Intel Yonah Performance Preview - Part I: The Exclusive First Look at Yonah
by Anand Lal Shimpi on November 30, 2005 2:50 AM EST- Posted in
- CPUs
Yonah vs. Dothan
We didn’t have much time to put together this piece, but at the same time we wanted to present the most complete picture of Yonah as possible, so we went back to our last Pentium M on the desktop article and configured our Yonah system identically so we’d have as close to an apples-to-apples comparison as possible. Of course it is impossible to use the same motherboard, due to the socket differences we’ve already mentioned, but the rest of the systems are configured identically. We apologize in advance for the brevity of the benchmark suite, in due time we will present an even more thorough look at Yonah, but for now we are working with what we’ve got. Also keep in mind that the platform and processor are both pre-release samples, so performance could change, most likely for the better.
With that said, we've got a question and that is: how does Yonah stack up to Dothan?
Unfortunately, our Yonah only runs at 2.0GHz, and our reference Dothan numbers are from a 2.13GHz CPU - so we don’t get the clock for clock comparison we were hoping for, making it even more difficult for Yonah to impress. Thankfully our first benchmark is clock speed independent as we look at how cache latencies have changed from Dothan to Yonah using ScienceMark 2.0:
| L1 Cache Latency | L2 Cache Latency | |
| Dothan | 3 cycles | 10 cycles |
| Yonah | 3 cycles | 14 cycles |
And changed they have indeed. If you’ll remember from our earlier desktop Pentium M investigations, Dothan’s very quick 10 cycle L2 cache allowed it to be competitive with AMD’s Athlon 64, despite lacking an on-die memory controller. With the move to Yonah however, the L2 cache latency has gone up a whopping 40%. While we’re still dealing with a lower access latency than the Pentium 4, this increase will hurt Yonah.
We’re guessing that the increase in access latency is due to the new dynamically resizable L2 cache that’s used in Yonah. In order to save power as well as maximize the use of the shared L2 cache between cores, Yonah can dynamically adjust the size of its L2 cache, flushing data to main memory when faced with low demand. The associated logic is most likely at least partially to blame for the increase in L2 cache latency.
So Yonah has a slower L2 cache working against it, but two cores and a handful of architectural enhancements working in its favor - let’s see how they stack up in the real world.
First up, we’ve got our business application tests:
| Business Winstone 2004 | Communication (SYSMark 2004) | Document Creation (SYSMark 2004) | Data Analysis (SYSMark 2004) | |
| Dothan (2.13GHz) | 24.3 | 129 | 202 | 118 |
| Yonah (2.0GHz) | 21.6 | 146 | 215 | 138 |
Dothan has a sizeable lead in Business Winstone 2004, which we’ve always attributed to its low latency L2 cache. Since the benchmark gets no benefits from dual core, and doesn’t take advantage of any of the SSE improvements to Yonah, the advantage is clearly in Dothan’s court.
The SYSMark tests paint a different picture, with Yonah outpacing the faster clocked Dothan by 6 - 17%. What’s interesting to note is that in these tests, the performance advantage isn’t exclusively attributable to the advantage of having two cores - Yonah’s architectural advancements are at work here as well.
The digital content creation tests are where Yonah’s improvements should shine:
| Multimedia Content Creation Winstone 2004 | 3D Content Creation (SYSMark 2004) | 2D Content Creation (SYSMark 2004) | Web Publication (SYSMark 2004) | |
| Dothan (2.13GHz) | 29.8 | 188 | 255 | 169 |
| Yonah (2.0GHz) | 34.7 | 264 | 323 | 236 |
And shine they do; thanks to a combination of the move to dual core as well as the architectural improvements over Dothan, Yonah shows anywhere between a 16 - 40% increase in performance.
| DivX | Doom 3 | |
| Dothan | 39.7 fps | 95.5 fps |
| Yonah | 57.5 fps | 93.8 fps |
The DivX test shows what we’ve pretty much seen across the board from dual core scaling in video encoding, so there’s no surprise there. Our only gaming benchmark, Doom 3, shows a hazier picture with Dothan on top, and Yonah close behind. We will investigate gaming performance of Yonah much closer later on.
What we can walk away from these benchmarks with is an idea of the level of improvement to expect from Yonah, but now comes the real test - how does it stack up against other desktop processors, especially the Athlon 64 X2.
Facebook
Twitter
RSS
135 Comments
View All Comments
JoKeRr - Friday, December 9, 2005 - link
http://www.notebookforums.com/showthread.php?p=147...">http://www.notebookforums.com/showthread.php?p=147...a new model of toshiba with yonah @ 2ghz, 2x512ddr2 667 annd x1400 graphics. Very interesting.
Obviously yonah with 2x512mb ddr2 667 should perform better overall than ddr2 533.
Similar to how dothan with ddr2 533 completely smokes out dothan at same speed but ddr2-400 or ddr333 (and surprisingly many manufacturers adopted that, example: sony FS toshiba m50 etc... stupid choice for the manufactures, not to mention turbocache or hypermemory will suffer a lot as well due to slower access to system memory).
So when is the official release date for Yonah based lappy?? and where's the Part II of Yonah article??
Thanks again AT, great job.
IntelUser2000 - Monday, December 12, 2005 - link
It was weird how single channel DDR2-533 is faster than dual channel DDR2-400 for Dothan. It seems the chipset doesn't take advantage of dual channel at all, so getting faster single channel stick is better than slower dual channel. So I assume it will be same for Yonah.
Yonah will be released Jan 5, 2006 at CES(some trade show I heard).
Furen - Thursday, December 15, 2005 - link
Two DDR channels are bottlenecked by the FSB. The P4 can use dual-channel DDR effectively because of its much higher FSB.cryptonomicon - Saturday, December 3, 2005 - link
I think yonah is a nice option for low power compact notebooks. As we can see here it doesnt really compare against the desktop amd X2 parts, but will line up good against mobile duel semprons and turion right? Also to be noted is that it has a low clock speed. How will it overclock?.. this interests me greatly.IntelUser2000 - Monday, December 5, 2005 - link
Actually the highest end will initially be available at 2.17GHz, and a quarter later there will be 2.33GHz. There will be "Extreme Edition" like versions that are clocked 1 or 2 speed grades beyond the 2.33GHz, so 2.5 or 2.67GHz. Of course the "EE" versions will be higher power consumption(Actually its officially called E).
snorre - Saturday, December 3, 2005 - link
Are you trying to make "Yonah" look better than it really is (another Intel lemon) perhaps?And did you check for CPU throttling activity in your power consumption tests?
Zebo - Friday, December 2, 2005 - link
Everyone keeps talking about power consumption..I for one am not impressed. First of all with AMD - X2 in these tests is basically a overvolted part.. will run just fine at 2.0Ghz with 1.15-1.2V rather than 1.4V used here, lowering power signifigantly in "X2 turion" form when it gets here. Second, even so, the power consumption difference was'nt that great 109W vs 144W loaded.. It's not like comparing a P4 setup which sucks over 300W or double X2 power just to run here - you're talking 30-40W differential which will surley be addressed with the addition of DDR2 and unvervolting even AMD's 90nm processes. Nevermind what 65nm should add. AMD's not in trouble in this sector and Yonah gods greatest gift to the sector either. Since we can't buy product now it all moot anyway but either should make a great notebook.Since I don't like notebooks I wish anand had done some serious clocking here - after all I/we really care about is can intel reach 3.0+ Ghz with this process and Yonah for those of us who want to slam these chips on the desktop and as a precusor to Conroe.:) That's what intel really needs to perk my intrest again. 3+ Ghz to really compete again on desktop - I don't really see them getting back to Northwood vs Athlon XP day with this chip though no matter how high it clocks.
Marmion - Thursday, December 8, 2005 - link
Ever thought that speedstep wasn't used when compiling this test. Also we're basing consumption on the entire platform. This platform would differ compared to the Centrino platform - low power graphics, wireless and chipset. When idle, the Yonah core shuts down a core. It is fairly obvious that the consumption figure shown for the Yonah does not utilise speedstep. You can also underclock a Dothan if you want.I also find it difficult that the Turion X2 being cheaper than an equivelent Yonah, as shown by the price difference with AMD64 X2 over the single core equivelent, however the Yonah cost is equal to the Dothan at the next highest frequency, but the yonah has 2 cores - ie Yonah @ 2Ghz = Dothan @ 2.13Ghz.
A Turion X2 will give up performance of the desktop equivelent to save power - its to be on the 90nm platform. So Yonah compared to a Turion X2 looks very good in all rspects (except 64bit but I don't need it so I don't care)
IntelUser2000 - Monday, December 5, 2005 - link
Who cares if X2 can be undervolted or overvolted?? Its not. That's default voltage of X2's. By under or overvolting it you are essentially putting it out of spec. Probably Yonah can be undervolted too. Whats your point here??? Plus DDR2 does not consume that much less power, since DDR2 clocks higher. I have even seen presentations that say 533MHz DDR2=400MHz DDR. Though its likely that's wrong, it shows that DDR2 isn't low power as you think.
For those who thinks dual core Turions with DDR2 would make it lower power. No, because the DDR2 would be at 667MHz, NOT consuming less than DDR400, plus it will be dual channel, unlike the single channel on the current Turions, ACTUALLY consuming MORE power.
People are essentially dumb here since most people almost assumes Yonah is a desktop chip. And people whos saying Yonah MAY be competitive at 2.6GHz, there WILL be 2.5-2.67GHz versions on "Extreme" ones. That's gonna be something as it would be higher clocked than X2's.
DDR2 DOES NOT have power consumption advantage since it CLOCKS MUCH HIGHER THAN DDR. The tested configuration DOES NOT use a mobile chipset.
AMD has NO PROCESS advantage by using SOI, since at 90nm and 65nm process, Intel is superior in leakage current at SAME transistor speeds for BOTH process WITHOUT using the wonderful SOI some are touting.
65nm DOES NOT NECESSARILY LOWER POWER OVER 90nm, look at initial AXP's at 130nm, called Throughbred, they were about 10% less power and no overclock.
Zebo - Wednesday, December 7, 2005 - link
Claim down fanboy.. geez even your user name says it you don't have to keep broadcasting it. It's default for a desktop chip, it's overvolted in the sense that anands substituting a desktop chip and competing with a notebook product here...as in the dual turons slated for lappys will be 1.2 or 1.25, not 1.4 futher reducing power..just like turions today vs regualar desktop A64's. If you really think Dual turions will be 1.4 then we have nothing further to discuss.And you're just wrong about DDR vs DDR2
http://www.samsung.com/Products/Semiconductor/DRAM...">http://www.samsung.com/Products/Semicon...Info/101...