@whs, Intel vs AMD CPUs(16 posts)
whs, let's discuss this here to avoid hijacking other threads. You have used every opportunity to blast the Intel Q6600 as an inferior processor, but have not shown any proof. You made it clear that you are disappointed with the performance of your new Q6600-based computer, but the CPU is not the only thing that can slow down a system.
Even if you are wary of benchmarks (I am too) there has to be some basis for comparison. Can you provide an example of a CPU-bound operation that is slower on your Q6600-based computer, but faster on your old AMD 4800+? For example, I do video encodes that used to take all night -- about 11 hours -- on my old P4 but my Q6600-based computer flies through these in about 2 hours! This task is completely CPU-bound and a good test of CPU performance.
It is fair to say that I have not measured any applications. But I have noticed on a day to day basis, that my Q6600 (Dell) with 4GB's runs slower than my 4800+ (HP) with 3 GB's. Both systems are new (less than 6 months old). As an example, just launching the IE to get the "about blank" takes at least twice as long on the Q6600. Maybe it is that I was disappointed because I expected a lightening fast speed. But then I realized that a single core on the Q6600 is slightly slower than on the 4800+ (2.4 vs 2.5 GHz) and that the FSB of the AMD is nearly twice as fast. Maybe that is the reason.
Maybe if you load the system up with a lot of parallel tasks, and even dedicate cores to tasks (like the Geek said he is doing), you would get a better overall throughput than on a dual core. But those are usually not the bread and butter applications one does.
Bottom line: I would not buy another Q6600 - at least not for what I am doing.
PS: Forgot to mention that I also noticed it to be quite slow on my video work - e.g. saving a clip (after trimming) seems slow.
Here's my problem with your argument. You say that your Dell computer is slower than your HP computer, but then you continue to say that therefore the Q6600 processor must be slower than the 4800+ processor. Non sequitur -- it does not follow.
Let's consider the speed of a single core. First of all, the clock speed has never been a meaningful measure of performance. All Intel Core2 processors are clocked slower than the Pentium 4 series, yet the Core2 provides more performance at a lower clock speed. So the 4800+ cores being clocked at 2.5 GHz does not make them faster than the Q6600 cores clocked at 2.4 GHz. As for the FSB, AMD certainly posts some very high numbers, but what is the actual throughput? The FSB transfers data between the CPU and the Northbridge, everything else is on the other side of the Northbridge, including the memory bus so a faster FSB does not mean superior CPU performance either. Finally, the Q6600 has a larger L2 cache, 2x4MB to the AMD's 2x1MB. Assuming the same cache hit rate, the Intel will spend less time on the FSB and memory bus resulting in better throughput overall.
These are all good arguments. I can only report what I experience in my little world. Maybe there is something else different in the Dell vs. the HP. But comparing the other components I could not really find a big difference. Here is another example. With the AMD the installation of SP1 (without the download) took 35 minutes. With the Q6600 it took 56 minutes. Where does the time go? The disks are comparable - I even have 2 disks on the Dell (ordered it that way for Ghost). But the disk specs look the same. So something seems to give and maybe it is not the processor alone.
whs, I know what it's like to be burned -- to buy a heavily praised product or service and find it not to my liking. I can understand the idea that you may decide to never buy a Dell again, a Q6600, or maybe any Intel Core2 processor. The experience you have in your little world means a lot to you, and my experience in my little world means a lot to me.
However, I disagree with telling others that the Q6600 is dog slow as if it's a fact or a widely held belief, instead of your personal experience. Also, to go from the general (your Dell) to the specific (the Q6600) remains a non sequitur. I would have no problem if you told people, "my Dell" or "my Q6600-based computer" is dog slow.
To the question, where does the time go? I have no idea. Why *is* your Dell PC slower than the HP? It doesn't make sense. Also, when you consider the operations you have mentioned -- web browsing, saving a video clip to disk, installing SP1 -- none of these are CPU-bound tasks, meaning the processor is not the main bottleneck in the overall execution. I would bet that if you look in the Task Manager Performance tab, that tasks like these don't push even one core, on either system, to 100%. Most of the CPU time is spent idling waiting for us users to press a key, waiting for the hard drive to fetch data, and so on.
ScottW, I always like your crisp logical approach to things - I really do. But in this case it is like buying a car where the sticker tells you 40MpG and after half a year your records show 22MpG. I probably do not have the right application mix for a quad, but then who has apart from the Geek.
I will take your suggestion referring to "my experience" rather than generalizing what I see happening. That is a fair point. Btw: I watch my CPU meter (4 cores) in my sidebar all the time. There are times when 2 or 3 cores go high up to e.g. 80%.
What is the model number of your Dell and HP machines? I can use that to look at the specifications and see how else they differ then processor. That will tell us more on the validness of this comparison.
My guess is that the dell is using a really inferior motherboard/chipset... which in my experience is really critical for the overall speed of the system. The HP machine is also using an NVIDIA video card as opposed to the intel integrated card on the dell, which could also make the system slower.
My personal experience: I've got 3 vmware virtual machines running XP, Ubuntu and Vista Home, and 102 other processes going (actually a light load for me), and the system is running so fast that I've been trying to think of things to slow it down so I feel like I got my money's worth =)
I might go for a 4th virtual machine...
The Geek, it only takes one application to bring a quad-core to it's knees. The video encoding I mentioned above is one reason I got my Q6600-based system. I'm encoding high definition video -- 720p or 1080p -- to VC-1 for playback on my Xbox 360. On my old P4, I would leave it to run overnight then come back the next day and it was still running! And that was all the P4 could do, so I had to wait for the encode to finish to use my system.
Now, I have 4 cores all working on the video encode and they run up around 95% and do the encode in a fraction of the time. Of course, there are all of these high quality settings that I never used before. If I turned them all on, the encode would still take overnight so there is a balance to find. Of course the remaining 5% of CPU is enough for simple tasks like e-mail and web browsing, so even while I'm encoding on this system I can still use it.
Having multiple VMs gobbles up memory quick, but unless there is a CPU-bound task running in all of them, the processor still has lots of overhead!
This topic has been closed to new replies.