Computer upgrade time
 

I'm thinking of upgrading to either an Intel Core i7 920, or Core i7 860.

According to benchmarks, they're about the same speed. There's only two differences I can tell; one is the platform, LGA 1156 (860) or LGA 1366 (920); and the second is dual channel memory (860) vs. triple channel memory (920).

The 860 ends up being cheaper when figuring in motherboard, memory, and processor. I don't think the Intel i5 750 is an option -- I'm not one for overclocking. System stability is paramount for me.

Is there any compelling reason to go with the 920? What are your thoughts on this?

My only concern is if Intel plans to support two platforms from here on out. I mean... if I intend to upgrade in two years, will I be able to on the 1156 platform?

one thing to rarely worry about is that the motherboard you bought today will work tomorrow, it wont :-) unless you bought DOWN in the chip with an obvious processor upgrade path in existance that day.
Out of any motherboard i get 1 update of the processor and/or video or something, and the Differeces arent squat. then the motherboard itself is obsolete for further updating.

for video work the processor is a lot of it, but for computer upgrade paths the busses, the memory the hard drive connectability, even the power supply connectibility becomes an issue WHEN you really want to 2TIMES speedup your computer.
without a new motherboard, the most overall speedup i have ever gotten was some worthless 10-20 percent (overall speed).

Every time i purchaced the most upgradable most diverse most open ended motherboard at very high prices, and the next real advancement came from Tossing it all in the trash again. They even tried to come out with Processor socket adapters to provide an update path, and that didnt work well either.

remember video cards? PCI AGP PCI-E , Chipsets 35 - 66 - 100 -133 - 400 , hard drives 33 66 100 2g 3g , then count the number of "Sockets" that has come out in the last 20 years, and do it without using your Toes :-)
i am not saying its a conspiracy, and that they manage to upsell a rack of $300 motherboards with ever $300 processor, but they dont seem to fit things in the same package very often when the technology itself changes 180*

12GB of ram is fairly cheap on the 1366 whereas the 1156 can only do 8GB for a low cost. 1366 can support a max of 24GB and 1156 is 16GB but 4GB sticks of ram are still $200+ each.

Also, Intel will only be releasing their 6 core 12 thread i7/i9 for the 1366 due out this summer. 1366 is Intel's performance chip and 1156 is their mainstream chip.

This is not quite the case right now. Memory prices, in fact, have actually increased for all sizes. As such, even 3GB of memory (via three 1GB modules) is now more expensive than it was just a couple of months ago.

And for all of the triple-channel hype, the 1366 actually performs slower than the 1156 per CPU clock cycle (or put it this way, the 1366 is less efficient than the 1156) with the same number of cores. This is because although the memory controller in both the 1366 and the 1156 processors are on-die, the 1366 still has to rely on external buses for everything else while the 1156 puts some of the features that were previously handled through external buses on the CPU die. I am suspecting that Intel has been artificially limiting the current maximum clock speed of the 1156 as to not take sales away from its top-end 1366 processors. What's more, the 1366 still officially supports only DDR3-1066 maximum while the 1156 officially supports DDR3-1333. And current JEDEC standards for DDR3 memory allow only one memory module per memory controller channel at the controller's maximum officially supported speed. This means that installing 12GB of memory using 2GB modules on a 1366 motherboard might force the memory to operate at a slower DDR3-800 or even DDR3-667 (if your DDR3 motherboard's memory speed could even be set that low) - one to two speed grades lower than the maximum supported by the memory and memory controller combination. One might not currently notice much of a difference even at that since few applications are especially sensitive to deficiencies in memory performance.

And even with the $200+ price for 4GB DDR3 memory modules, the 1366 rig still costs more money on average than a 1156 rig of the same performance level. This is because on the whole motherboards for the 1366 remain significantly more expensive than their 1156 counterparts right now - largely because the only chipset that's currently available for 1366 motherboards is the expensive-to-manufacture X58 while the 1156 can use any of the less expensive P55 series chipsets. (And yes, the relative prices are for an entire system, with a CPU, motherboard, memory, hard drive, video card, etc.)

Randall,

There's no problems getting DDR3 to behave at or above spec on a 1366 system. I have 12GB running @ 1600mhz and my 920 CPU running at 4.2ghz.

Probably with better-than-average cooling.

After all, I have had trouble getting anything stable at anything above stock speed lately due to heat issues, especially in summer.

And I still stand by my finding about the relative cost of a 1366 system versus a 1156 system, especially in the cost of the motherboards for such systems. And the memory speeds are merely based on the officially supported specs, not the "maximum stable speeds" that many users have gotten.

In any event, I would have picked either one if I were building a new system today. But neither system would provide sufficiently improved performance that's worth the cost of its core components (primarily the CPU, motherboard and memory) if I already have a Core2 Quad system (I have a Q9450 with 4GB of DDR2-800 memory myself). And right now DDR3 system memory (required by all i7 and i5 systems regardless of CPU socket type) still costs nearly double that of DDR2 system memory per GB. An upgrade to Socket 1366 would have cost me a minimum of about $600 with only 6GB of memory - and that's more than high enough even with an i7-920 that I would have needed a motherboard that's an especially good overclocker in order to be worth anywhere near the cost. Many of the 1366 motherboards are mediocre to poor in overclocking (a poor overclocker crashes or locks up at clock speeds slower than what most other motherboards have no trouble reaching), so I would have had to select the right mobo. (Intel would also not be my first choice in an enthusiast motherboard since that brand has limited or no overclocking capabilities. After all, if I can get a given i7 CPU to run stably on an Intel Extreme Series mobo at the maximum selectable clock speed, only to find that the motherboard's highest available CPU speed settings are limited compared to the maximum reliable speeds attainable from some other brands of motherboards, then it's not worth it getting that Intel-branded motherboard in the first place especially since that mobo is on the expensive side for the features and performance it offers.)

I have to disagree with the theory that it's not worth going from a Core2 Quad to i7. I just did that and get HUGE improvements, especially when overclocked. And my stuff is crammed into a 4U rackmount case, though I upgraded to a Noctua CPU cooler. Well worth the money.

If you have Premiere CS4, try running the PPBM4 benchmark and see what you can come up with. If you look at the benchmark results, mine is the "Wild Fly" machine that has the second highest score, and that's with minimal tweeking, not hours of messing with overclocking settings.

PPBM4 home page

The problem is that I simply don't have anywhere near $600 to spend on such an upgrade all at one time. In fact, I don't have even $100 right now. Plus the fact that I have recently upgraded operating systems resulted in me having to wait a few months longer before spending a single cent for upgrades.

And no, I don't have Premiere CS4. This suite costs another $800 on top of any hardware upgrades. It could be years before I could earn the $1400 needed even if I work for it (the job market is virtually nonexistent right now in my area except for high-level positions which require extensive education and experience). And every month that I wait for this upgrade makes me fall that much farther behind.

So your reservations about upgrading are more a matter of money and less a matter of you thinking that an i7 upgrade is inconsequential when one has a Core2Quad? I can understand the former, not the latter. The i7 upgrade is a BIG upgrade performance-wise.

To the original poster, Devin, I would definitely suggest going with the 1366 platform. It offers more for a very small amount more money, especially considering the new i9 processors will run on 1366 and not 1156. If you need more info, check out some of the posts that Harm Millaard has made. He knows a LOT about hardware, and researching via his input has gotten me to where I have a dynamite editing system.

In my particular situation, it's much more a matter of money. Also, I was speaking of everyday office/Web/e-mail apps which do not take anywhere near full advantage of even a Pentium II processor, let alone current processors. For video editing apps, then yes, the i7 and i5 do provide quite a significant improvement over the older Core2 series processors. I was only trying to say in my earlier post about this, that if I were going to spend that much money on such an upgrade I might as well spend more for an even higher-end processor than the 920, especially if I end up getting a motherboard which turns out to be relatively unstable compared to most other motherboards when the processor is overclocked (and yes, I would only consider such a motherboard if its price is remarkably low compared to the other motherboards for that particular socket out there). IMHO the bottom of the high-end processor line is not worth the price if it is priced nearly as high as the next speed grade higher (for the entire processor/motherboard/memory combo, not just for the CPU).

In the case of the 920, it's the best bang for the buck. It is stock 2.66ghz, mine is OC'd to 4.2ghz on a middle of the road motherboard. If you go with the i7 920 on an Asus P6T motherboard, the cost is well below what you would spend going to a 940 or higher and getting a cheaper board. I am not a fan of skimping on the motherboard either. If the board won't OC well, that means it's most likely built with crappy components. In this day of online retailers and online reviews, it's pretty easy to discern what is good and what is not.

As far as everyday office/web/email performance, you won't find that people on DVinfo.net are asking for recommendations for machines that are designed for that. This site is for people that edit video, and they want video editing systems.

I've double-checked the official JEDEC standards for DDR3 memory and memory controllers again, and I was mostly incorrect with that statement. It seemed as though I carried over the maximum number of memory banks per channel from the official JEDEC standards for the highest official JEDEC speed grade for DDR2 memory: JEDEC only allowed two banks of memory per channel at DDR2-800 speed (one double-sided module or two single-sided modules). The official DDR3 specs now allow four banks of memory (two double-sided modules) per channel at up to DDR3-1333 speed. The maximum current JEDEC-standard DDR3 memory speed grade is DDR3-1600, which the official specs allow only two banks (one double-sided module) per channel at this speed. If for some reason you have installed two DDR3-1600 modules per channel in a system with a fully JEDEC-compliant DDR3-1600 memory controller, the official JEDEC standard dictates that the memory controller speed be automatically downclocked to DDR3-1333 speed. (For the record, none of Intel's current consumer CPUs officially support memory speeds higher than DDR3-1333 - and the official DDR3-1333 memory speed support on Intel's consumer CPUs (as opposed to Intel's motherboard-based memory controllers such as the ones used in the X38 and X48 chipsets for Core2 systems) is currently limited to the "mainstream" Socket 1156 platform.)

Thus, at the official Intel Socket 1366 maximum of DDR3-1066, up to six double-sided memory modules can be operated at that speed. Some Socket 1366 motherboards have as many as nine DIMM slots - but what good is having all those slots when for all practical purposes you can use only six of them at a time at full memory speed?

"for all practical purposes"

Are you bent on confusing people? FOR ALL PRACTICAL PURPOSES you can populate all slots at the full speed the memory supports as long as your motherboard supports it. Keep in mind that sometimes this means going in and manually configuring the memory settings in the BIOS, but it is a completely normal process.

Again, I'm running 12GB (six slots filled) of DDR3 1600 on my socket 1366, and it's running AT 1600mhz. You aren't limited to 1066mhz.

I'm starting to get the feeling that you have no "practical" experience with the i7 platform and are formulating your opinions based on your understanding of previous generations of hardware and a misguided understanding of a few older spec sheets you've read.

That's because most Socket 1366 motherboards have only six memory slots total.

And as for my post above, I was only stating the official JEDEC standards and not individual motherboards' communication between the motherboard BIOS and the CPU's memory controller. (And indeed, Intel's reference X58 motherboard, a laboratory-only motherboard which differs greatly from the same company's Extreme Series motherboard using the same chipset that's sold at reseller level, does strictly follow official Intel and JEDEC specs.) Individual users may have better luck with higher memory speeds than what JEDEC officially supports.

In other words, I am stating (but did not previously) that YMMV. And if I had a 1156 or 1366 CPU which did not work reliably when the memory speed is overclocked beyond Intel's and JEDEC's officially supported specs, either I have a relatively bum CPU memory controller or a relatively bum motherboard which barely achieves minimum performance and reliability standards for that CPU type.

As a general rule go for the less expensive system and use the money you save for your next system. In all my years of owning a PC, I have never upgraded the cpu on an existing motherboard. People including myself get bogged down in the little details, in 3+ years from now, no one will care whether you have a i720 or i860, because there will be a whole new cpu with a different socket that blows both of them away.

Randall, I'm not going to argue it further. You are spouting off JEDEC standards and Intel initial specs that have nothing to do with real life experience with the hardware. Let's get back on topic with Devin's original question.

Devin,

The 920 is better for a video editing system. Aside from the fact that it overclocks easier, the fact that it can cost effectively support 12GB RAM vs 8 GB RAM is a BIG thing when video editing. It is the preferred choice by many professional video editors. If you look at the PPBM4 benchmarks (PPBM4 home page), you'll see that most of the top scores come from the 920 because it overclocks so well. Harm Millaard has the top score, and he's on this forum. He has put a LOT of time in to researching and building the best machine, and he has no issues helping others optimize their systems.

Another good resource is the VideoGuys DIY article. Videoguys Blog - Videoguys' DIY7.7: Intel Core i7 with Vista 64 AND Now Windows 7

In the end, the best thing to do is to follow others with experience building great editing systems. I actually have one and have built multiple i7 systems. My 920 is faster than any other I've built, including a 965 extreme that cost 3 times more for the entire system as mine did.

If you are looking to do more than just edit DV on your mom's consumer Sony cam, then you'll want to get the most out of your system that you can. You'll skip the cheapest options and go for a mixture that gets you the best performance/cost balance.

My suggestion is to base your system on a i7 920 on an Asus P6T board.

OC'ing and Video Editing
 

I'll be building an i7 920 system soon. Based on my many years of system building, I'd say you're right on with your suggestion, Craig. "Cheap" usually menas poor quality with computer components.

A few places you don't want to "cheap out" on:
1: Motherboard. Cheap ones will die early and cause stability problems.
2: Power supply. Good power supplies are not cheap unless they're on sale. Read reviews on enthusiast sites (like hardocp.com, or overclockers.com) before buying.

Poor-quality motherboards and power supplies often fail due to low-quality capacitors. I find the cheaper motherboards are often physically thinner than a good board by Asus or Gigabyte. Thin boards flex and crack solder joints.

I stopped overclocking my video editing systems. I've had problems in the past with oc'd systems. They seem to work well for just about everything except heavy duty video encoding. Overclocking is an art, though. Maybe I just don't have the touch. ;-)

i agree with ken, get a good brand of mobo (asus, gigabyte, intel) and power supply (Antec, Seasonic, Corsair, etc)

The only series of Intel-brand motherboards that I recommend, especially if one is tweaking his or her rig's settings, is the Extreme Series (the other series of Intel-brand motherboards, such as the Media Series and the Essential Series, permanently lock your system to stock speeds, voltages and timings). Otherwise, some (but not all) Asus and GigaByte motherboards are better buys.

And poor quality can be found at all price points, not just in the cheapies. You can find a motherboard that's astronomically expensive that's of extremely poor quality, especially at the price which it goes for (or put it this way, they spent way too much money on the components but are very careless in assembling the board).

Although I wouldn't get an Intel because they're overpriced, they're still a quality name, and thats why I listed them. In addition, I'm not into OCing.

I would also spend extra for mobo with built-in firewire, since it will cost you more to purchase an expansion card separately.

I play it safe these days. I used to oc my systems when I used them to play games, but now that I'm using them to make money, I want to avoid any down-time. Running everything at it's rated speed, timings, and voltages provides the best stability (in my experience, anyway).

The last thing I want is a crash while rendering a video.

At Craig's suggestion, I am getting back to this original topic.

Actually, the 860 costs more than the 920 (when comparing CPU prices). What's more, a P55 motherboard can cost nearly as much money as the average X58 motherboard these days (assuming that both motherboards have the same amount of features, but not necessarily the same level of quality). This is because the 860 actually runs at a higher base core clock speed of 2.8GHz while the 920's base core clock speed is only 2.66GHz. These offset each other; as a result, the total cost difference boils down squarely to the memory. Unfortunately, one cannot directly compare prices here, either, because the maximum amount of memory which can be installed on a typical 1156 motherboard without having to spend an astronomical amount of money per GB is 8GB (remember, 4GB memory modules currently cost 2.5 to 3 times more money per GB than 2GB memory modules), or on Socket 1366, 12GB.

Secondly, if you plan to upgrade in two years, the new CPUs would likely (though not certainly) be on an entirely different, non-compatible socket.

What's more, the forthcoming new 32nm CPUs for Socket 1156 will initially be available only in dual-core form (the relatively low-end i3) while the six-core processors will be available only for 1366. (The quad-core processors for Socket 1156 will continue on their existing 45nm process.) This further separates the market for the two platforms. Once the low-end processors for LGA1156 get produced in large quantities, Intel can then finally pull the plug on LGA775.

In other words, right now a combo consisting of an i7-920 processor with a compatible X58-based motherboard costs more than a comparable combo consisting of an i7-860 processor and a comparably-featured P55 motherboard only because the 920 system has 2 extra GB of memory. If both systems come equipped with exactly equal amounts of system RAM, once again the total cost of both platforms comes out about equal.

Under these circumstances, the i7-920 combo is the better choice for now.

Randall,

I agree with your last post completely, but wanted to point out that the 1366 would have 4GB more memory than the 1156 (2 extra slots x 2GB, unless you are populating with 1GB chips?). That extra RAM is a HUGE thing when dealing with editing systems like the Adobe Suite. Most editing apps are very memory hungry and the more the better as far as RAM is concerned.

Craig,

I said "2GB more RAM" because many users actually purchase only one 2GB module per memory controller channel - whereas your post assumes that one would be filling up all of the memory slots on the motherboard. (In other words, the 1156 system would have 4GB of RAM while the 1366 system would have 6GB of memory under the scenario that the majority of computer users would encounter.)

I merely wanted to clarify that last post.

Most interesting thread. I too am looking at these i7 versions but I am curious regarding Sata 3 hard drives. Surely these would improve editing performance further. - Are they readily available yet ?

Over here in Oz land I was advised to-day by a supplier that they won't be available 'till towards the end of the year. I find this very surprising.

Ron C.

Video editors don't fall into that majority. We are a minority that almost always will max out the RAM available on a mobo. I'm pretty sure everyone that posts questions about a build in this subforum is asking about building an editing machine and not a word processing/email/gaming machine.

Just an update on my re-build with the i7-920 (on Intel's reference DX58SO motherboard):

When I first did the upgrade from my previous Core 2 Quad Q9450, I experienced crashes even after a complete re-install of Windows 7. The crashes came via blue screen and rebooting, and the error message was a memory dump. That told me right away that I had to replace the system memory.

The first memory I got was an OCZ Gold, which I got only because it was cheap after a rebate. But my new system crashed frequently with this memory even at stock memory clocks and default loose timings, especially when I tried to do anything that's more strenuous than simple Web surfing and e-mail with the system. That was unacceptable stock-speed stability, in my book, especially since the three memory modules checked out error-free in Memtest 86+ 4.00 when tested individually in single-channel mode. In addition, when I booted up my system with each module, I noticed that one of the modules used a different SPD (Serial Presence Detect) chip than the other two, which set the default memory clock to a different speed (in this case, the one module sets the memory clock to the correct DDR3-1066 speed that's the maximum official support on the i7-920, while each of the other two defaulted to DDR3-1333 speed simply because those two modules lacked any SPD programming data at all whatsoever for DDR3-1066 speed). I ended up not claiming the rebate at all and decided to return that memory to the MC store and replace it with a slightly more expensive Corsair kit purchased elsewhere. (MC for memory upgrades seemed bullish on OCZ, and as I had suspected even makes a commission on the sales of their largely bum memory.) If I did the exchange for the same-model kit instead, there would have been no guarantee that I would avoid such a mismatch between modules: It appears that the Gold and lower lines of their kits might not have been perfectly matched between modules; if such precision is required, I would have had to spend extra money on their Platinum line which currently consists primarily of modules which require a significantly higher voltage than the maximum that's supported by the i7 memory controller (in other words, they're only usable on a system with an on-motherboard, away-from-CPU memory controller such as the outgoing X48 chipset for Core 2 processors).

The moral of my upgrade story: Don't go cheap on the memory if you are building a system based on an i7 or i5 platform. If you absolutely must go inexpensive on the memory, better do it with memory modules rated by their manufacturer at relatively low speeds (DDR3-1333 instead of DDR3-1600) and relatively high latencies (9-9-9 instead of 7-7-7 or 8-8-8). Good low-latency memory does not come cheap - but the OCZ Gold claimed to be rated for 8-8-8-24 timings at DDR3-1600 speed at the 1.65V maximum that Intel officially supports for its i7 and i5 processors. The truth is, you can't reliably do both low-latency and low price at the same time. (For the record, the Corsair kit does 7-7-7-20 timings at DDR3-1600 speed with the same VDIMM of 1.65V and costs just $30 more than the regular price of the shoddy OCZ kit - both kits in a matched triple-channel kit of three 2GB modules, or 6GB total.)

I should have heeded the advice of the owner of a small computer shop when it comes to the choice of the brand of memory. He has experienced serious reliability issues with recent OCZ memory modules (yes, even their top-of-the-line models).

As I type this post, I am on an older computer running (you guessed it!) OCZ memory modules. But this is a 2.80C GHz Pentium 4 system with 2GB of DDR400 (PC3200) memory rated for 2-3-2-5 timings. And since this is strictly a 32-bit system, it is running a 32-bit version of Windows 7. OCZ has not made anything really good (memory-module-wise) since the last days of DDR1 memory, IMHO. In fact, only their PCP&C power-supply subsidiary (which still makes good if relatively expensive units) has kept OCZ from becoming "craptacular" in my book.

As for the DX58SO, it does allow overclocking and voltage tweaking. But the VDIMM can only be set in increments of 0.04V (the stock default VDIMM setting appears to be 1.54V). With most of the high-performance memory modules, a setting of 1.66V (the setting closest to 1.65V) is optimal.

Randall,

You upgraded from the Q9450 to the i7 920? If so, did you do a clean install on your OS drive afterwards or just move your OS drive from the Q9450 to the new system?

I'm not sure how Windows 7 deals with chipset differences (I haven't tried moving an installed OS from one chipset to a machine with another), but doing this type of move on older OS's would cause HAL issues resulting in BSOD.

Yes, I did a fresh (clean) install. My stability problems were caused by bad memory.

Currently, thanks to XMP, the CPU is running at stock while the memory is running at 1600 speed. Without XMP mode, that memory would have run at only 1066 speed with the CPU speed still at stock.

I am testing my config now with recompression of Lagarith YV12 material to 25Mbps AVC in DVDA. So far, it has saved me all of two hours of re-encoding (per two hours of actual video content) compared with the C2Q (16 hours versus 18 hours, according to my estimates) - and this is with both CPUs at stock clocks. Moreover, with the good memory now in place, it has been four hours and not a single crash whereas with the original bad memory either Windows or Vegas/DVDA/VirtualDub crashed at between one and two hours of rendering.

I have taken this 920 to as high as 3.66GHz on stock cooling at stock 1.2V CPU core voltage; however, when I set the clock to 3.8GHz, Windows crashed on boot with the bad memory. But right now, I'm at stock clocks with the good memory.

I know a few people have said that they like to run at stock speeds for stability, but I really feel that you are underserving yourself if you do not overclock the 920. I have been running at 4.2ghz (on a Noctua D12 cooler) and have not had ANY problems in Premiere CS4, and the performance jump is absolutely huge.

If you go to http://ppbm4.com/Benchmark.html you can see the performance difference I got in Premiere compared to a stock 2.66 speed on that chip. My machine (ranked second, the wildfly one) benched higher than a dual Xeon W5590 system, which is pretty cool. I too was in the "better safe than sorry" boat in regards to overclocking on past processors, but Intel really has stepped up the stability of their processors when overclocked, and it's almost expected that the 920 will be overclocked.

Good luck with it, and I hope you enjoy the new system. I'm having a blast with mine!

Thanks for the encouragement.

My render was a semi-failure, however: Although it completed the render in just over 16 hours, the result was an unacceptably flickery video. It seems as if DVDA does not properly handle lossless compressed or uncompressed AVI in any color space other than 4:4:4 RGB: When I stepped through the AVI file, some of the frames were blank. Thus, if I want to preserve the original YV12 (4:2:0) color space, I would have to somehow transcode it to a lossy compressed MPEG-2 or AVC type format with a minimum of conversions before I import it to DVDA. Unfortunately, Vegas and most other NLEs convert YV12 content to RGB as the video is being imported into the NLE's timeline, then re-convert the color space to whatever is specified in the final format. This can adversely affect color reproduction quality.

Right now I'm still on the stock boxed processor cooling. Unfortunately, with the stock cooler, I could not get the processor to operate reliably at 3.8GHz at stock voltage (1.2V); Windows crashed on boot-up at those settings. I had to bump up the core voltage to Intel's recommended maximum of 1.35V before I could boot into Windows reliably, and then I stopped my overclock at 4.0GHz. I do not want to risk destroying a $290 (USD) processor, especially in the hotter months of summer.

I am currently running the 64-bit version of Prime95 on that system at 4.0 GHz. If it passes the Prime95 stress testing for several hours, then I might still have more overclocking headroom. But if it crashes in less than an hour, then I have taken the overclock a bit too far.

UPDATE #1: Looks like I had taken the overclock a little too far on that stock cooler. Prime95 testing showed half the workers stopped working at 4.0GHz. I am retesting the system at a more conservative 3.8GHz, and report my results later.

UPDATE #2: At 3.8GHz with 1.35V VCore, it was stable for an hour (no errors; however, I stopped the test after an hour due to heat issues, as mentioned below) in Prime95 - but the core temps as measured in RealTemp were right at the maximum. So I backed it off a bit, and then experimented with overclocking at the stock 1.2V VCore. At that VCore setting I could only muster about 3.5GHz before Prime95 reported errors.

Based on these results, assuming that I have the turbo switched on I feel safe with an overclock to only 3.44GHz at the higher 1.35V VCore. At the lower 1.2V VCore, I might have to settle for an overclock to only 3.2GHz for best reliability. (And these results are with the stock boxed Intel CPU cooler.) If I decide to use 3.5GHz @ 1.2V or 3.66GHz @ 1.35V, I would have to disable the Turbo feature for best stability.

The conclusion in my case is it's the stock Intel boxed CPU cooler that's holding it back. But these overclocking results were actually quite good without having to spend a single penny more for an optional high-performance cooler.

Have you tried bumping up to 1.25 or something incrementally instead of heading right to 1.35? I don't have much experience with that Intel board. Almost everyone has been throwing the 920 onto the Asus P6T platform (one variation or another).

Have you seen the overclocking the 920 walkthrough that Harm Millaard found and posted?

Adobe Forums: Overclocking the i7, a beginners guide

I read that. As it is (at 3.5GHz and 1.20V with the stock Intel boxed CPU cooler), I'm already getting temps in the mid-70s C at 100% load - and any raising of the core voltage (VCore) would raise the temps further even at lower core clocks. Intel recommends a maximum TJunction (core temperature) of 70°C under continuous conditions.

I am also checking the memory settings. It seems like I had the QPI/Uncore voltage (VTT) also set at 1.35V (as the XMP memory profile dictated). I will reset everything back to stock for about half an hour until the CPU has cooled down sufficiently, and then redo my overclocks.

UPDATE: Upon further checking of that guide it seems like 3.5GHz is about the highest overclocked speed that I can safely go to with the stock Intel cooler. This is because the core temps at that speed get to just above 70°C at 100% load. (Speeds higher than 3.5GHz with the stock cooler are stable under Prime95, but causes the core temps to rise to potentially harmful-to-components levels - above 80°C - under anywhere near a full load.) In other words, the Intel board is very capable but the stock CPU cooler is holding it back (in the case of my particular system).

Gosh, I need a better CPU cooler before I can overclock any further than 3.5GHz...

As an addendum, the DX58SO is currently the only "consumer" Intel-branded motherboard for Socket 1366. The other Intel-branded Socket 1366 motherboards are sold for workstation and small-server use and may provide incomplete or no tweaking capabilities. The DX58SO is part of the "Extreme Series" of Intel-branded motherboards, which also includes a standard ATX and a microATX motherboard for Socket 1156 with the company's P55 chipset (however, I have been unable to locate either of those other two motherboards which tipped the scales towards 1366 for me). Intel also distributes various lower lines of motherboards including a microATX and a standard ATX Socket 1156 P55 motherboard in the Media Series. Unfortunately, none of the Media Series, Classic Series or Essential Series lines of Intel-branded motherboards provide any overclocking capabilities at all whatsoever, and most motherboards in those three lines provide minimal (if any) memory or voltage tweaks.

And the biggest reason why I picked Intel over Asus or GigaByte is the LAN. The boards from Asus and GigaByte people have talked about all use a Realtek LAN chip for their primary LAN ports (and I have had nothing but bad luck with Realtek LAN controllers in the past, such as very slow transfer speeds and recalcitrant wired connection reliability, which made me very leery about buying anything with a Realtek LAN controller to this day). The lower-end Intel-branded motherboards also use Realtek LAN controllers - but the Media and Extreme Series lines of Intel boards use the company's own (IMHO superior) LAN controllers. (Speaking of Asus, the last board I purchased from that brand had an Attansic L1 LAN controller chip, which had compatibility issues of its own, and also has especially poor driver support since Attansic was bought out by Atheros back in 2007.)

If you decide to drop money on an aftermarket cooler, my strong suggestion is the Noctua NH-D14. It's absolutely amazing. Make sure to use their provided thermal paste, as it is better than Arctic Silver's offerings when paired with this particular cooler (my own tested experience). Just make sure you have enough room in your case for this one. I have everything in a 4U height rackmount and had to dremel the top of my case to get it to fit right.

Craig,

I did some additional experimenting with my system. It turned out that an overclock to 3.5GHz with the Vcore and VTT voltages only turned up slightly to 1.2V from the stock 1.15V and the FSB:DRAM ratio remaining at 2:8 (this results in my memory now running at an effective 1400MHz instead of the stock 1066MHz) is the best compromise between performance and stability/heat (with the stock cooler). I did short Prime95 runs while also running TrueTemp with all three combinations - stock CPU speed and stock 1066 memory speed and timings at stock voltages, stock CPU speed but using XMP-1600 memory settings and the aforementioned overclock mentioned above. It turned out that the overclocked-to-3.5GHz config ran only 2°C higher on average at 100% load (in terms of the average of the temperatures of all four CPU cores) than the stock setup with stock memory! In fact, running a stock-speed CPU with overclocked XMP memory actually produced higher temps than the overclocked setup. 3.6GHz booted into Windows at these same voltage settings, but Prime95 testing throttled back one core - this meant that I would have needed to raise the Vcore and VTT a little higher than the 1.2V that I used at 3.5GHz in order to get things stable at 3.6GHz (which, in turn, would have raised the CPU core temps even higher, necessitating the use of a higher-performance CPU cooler for such a level of overclocking).

As for the actual Blu-Ray re-encoding from XDCAM EX MXF to AVC in DVD Architect, re-encoding 2 hours worth of content took nearly 4 hours longer with the stock-speed setup than with the overclocked setup - 18+ hours versus just over 14 hours. (The stock-speed performance is nearly as slow as my old Q9450 with 4GB of DDR2-800 memory could muster; that older setup now resides in my auxiliary system which is currently in use for some of my standard-definition video work.) And yet the total CPU utilization barely reached 25% even with a stock 920.

I also looked at the table of rendering performance with the different systems. A configuration that's closest to that of my old Q9450 setup did respectably well in the tests (a Q9550 on a non-overclockable Intel-brand DDR2 motherboard with a G45 chipset that's similar in performance to the P35 chipset Intel-brand motherboard which my Q9450 resides on), in fact beating a few of the stock i7-920 configurations. And an Asus motherboard is no guarantee of superior performance; other components such as multi-disk RAID setups also factor into the overall performance picture. Most (though not all) of the very top-performing systems used AMD/ATi HD4870 graphics cards (I have an HD4850, one step below the 4870, in my overclocked 920 system).

Have you ran a speed test on your drives? You should be getting much higher than 25% utilization during encoding I would believe. I am not a DVD Architect user though, so I don't know how that software utilizes the processor.

As far as Asus/superior performance subject, I believe that Asus delivers a product that is very easy to configure to get the maximum performance and they are very reliable. Yes, a motherboard does not make or break system performance. Bottlenecks can occur in many different areas, drives being of critical concern.

And video cards: I've built an i7 965 machine with the 4870, and it was good. I like my GeForce 285 2GB better though =) Especially when I factor in the Mercury Playback Engine by Adobe, which will not be supporting ATI graphics cards.

Maybe this is why I was not seeing as much of an improvement of a performance as one would expect: Almost all of the other systems on that list have at least one additional drive outside of the internal Windows boot drive. (Sadly, consumer external hard drive kits are ill-suited to being the project disk due to their not only being slower than internal drives in external drive enclosures, but also due to their power management features which often cannot be disabled at all.) My times were based on just using the same Windows boot drive for both my project and the system files. Adding enough drives to make a substantial difference would have cost me far more money than I can currently afford - at least $1,200 worth of drives plus another $200 worth of add-in controllers. And $1,400 is a lot to spend just to get a relatively small amount of performance gain with everything else.

Also, the Vegas' and DVDA's Sony AVC rendering engine is relatively old (even in the most recently released versions, Vegas 9.0c and DVDA 5.0b): It still cannot take full advantage of the new features in the i7 processors; after all, it was designed during a time when the fastest consumer processor available was the Core2 Extreme QX6850. This is why a stock-clocked i7-920 came out as slow as a stock-clocked Core2 Quad Q9450 in the rendering tests (both processors run at a stock 2.66GHz). After all, Vegas and DVDA are better at MPEG-2 encoding than AVC encoding (performance-wise, even though lower-bitrate HD MPEG-2 renders produce lower image quality than AVC renders of half the given MPEG-2 bitrate). Plus, even the 64-bit versions of Vegas are still not fully 64-bit; in fact, all of the plug-ins which come with Vegas are still 32-bit.

That said, I would like to have at least one or two other drives added to my system. Unfortunately, the Intel ICH10R's built-in RAID feature requires that all hard drives connected to the SATA ports be RAIDed together - it's all RAID or no RAID. Although the improvement in DVDA recompression times are in line with what a mere increase in the processor's clock speed expected, the times are still quite a bit longer than it should have been due to the use of a single hard drive for everything.

By the way, you can still run CS5 on a system with a graphics card which does not support the Mercury Playback engine. (This includes all ATi graphics cards and all older and lower-end NVIDIA graphics cards.) It just means that the Mercury engine will not run with an unsupported graphics card, and the system will thus default to the slower standard Adobe playback engine. In addition, Adobe stated that 12GB is the minimum amount of RAM required in order to run CS5 with the Mercury Playback Engine; it will not run at all with less than that much RAM. And as Craig pointed out in another thread the Mercury Playback Engine will only work with an NVIDIA graphics card with hardware-based CUDA acceleration (so far, only the very high end of the GeForce GTX2xx line plus higher-end versions of the latest edition of the Quadro FX line support hardware-based CUDA acceleration although all GeForce and Quadro FX cards based on a G80 series or later GPU support software emulation of CUDA).

And given my current equipment even after the i7-920 upgrade, I will have to spend at least another $1,500 worth of hardware upgrades plus another $800 or so on Premiere CS5 just to even run CS5 in its full glory. This is because my motherboard effectively has only three DIMM slots; thus, I would need three 4GB memory modules in order to run CS5. (Technically, the motherboard has four DIMM slots - but using all four slots will force part of the memory to run in single-channel mode instead of triple-channel.)

In your bios set your SATA controller to non-RAID mode. Add two identical drives (even 500 GB drives are fine, and you could get two of them for just over $100). Then use Windows to create a RAID 0. It will work wonders for you.

Craig,

I did not even think about that. The separate software RAID will speed up whatever projects that I have running through my current NLEs.

Does it matter which RPM speed the two drives spin at?