350V. Note that our 90nm 5000+ ran at 1.300V, a lower voltage than the newer 65nm core. The fact that many 65nm parts aren't at much lower voltages yet is why the highest clocked Athlon 64 X2s are still 90nm CPUs, such as the 5600+ which runs at 2.8GHz.
At some point in the future, AMD will hopefully be able to tune its manufacturing so that we will get lower voltage, lower wattage 65nm parts. This is also part of the reason why we encountered such dismal overclocking results with our 5000+. The 4800+ we tested fared no better, with our best overclock on stock cooling ending up at 2.837GHz (227 x 12.5) - not terrible for stock cooling, but not great either.
The impact of higher voltages on power consumption also applies to Intel as well. As you will see in our power comparison, in a number of cases our Core 2 Duo E6300 required even more power than the E6600 we tested last time. The reason being that our E6300 sample runs at a core voltage of 1.325V vs. 1.2625V for our E6600 sample. Just things to keep in mind as you look at the power results over the next few pages.
| CPU | Core Voltage of our Test Chip |
| AMD Athlon 64 X2 5000+ (90nm) | 1.3000V |
| AMD Athlon 64 X2 4600+ EE (90nm) | 1.2500V |
| AMD Athlon 64 X2 3800+ EE SFF (90nm) | 1.0750V |
| AMD Athlon 64 X2 5000+ EE (65nm) | 1.3500V |
| AMD Athlon 64 X2 4800+ EE (65nm) | 1.3500V |
| Intel Core 2 Duo E6600 (65nm) | 1.2625V |
| Intel Core 2 Duo E6400 (65nm) | 1.3125V |
| Intel Core 2 Duo E6300 (65nm) | 1.3250V |
Facebook
Twitter
RSS
52 Comments
View All Comments
Spoelie - Thursday, December 21, 2006 - link
This is not the first time this has happened, it may be easy to forget, but do you guys remember the thoroughbred?Thoroughbred A was the first 180nm to 130nm shrink and had a hard time reaching the speeds the mature 180nm cores were getting. It wasn't till AMD added another layer to the core (Thoroughbred B) that we saw the expected speedups from a die shrink.
PetNorth - Thursday, December 21, 2006 - link
Anand:Why don't you set manually the voltage, to know really what's the improvement with 0.65 transition?
1.30v to compare it with 5000+ 90nm, and 1.25v to compare it with 4600+ EE 0.90nm.
It would be a good thing IMO.
yyrkoon - Thursday, December 21, 2006 - link
There are already people who believe that odd numbered multipliers offer worse performance compared to even numbered multipliers. I cant help but wonder why AMD chose to start implementing floating point multipliers now. The first thing that comes to mind, is maybe to refine their pricing ? Although, I've never really noticed much performance (if any) difference using odd vs even numbered multipliers, I can not help but wonder if floating point multipliers will play a factor in performance.Regs - Thursday, December 21, 2006 - link
AMD has been stepping in baby steps in their innovation merits. Ever since the IMC and the enhancements from K7 to K8 it seems like they improve little by little. I hope this gives them a rude awakening to how competitive the market can or could be in future. If they did it before they can do it again.As for the transition to 65nm, it was no surprise that these parts could not over clock very well. The K8 is showing its age and I think there are no more ways you can breathe life back into it especially when Core Duo is out in the market.
mino - Thursday, December 21, 2006 - link
Why awekening, and why rude? The fact is AMD kept PARITY with intel on power AND performance inthe lower end with 90nm!!! part with Intel beeing at 65nm for a year allredy!In other words, When AMD's 90nm process is FAR better that Intel's ever was. Same happened with 130nm. Two words: SOI,APM.
No confusion, all thi means no one should avaluate AMD vs. Intel on process_used base. Simply put, as of now(at stock) Intel rules on perf&power while AMD rules on idle_power and price(up to 4200+/E6300 combo).
IntelUser2000 - Thursday, December 21, 2006 - link
Intel bins Core 2 Duo by power consumption.
xsilver - Thursday, December 21, 2006 - link
just to clarify further; all e6600's will have lower stock voltages than e6400's and all e6400's will have lower stock voltages than e6300's?at both idle and load?
how successful are the conroes at undervolting?
Accord99 - Thursday, December 21, 2006 - link
Pretty good, my week 25 E6600 is stable at 2.6GHz/1.1v (My P5B-dlx doesn't go any lower) with dual-P95. The heat output is easily cooled passively by a Scythe Ninja.Here's a thread, one person has a E6600 that does 2.4@/~1v
http://www.xtremesystems.org/forums/showthread.php...">http://www.xtremesystems.org/forums/showthread.php...
blackbrrd - Thursday, December 21, 2006 - link
I have seen a E6600 running at 1,0v at load... It was obviously very cool running :)My E6400 is running at 1,15v at idle (2133MHz) and 1,25v at load (2133MHz)
Power saving features were off in both instances...
haugland - Thursday, December 21, 2006 - link
AMD win in one aspect...I you really consider power consumption to be important, it is much more important to look at idle power consumption than power consumption at full load. Most business PCs idle a lot of the time, and AMDs CPUs are much better at saving power at idle.
EIST was designed for P4, and for a 3+ GHz P4 it makes sense to drop the multiplier to 6. However when the E6300 normally run at a multiplier of 7, you don't get much of a power saving by dropping the multiplier to 6. AMD C'n'Q allows for much lower settings.