Dual Core Intel Platform Shootout - NVIDIA nForce4 vs. Intel 955X
by Anand Lal Shimpi on April 14, 2005 1:01 PM EST- Posted in
- CPUs
SATA Controller Performance
Both NVIDIA and Intel offer support for NCQ in their SATA controllers, and given our recently renewed interest in NCQ performance, we decided to find out if there were any performance differences between the two SATA controllers. However, as we've found in the past, coming up with tests that stress NCQ is quite difficult. Luckily, there is a tool that works perfectly for controlling the type of disk accesses that you want to test: iometer.
An Intel developed tool, iometer allows you to control the size, randomness and frequency, among other things, of disk accesses, and measure performance using data generated according to these specifications. Given that NCQ truly optimizes performance when disk accesses are random in nature, we decided to look at how performance varied according to what percentage of the disk accesses were random. At the same time, we wanted the tests to be modeled on a multitasking desktop system, so we did some investigation by setting up a computer and running through some of our multitasking scenarios on it.
What we found is that on modern day hard drives, the number of outstanding IOs (IO Queue Depth) is rarely above 10 on even a moderately taxed system. Only when you approach extremely heavy multitasking loads (heavier than anything that we've ever tested) do you break into queue depths beyond 32. So, we put together two scenarios, one with a queue depth of 8 and one with a queue depth of 32 - the latter being more of an extreme condition.
In each scenario, we sent the drives a series of 64KB requests, 75% of which were reads, 25% were writes; once again, derived from monitoring our own desktop usage patterns.
We then varied the randomness of disk accesses from 0% (e.g. 100% sequential) up to 100% (0% sequential reads/writes). In theory, the stronger NCQ controllers will show better performance as the percentage of random accesses increases. We reported both Average IOs per Second and average IO response time (how long accesses took to complete on average):
With a queue depth of 8, the two SATA controllers offer virtually identical performance.
Looking at latency, Intel actually offers a very slight performance advantage here - nothing huge, but it's definitely there.
The results get much more interesting as we increase the queue depth to 32:
Here, NVIDIA starts to pull away offering close to a 20% increase in average IOs per second as the access patterns get more random (e.g. as more applications running at the same time start loading down the hard disk).
What's truly impressive, however, is the reduction in average response time - up to a 90ms decrease in response time, thanks to NVIDIA's superior NCQ implementation.
But stepping back into reality, how big of a difference NVIDIA's NCQ implementation makes depends greatly on your usage patterns. Heavy multitaskers that are very IO bound will notice a performance difference, while more casual multitaskers would be hard pressed to find any difference. For example, Intel was actually faster than NVIDIA in our gaming multitasking scenarios from our dual core investigation.
Facebook
Twitter
RSS
96 Comments
View All Comments
segagenesis - Thursday, April 14, 2005 - link
#15 - Score: -1, TrollWhy are you here then? Hell, you cant even read I guess. I gave you a link talking about heat output when you said it was "opinion" when it was stated Intel runs hotter. I can tell you that from fact from the 25 prescotts sitting in a lab here and when they are all running the A/C better be running too!
Why dont you come up with some facts yourself instead of insulting both the site and others?
Questar - Thursday, April 14, 2005 - link
#11,What makes you think I would care about "a huge landslide of flames regarding your post in this comment section"?
99% of the people on this site are ignorant cattle without the ability to think for themselves. They are here as part of a communal circle jerk over AMD cpu's. I care as much for their thoughts as I care about the thoughts of the cow that gets slaughtered for my dinner.
jimmy43 - Thursday, April 14, 2005 - link
I see we got an Intel fanboy posting in the forums :]Questar - Thursday, April 14, 2005 - link
#10,I see no information in your link that #8's argument. He specifically said that an A64 is faster than Intel in all applications except for video encoding. The link you provided actually proves him wrong, as there are other applications in which the comparable Intel CPU is faster, or the difference is insignificant between the two.
mlittl3 - Thursday, April 14, 2005 - link
Oops, sorry...correction"...release of the K8 architecture 8 years ago..." should read "...release of the K8 architecture 2 years ago..."
Sorry about that. AMD's 2nd year anniversary of the release of the Opteron is next Thursday where they will introduct dual core Opterons to the public. Can't wait.
mlittl3 - Thursday, April 14, 2005 - link
Oh my God!!! Questar where have you been?Go read all the reviews of Intel and AMD processors since the release of the K8 architecture 8 years ago. You have a lot of reading to do.
Don't give the editor-in-chief of a 8 million plus readership hardware review site advise about getting his facts straight. Anandtech receives and reviews hundreds and hundreds of hardware that you will never even dream of owning.
If a statement is made in a review from a site as reputable as Anandtech, it is not made lightly. You have all the right in the world to question it and seek a second opinion elsewhere, but it is COMMON knowledge among those reading CPU reviews over the last two years that AMD CPUs are faster in games and computational number crunching whereas Intel excels in audio and video encoding PERIOD.
You can easily explain these findings. Games and computational number crunching take low latency, high memory bandwidth to work well. Audio and video encoding need fast processor speeds.
Wait about five hours after the release of this review and you will soon be finding a huge landslide of flames regarding your post in this comment section.
Have fun! :)
PS. You must not read at all if you think anyone has found a performance advantage of PCI-E and DDR2 over AGP and DDR for current software applications. Not tomshardware, Hardocp, Xbitlabs, Anandtech, etc. have found a performance increase between AGP 8X vs. PCI-E 16x or DDR400 vs. DDR2400/533/667.
segagenesis - Thursday, April 14, 2005 - link
#9 - You havent been here long have you? Intel does have a performance advantage in encoding applications... but your claims of without proof to others?http://www.anandtech.com/cpuchipsets/showdoc.aspx?...
Umm yeah. Pay attention 007.
And why even claim AGP has no performance benefits over PCI? Ever heard of a shared bus (PCI)?
?LOGIC ERROR
READY.
Questar - Thursday, April 14, 2005 - link
#8,Without proof, it's your opinion. So, find some reviews that showed an Intel cpu being significantly slower than an equivalent A64 in all applications except video encoding.
I assume you are using an AGP video card. Why? It has no performance benefits over a PCI card. Or is FPS in a game the only way you measure perfomance?
n yusef - Thursday, April 14, 2005 - link
#5"Honestly, Intel processors and even the platform haven’t been interesting since the introduction of Prescott. They have been too hot and poor performers, not to mention that the latest Intel platforms forced a transition to technologies that basically offered no performance benefits (DDR2, PCI Express)."
Your opinion only, don't make this out to be fact.
That is pretty much fact. In all areas except encoding, they were worse performers than their competiton (Athlon 64). The extra heat sure didn't improve that either. As far as forcing DDR2 and PCI Express, when they didn't improve performance, you can't disagree with that.
radx - Thursday, April 14, 2005 - link
QuestarI'm happy you're not the one writing these reviews here at anandtech. :-)
Go Anand!