New i5 Chip Improves Things Considerably: Stable 3.44 Ghz With Stock Cooler

My previous i5-750 processor turned out to be defective. I sent the chip back to NewEgg and got a prompt replacement. The new processor works much better and well within Intel's specs.

With a properly working processor, it is much easier to see the typical i5/i7 behavior when overclocking. It makes an interesting comparison with AMD. Both Intel and AMD processors are unstable when the voltage is too low. This can be anything from a single thread halting with an error under Prime95 to a hard crash. However, at the top, there is a distinctly different behavior. AMD processors tend to become unstable at some temperature threshold and exhibit very similar behavior to undervolting. The variability is the same, you can see one thread halt or the screen turn to multi-colored gibberish or just go black with a hard crash. Intel processors however don't do this. The problem for Intel is thermal throttling when the chip overheats. Unlike AMD, there is no outward sign that anything has happened. The clock drops to a much lower frequency but the chip just keeps chugging along with complete stabillity. However, the fundamental rules are the same: increased clock requires more voltage to remain stable. More voltage means more heat and eventually both types of processor will reach a point when the voltage is not enough to remain stable but more voltage is too hot.

However, Intel's thermal specs are somewhere between cryptic and endecypherable. If my understanding of these specs has improved then the maximum temperature Intel gives would be the center top of the heat spreader (which we have no way to measure without adding our own probe). Secondly, this spec only defines Intel's guarantee for normal operation within the rated wattage. I suppose this might be useful for an OEM but it doesn't help us much for overclocking. The only real temperature we know for certain is when the chip reaches its maximum Tj and thermal throttles. However, the indicated temperatures of when this happens seem to vary from 5 C low for the stock cooler to 30 C or more low with a good third party cooler. Indicated temperature aside though, for all practical purposes, we can use thermal throttling the an upper limit much the same as an AMD error or crash.

What is Stock?
The first question that usually comes up is what a given chip can do at stock settings. However, one needs to be aware that stock is not exactly defined. Suppose we use 1.225 volts VCORE and 1.1 volts IMC as stock. That is fine; I can set both of these manually in the BIOS. So, I save and exit and then re-enter the BIOS. I bring up the Hardware Monitor and find out that the indicated voltage is not what I set. Likewise, I can boot up Windows and check the voltage with CPU-Z. And, I find out that it says something different from both the BIOS setting and BIOS Hardware Monitor. I'm sure one could argue that one of these is more accurate than the other. However, I chose to use the BIOS setting since it is the only thing I can control. What this means is that when I set the BIOS to 1.225 volts this may be a different voltage than what would be available on a different manufacturer's motheboard or even a different model within the same manufacturer. Nor is there any indication that the Auto setting is a common stock value. Perhaps it is most of the time, but I know that my Asus 785 and 79x boards use different stock voltages. Also, what you test with will change the results. Being able to run SuperPi is no real indication of stability. The same system can boot Windows, complete a SuperPi run, and then crash a few minutes or few hours later with no additional stress. My indications are that:

Prime95 Small FFT < Prime95 Large FFT < OCCT Linpack

In other words, I've had settings that passed Prime95 Small FFT but got errors under Large FFT. And, I've had settings that passed Prime95 Large FFT but got errors under OCCT Linpack. However, I've never had the reverse. I've never a setting pass OCCT Linpack but get errors under Prime95 or pass Prime95 Large FFT but get errors under Small FFT. From what I've seen, this relative stress is the same with AMD and Intel. For example, with Prime95 my Phenom II 965 can pass Prime95 Large FFT at 1.4 volts at 3.8 Ghz. However, my Phenom II 965 cannot pass OCCT Linpack at 3.8 Ghz and 1.4 volts. It needs 1.42 volts to be stable. So, I consider OCCT Linpack to be the definitive test. If it can't pass Linpack then I don't consider it to be stable.

What is Stable?
Once you've tested with Linpack and found the point where it passes but increasing clock will either cause thermal throttling with Intel or thermal instability with AMD then where do you go? I have never considered these points to be usable. In other words, I require some margin between where I will run a system and where it has reached its limits. For example, if the system passed Linpack at 180 Mhz x 20 and got errors at 181 Mhz x 20 then I wouldn't run at 180. I would reduce clock two steps and run at 178 Mhz x 20.

We also need some similar margin with Intel's thermal throttling. My preference would be 5 C. In other words, if a given setting passes Linpack without throttling but higher settings do throttle then I would want to reduce clock and voltage enough to drop the temperature 5 C. This should be enough margin to ensure that with changing room temperature, thermal throttling is never an issue. However, this 5 C number can be a problem. If thermal throttling occurs around 95 C then reducing 5 C seems reasonable. However, using a better cooler can drop the indicated throttling temperature down to the 60's. With this situation, dropping 5 C is not the same. The best I could do would be to try to calculate the same ratio. My room temperature is 23 C so if it throttles at 95 C then that is a 72 C range. That is about 14.5x the 5 C reduction. So, I would probably take the new indicated temperature, for example 65 C, and keep the same ratio. 65 C - 23 C = 42. So, I would only reduce 3 C indicated. These margins are the best guestimates I can come up with; if something turns up later to change these then I'll adjust them accordingly, but this is probably a fairly good starting point.

This is the defective i5-750 that I sent back.

And, this is the new one.

I'm going to do the testing in two parts: first with the stock cooler and then with the Freezer 7 Pro. My case has exceptional cooling with its two 200mm fans. So, I decided that to give a more typical result I would unplug the top 200mm fan for the stock testing. With the rear 120mm case fan and front 200mm case fan running, this should be closer to a typical case with three 120mm fans. When I switch to the Freezer 7, I'll plug the top fan back in. I put the side panels back to close up the case (which is how most cases would be). I also decided on a low tech solution to check temperatures inside the case. I have a clock that gives the room temperature but also includes an outside temperature unit. So, I simply set the outside unit on top of the HD 5770 card which is just below the i5 cooling fan.

This is the outside unit setting on top of the video card.

So here the Indoor temperature is room temperature and Outdoor temperature is the temperature inside the case. The screenshot is convenient because it also documents the time and date. However, even with the top fan turned off my case might still be cooler than average.

My case has the vents on top for the top fan.

And, the left side panel is well ventilated.

Aside from the case ventilation there is also the question of CPU fan setting. My BIOS allows setting for a slower, quieter fan or a faster, noisier fan. I decided to leave the setting at Auto since I don't think all BIOS's have this range. We know that at some point, the stock cooler will be inadequate and the highest fan setting might allow just a little more clock but I don't think this distinction is worth testing since the third party cooler is well above this.

My Results So Far
I'll keep adding results here until I reach thermal throttling with the stock cooler. I would guess that thermal throttling probably won't occur until 3.45 - 3.5 Ghz.

1.225 volts VCORE, 1.1 volts IMC BIOS setting
148 Mhz x 20 = 2.96 Ghz - Linpack stable

1.25 volts VCORE, 1.15 volts IMC BIOS setting
168 Mhz x 20 = 3.360 Ghz - Linpack stable

This is a typical screenshot with the 1.25/1.1 volt test at 3.36 Ghz showing the indicated temperature running about 81 C.

And, this is the clock at the same time showing the case at about 16 C above room temperature.

And, this is the screenshot of the completed Linpack test.

My i5-750 completed Linpack at 1.28125 volts VCORE, 1.20 volts IMC at 3.42 Ghz.

And 3.48 Ghz at 1.30625 volts VCORE and 1.25 volts IMC.

And, finally we find the top:

12/17/09 19:04:15 9.000 0.000 9.000 100.00

We've clearly found the ceiling with the stock cooler. At 1.34 volts VCORE, 1.3 volts IMC at 3.56 Ghz everything shows thermal throttling. With the OCCT graph reading 93 C max we will have to bring it down to 88 C to have a 5 C cushion.

And here is the clock shot showing the temperature inside the case about 20 C higher than room temperature.

I have the OCCT temperature graph handy from the previous 3.48 Ghz speed.

However, this graph touches 90 C. So, I had to come down one step to 20 x 173 Mhz = 3.46 Ghz to pass Linpack with 88 C. Next is the issue of stability. Normally I would drop down 2 steps to 171 Mhz while leaving the voltage at 1.30625 VCORE and 1.25 volts IMC to ensure stability. However, I've already established that this voltage is stable for 174 Mhz. So, using this number we drop two steps to 172 Mhz. And, this gives us our official stock cooler testing speed: 20 x 172 Mhz = 3.44 Ghz. This is a 29% overclock with the stock cooler. Not bad at all.

With Turbo turned on, using the stock cooler, the maximum clock is 165 Mhz which gives about the same speed.

Turbo Off : 20 x 173 Mhz = 3.46 Ghz
Turbo On: 21 x 165 Mhz = 3.465 Ghz

Now it's time to reinstall the Freezer 7 Pro.

My Impressions, So Far

I sent my i5-750 processor back to NewEgg last Saturday. I sent it Priority Mail with Insurance and Confirmation. It arrived in Whittier, CA and was accepted by NewEgg yesterday. Hopefully, I'll have a new one in about three days.

Since none of the test data for the i5 was worth anything, I'll just go over my general impressions so far.

Processor and heatsink retention.
AMD processors have pins whereas with Intel the pins are in CPU socket. I prefer the AMD arrangement. An AMD processor should just drop into the socket by gravity. If it doesn't drop in then you probably have a bent pin. I know this well because one of the pins on my PII 965 was bent and I had to straighten it before it would drop in. With the Intel arrangement, half the pins could be bent and you would never know. So, for detecting bent pins, AMD wins.

The second item is the way that the processor is retained in the socket. AMD uses a latching mechanism that grips the pins whereas Intel uses a clamping mechanism that holds the processor down. Here, Intel clearly wins. The clamping mechanism is all metal parts with a metal backing plate; there is no way that this processor is budging. AMD however is a bit worse off. The pin gripping mechanism is not designed to handle pulling loads. Simply put, you must twist the heatsink to break the seal with the processor before you try to remove it. If you just tug on the heatsink, you could end up leaving pins behind in the socket. And, this would almost certainly destroy both processor and motherboard. With Intel, it should be nearly impossible to damage anything no matter how tight the thermal compound is stuck. Be careful with the AMD heatsink and you should be okay but Intel is the clear winner.

Next is the heatsink retention mechanism. I've had both experiences with AMD's. For example, the X3 720 stock heatsink was great; I just put it on and clamped it down. However, the heatsink for the X4 965 was not so nice. The metal loops were too short and I could barely get it on. There was no slack to tighten the clamp. And, you cannot be ham fisted with the clamp because the cam is just made of plastic; I'm sure it will break if you try to force it. The metal loops for the Freezer 64 Pro were just as good as the X3 stock unit; it just went on and clamped into place without any trouble. AMD's mechanism works the same whether it is third party or stock and the retention frame uses a metal plate on the back of the motherboard for extra strength much as Intel does with the CPU clamp.

Intel however seems to have more than one way to hold the heatsink on. The stock unit uses pins that push down through the four holes surrounding the rentention frame. You push opposite diagonals at the same time. I unlocked the pins, then pushed the pins through and twisted the knob on top of each leg to clamp the pin in place. Apparently, however, you can just push the knobs down while in the locked position. This is maybe a tiny bit easier than the way I did it. Either way though, you want to make sure that each knob is down, locked, and the pin firmly attached to the motherboard. The Freezer 7 uses a completely different method. it has a plastic frame that goes around the CPU clamp frame. You have to push plastic pins through the holes in the frame and through the holes in the motherboard. Then you push expander plugs into each pin to keep them from coming loose. This operation is much more than a pain; it is sloppy and includes far too many ways that you could break or damage the plastic parts. I had to push the retaining plugs in (and remove them later) with needle nosed pliars. And, they take a fair amount of force to push in and pull out. I don't know if these are all similar with third party coolers on Intel motherboards but the Freezer 7 requires patience and a fair amount of care before you apply force. Then, once you have the plastic frame mounted you attach the actual heatsink bracket with two screws. You have to remove the fan to have room to insert the screw on that side. And, fiddling with the other screw in the cramped space between the heatsink and rear case fan is also not much fun.

With AMD, as long as you have a third party cooler that uses the stock latching mechanism then the AMD system is much easier to change processors. With Intel, at least with the Freezer 7, the heatsink bracket blocks the CPU clamp so you cannot change processors without first removing the plastic bracket. This is a major pain. Still, most people probably wouldn't be swapping processors very often so it's probably a hassle that one can live with. I think both of the systems can work with a 500 gram cooler but I wouldn't trust either one of them with anything heavier. Both designs have shortcomings which could result in damaged parts if you aren't careful. However, both systems can (and I'm sure they do) work most of the time.

Changing Clock Speed
I've only worked with Asus motherboards. It is interesting though how different the flavor of the BIOS is from AMD to Intel. My 785 board has few options even though my X3 is Black Edition which allows easy overclocking. The BIOS works fairly well with the X3 just in terms of overclocking. There are items that are mislabeled in the BIOS, for example, calling something a clock setting when it is actually a multiplier setting. And, I have to say that I don't understand why the Intel BIOS nicely calculates memory speed based on actual base clock whereas the AMD memory speed always uses a base clock of 200 Mhz for its calculations. Someone is clearly dropping the ball here.

However, with the X4, it was another story; it wasn't until I realized that the BIOS was setting the voltage too high and reduced it that I could get it stable at higher settings. This amounted to a bump of at least 100 Mhz. The 79X board is a different flavor. It has more settings and is a little easier to use. However, it seems to set the default voltage lower than on the 785 board. I couldn't get 3.9 Ghz stable running OCCT Linpack without bumping the voltage to 1.425 (when 1.4 volts is AMD's max for this chip). The 785 BIOS is adequate but, when using the PII 965, I like the 79X BIOS a little better. I would guess that this BIOS is more like that of the older 78FX boards.

The Intel BIOS on my P55 motherboard is a bit more polished. As I've already mentioned it correctly displays memory speed even when you change the base clock from 133 Mhz. The disappointments come more from the processor itself. Since regular Intel processors have a locked multiplier they aren't nearly as easy to overclock as AMD processors. You really have to stretch that base clock setting to hit anything substantial (at least with the i5-750 which only has a base clock of 2.66 Ghz). For example, to hit 3.32 Ghz you have to bump the clock from 133 to 166 Mhz. This wouldn't be so bad if it weren't for memory. For example, with the base clock at 166 Mhz you can set the memory multiplier the same as for 1066 Mhz and it will actually be at 1333 Mhz. Likewise you would hit the same 1333 Mhz speed again with a base clock of 222 Mhz if you set for 800 Mhz. Of course, with a standard multiplier of 20, this would be 4.44 Ghz on the processor which is out of the question. However, since Intel has no problem with reducing the multiplier you could just set this to, say, 16 or 17, and get an accessible processor clock. It does work, just not as well as with Black Edition. Naturally, not all of AMD's processors are Black Edition but Intel has nothing remotely like this below the outrageously priced i7-965. Maybe Intel thinks that Turbo makes up for some of what is lacking but I would rather see unlocked multipliers. Finally, I'll mention the x21 multiplier because I've seen this mentioned in several reviews. My BIOS can be set to a multiplier of 21 but is doesn't work; the highest multiplier you can actually set is 20 (the same as the default).

Overall, AMD seems a little easier because of the unlocked multiplier, but if the Intel chipset and processor are able to tolerate base clocks way over the default 133 Mhz, Intel could still end up the winner. It will be interesting to find out.

The More, The Merrier -- Nine More Tests Showing The Chip Is Indeed Defective

After a suggestion from rge (who was involved with RealTemp), I decided I would try a number of variations on the tests. The results however are all the same as my earlier tests.

I ran tests with a number of variations: with OCCT, both the CPU test (which is basically Prime95) and Linpack (which is similar to Intel's Burn Test). If you have only 4 GB's and use just a fraction of the memory the test will cycle quickly. However, I have 8 GB's and used 90% of the memory so the cycle time is about seven minutes. I ran Prime95 itself both with large and small FFT's. I also ran i7 Turbo logger at the same time. Some of these screenshots are rather large. My desktop is 1920 x 1080 and I used most of it in the later screenshots to capture everything. Again, just click on the pictures to view them full size. I also included the data from the i7 Turbo log. It shows the data, time, three clock multipliers and the percent CPU load.

12/04/09 15:43:10 20.000 0.001 20.000 1.55
Everything here is normal. The i7 Turbo data agrees with the screenshot.

12/04/09 15:46:44 9.000 0.000 9.000 100.00
Here we have throttling with OCCT Linpack. The multipliers have dropped to the minimum value even though we have 100% CPU load.

At 3:43:10pm which is 4 seconds into the test, the i7 Turbo data agrees with the OCCT graph which agrees with the Windows CPU Meter and the live display of i7 Turbo.

Likewise at 3:46:44pm which is 3 minutes 38 seconds into the test, everything shows the same thing: throttling at 100% load.

12/04/09 15:48:20 20.000 0.000 20.000 2.09
Everything normal.

12/04/09 15:51:22 9.000 0.000 9.000 100.00

Throttling with 100% load running OCCT CPU Test.

12/04/09 15:53:04 20.000 0.000 20.000 100.00
Everything normal.

12/04/09 15:54:09 9.000 0.000 9.000 100.00
Thermal throttling with Prime95 running large FFT.

12/04/09 16:06:09 9.000 0.000 9.000 100.00
Some people seem to prefer small FFT; here is Prime95 with small FFT.
And, everything shows throttling.

12/04/09 16:14:44 9.000 0.000 9.000 100.00

Again, everything agrees. Throttling with 100% load.

12/04/09 16:17:42 9.000 0.000 9.000 100.00
And, in the last screenshot everything including the i7 Turbo data agrees: we have thermal throttling.

Bad Chip Test Results

After more testing I've come to the conclusion that I have a bad chip.

These are screenshots from my last test which I think is pretty conclusive. You can click on the images to see the full size.

This is the beginning of the test. Everything is normal here. The voltage and clock frequency are at normal values. This should be an easy test for the i5-750. I also included the Windows CPU Meter, Clock, and Calendar gadgets for extra verification. Note that the CPU meter agrees with OCCT showing both 100% CPU usage and 90% memory usage.

This period is the gap between test cycles. Note that CPU usage has dropped to 25%. Note that this occurs at 10 minutes and 16 seconds into the test.

The most important thing to note here is that the CPU clock has dropped to 1.2 Ghz but CPU activity is still 100%. This is a clear indication of thermal throttling. Note that this occurs at 12 minutes and 19 seconds into the test.

This is the OCCT graph. You can see the large drop at 10 minutes which corresponds to the second screenshot. The smaller dip at 12 minutes is the thermal throttling captured in the third screenshot. You can also see that thermal throttling had been occuring before the 10 minute period between tests.

This chip is definitely bad.

Intel i5-750, Bad Chip Nixes Tests

I don't think these tests were representative; I think the chip is bad.

Below is the original article. I'll have to do a new one after I get the replacement chip.


******************************

I could not get the processor to pass an OCCT Linpack test at 2.9 Ghz. The problem is that the processor heats up quickly to over 72.7 C which is the chip limit. But, when you reduce voltage, the procesor becomes unstable. At 2.9 Ghz the processor gets errors with 1.16 volts. So I tried increasing to 1.18 volts VCORE and cutting the clock down to 2.8 Ghz. The processor is stable at 2.8 Ghz but heats to 75 C. Frankly, I could not get the processor to stay within range higher than 2.66 Ghz with Turbo turned off. I did some screen captures to show this.

The voltage indicated by OCCT on the following images will be slightly less since it drops during testing. When idle, the voltage shows correctly as 1.18 volts.

The maximum temperature for this processor is 72.7 C. With Turbo off, it just stays within the range.

However, with Turbo enabled the processor increases clock and heats up to 77 C which exceeds the maximum temperature. The thing that bothers me is that I'm seeing it exceed its thermal limits with the voltage set at 1.18. With the voltage at the normal 1.25 volts this would be even worse.

We now have our baseline. With the stock cooler, the i5-750 cannot be considered to be overclockable or even within limits with Turbo enabled. This means that the vast majority of people who buy systems with this processor have a marginal configuration at best. Intel pulled the same stunt with its original Kentsfield quads which were also out of range at stock speeds. Anand Lal Shimpi was well aware of this but in true overclocker fashion he simply averted his eyes and pretended it didn't exist. I had hoped the problem had been fixed with the G0 stepping and the lower voltage 45nm Penryns but now we have Nehalem and the problem is back.

There may be some Intel fanatics who will think I'm being unfair to Intel but I tested my X3 720 and PII 965 the same way. The X3 720 is fine with the stock cooler while the PII 965 needs improvement. However, it should be noted that the PII 965 in stock configuration is more than adequate and could even tolerate a mild overclock of maybe 100 or 200 Mhz. So, in stock configuration AMD does what it is supposed to while Intel fails. The next step is to bolt on the heftier Freezer 7 Pro cooler and see what happens.

Intel i5-750, First Look

I now have my Intel i5-750 system up and running. This is the progress so far.

I have three very good processors to compare. I started with a three cored AMD X3 720 in my CoolerMaster Storm Sniper case with an Asus motherboard with AMD 785 chipset and 8 GB's of OCZ AMD gold edition memory (4 x 2 GB's). I never bothered changing to a third party cooler because the stock unit was adequate. I ran it with the integrated graphics which were not too bad but underpowered for games like Dawn of War. The thermal properties of this processor are good enough that I should be able to run it just fine in a small, tighter case with a moderate video card. I have handy a smaller case with neither front nor top fan (only the rear fan) and an HD 4650 video card. However, I do have an upgrade ready for the rear case fan if necessary. I'll probably end up running Linux on this system. The processor is rated at 95 watts TDP.

After the X3 720, I replaced the motherboard with an Asus board with AMD 79X chipset and 8 GB's of OCZ AMD Black Edition memory (4 x 2 GB's). The video was upgraded to an HD 5770 and the processor upgraded to Phenom II X4 965. This processor is the C3 stepping and is rated at 125 watts TDP. I also preferred a third party, Freezer 64 Pro cooler over the stock cooler. The stock cooler, in spite of the copper bottom and heat pipes, still blows hot air down on the board.

I have a second CoolerMaster case, a HAF 922. This has very similar front and top 200mm cooling fans with the same rear 120mm case fan. The box chasis is identical with the same drive bays in front. The power supply is a bit different. This system has a 700 watt CoolerMaster PS compared with the 850 watt Corsair PS in the Sniper case. However, neither power supply contributes to case heating or cooling because they both draw air from the case bottom and exhaust it out the back. My i5-750 is also rated at 95 watts TDP. I suppose the first question people might have is if it draws the same amount of power as the X3 720. No, it doesn't. Even at maxium overclock, the X3 720 barely throws any heat while the PII 965 throws a lot. The i5-750 is in between. It throws a great deal of heat at 3.4 Ghz, much more than the X3 720 does. If I were to make a rough estimate, I'd say the i5-750 must pull around 115 watts. This would put it a little closer to the PII 965 than the X3 720.

I've heard people say over and over that the Intel cooler is poor. It is actually quite a bit better than the stock X3 720 unit which has a smaller fan and no copper core. If the i5-750 really did draw as little power as the X3 720 then the heatsink would barely get warm. But this is not the case. Overclocked to even 3.4 Ghz, the heatsink blows hot air on the motherboard much more like my PII 965 than the X3 720. I'm sure there will be some skeptics who've read some whiz bang review somewhere that claims that is has accurately measured the CPU power draw at some incredibly low figure. Not a chance. The i5-750 undervolted to 1.12 volts still throws more heat at 3.4 Ghz than an X3 720 does at 1.325 volts and the same 3.4 Ghz. The X3 720 and i5-750 are absolutely not in the same thermal class.

So, Intel fudges the numbers? Yes. Does that bother me? No. Keep in mind that I'm not comparing the quad core i5-750 to the X3 720. That wouldn't make any sense because the X3 720 is much cheaper. No, I'm comparing the quad core, i5-750 to the similarly priced, quad core, Phenom II X4 965. The PII 965 is certainly not light on heat and, if the i5-750 can do the same amount of work, it doesn't have to be either. I have a very similar Freezer 7 Pro ready to install so I can at least stop the hot air blowing on the board. I'll add an update after I see how that works. You can see the pictures of the coolers below in my last article.

Note. I haven't switched to the Freezer 7 Pro cooler yet because I'm still trying to establish the baseline with the stock cooler. I have to admit that this is difficult for me. I didn't have any trouble pushing the X3 720 and PII 965 to their limits. And, when I knew what the limits were I accepted them. I ended up with 3.8 Ghz on the PII 965 and 3.4 Ghz on the X3 720. But, I tend to handle the i5-750 with kid gloves. It isn't easy being tough on i5 because doing thermal testing on an Intel Nehalem is like watching an eight year old trying to keep up with teenagers. Simply put, when I started holding the i5-750 to the same standards I used for my AMD processors, the i5 failed. So, I lowered the clock and tried again. I started at 3.6 Ghz then dropped to: 3.5, 3.4, 3.32, 3.154, and 3.078 Ghz without success. I was hoping it would pass at 2.988 Ghz but that just failed with an error. So, I'm trying again at 2.9 Ghz.

A First Look At AMD And Intel Stock Coolers

This is rundown of the stock coolers for the Intel i5 and AMD Phenom II processors that I have along with third party coolers.

Click to enlarge any of the following images.

On the left is the stock cooler for the Intel i5-750. On the right is the stock cooler for the AMD Phenom II X3 720. The Intel cooler weighs 250 grams while the AMD cooler weighs 280 grams.

The Intel cooler has a larger fan.

Both heatsinks are solid aluminum. You can see that that the AMD heatsink is taller.

On the left the AMD heatsink has the thermal paste removed down to bare metal. AMD has a solid aluminum heatsink while the Intel heatsink obviously has a copper core. Both processors are rated at 95 watts. The Intel processor probably draws a little more but the copper core and larger fan would take care of that.

The X3 720 cooler is on the left while on the right is the cooler for the AMD Phenom II X4 965. You can see that both fans are the same size. However, the X4 965 fan has deeper blades which probably help move air against higher back pressure caused by the closer spaceing on the cooling fins. The X4 965 cooler weighs 355 grams.

Here you can see that while the X3 720 heatsink is solid, the X4 965 heatsink uses two heatpipes on each side. This is necessary to cope with the greater 125 watt rating.

Update: I was not entirely happy with the stock HSF for PII 965. It is adequate at the stock speed of 3.4 Ghz and probably 3.6 Ghz. However, it does tend to blow hot air on both the motherboard chipset and the memory. And, with my particular case arrangement I have a large, 200mm fan right above the cooler that draws upward. However, because the fins run left and right this is no benefit.

Here you can see that the base of the X4 965 heatsink is all copper. On the right is a third party cooler. This is the Arctic Cooling Freezer 64 Pro. It weighs 530 grams. It has a copper base and three heatpipes on each side. The heatsink is obviously much larger. It also has the standard latching mechanism used on socket AM2/AM2+/AM3.

Update: The Freezer 64 seems much better. I was very concerned that the fan would not clear the DIMMs on the right side of the cpu socket. You can see this in the above picture. The problem with this is that it would have required me to turn the cooler around and then the fan would be blowing backwards. However, it does indeed clear the DIMMs.

Most people wouldn't have the very large fan right above the cooler; but you can see that even with the boxing effect of the video card, the air is directed properly toward the rear case fan. This is much better than the stock fan since it draws cooling air across the DIMMs and then blows cooling air across the chipset heatsinks behind the CPU socket. The air coming from this heatsink is much cooler than it was from the stock heatsink, presumably because of the increased airflow from the larger fan. There may also be some benefit from the vertical draw from the large fan since the fins are aligned vertically.

Here are the X3 720, X4 965, and third party Freezer 64 side by side. The Freezer 64 has a much larger 92mm fan.

On the right is the Arctic Cooling Freezer 7 Pro which can be used on both Intel and AMD sockets. It has the same size heatsink and same size fan as the Freezer 64.

A good thing to have handy for overclocking is a better case fan such as this Scythe unit.

Also good to have handy when swapping coolers is thermal compound cleaner and extra thermal compound such as these from Arctic Silver.

I'm current running my PII X4 965 at 3.8 Ghz with the NB at 2.6 Ghz. This seems pretty stable. I've been able to pass OCCT Linpack at about 60 C at 3.9 Ghz. I'm a bit disappointed that I couldn't get the NB up to 2.8 Ghz. I'm not sure how much of this is due to the BIOS. The version I'm using is the very first version that works with this processor.

My opinion so far is that if you have an X3 720 there is very little reason to replace the stock cooler, however I like the Freezer 64 much better than the stock cooler on the PII 965. The Freezer 64 matches the normal weight allowance of 500 grams and since it uses the same cam latch, it installs just like the stock unit. The extra airflow means that the exhaust air is much cooler and this is something I greatly prefer for the motherboard.

Laying The Ground Work For Proper Testing

I've seen reviews of AMD's Phenom II and Intel's Nehalem. These reviews have varied a lot in quality but none have really provided comprehensive results. It's time to find out.

I originally bought an AMD Phenom II X3 720 Black Edition. Mainly I did this to have something to try out while I was deciding what quad core to buy. The 955 BE looked pretty good. The 965 had a higher base clock but was also more expensive and was rated at 140 watts. On the Intel side, the i7-920 was still much more expensive than the PII 965. But, recently, this all changed. AMD released a new C3 stepping of PII 965 that is rated at 125 watts. Surprisingly, it was released at $200 instead of the $250 that most had been expecting. So, I ordered one. And, I ordered an Asus M4A79X motherboard which is similar to my MA4785 board but without graphics. I also purchased an ATI HD 4650 and couple of ATI HD 5770 graphic cards.

Then I noticed that the i5-750 was the same price, $200, and that I could get an Asus P55 motherboard without graphics that was almost identical to the 79X board and the same $120 price. I ordered those as well. This will give me two almost identical systems. Both systems will be native quad core with onboard memory controller and two memory channels. This should be an excellent head to head, dollar for dollar test. I'll use DDR3-1600 memory rated at CL 8. This makes the most sense because CL 7 memory is still less common, and faster memory rated at 1800 - 2100 Mhz tends to be twice the cost. I'm getting a couple of moderate sized third party coolers to test overclocking although I'm also interested in how much headroom there is with the stock HSF. Moderate sized is close to 500 gram weight, under 130mm's tall, and using 92mm fans. This compares with the heavy coolers which tend to be closer to 160mm's tall, weigh upwards of 700 grams, and use 120mm fans.

The hardware is perfect; this is the closest match of AMD and Intel hardware that I've seen in a number of years. I don't think we've had this close of a comparison since the K7/K8 single core days. The question now is how to test. I'm working on that. My game Dawn of War has a graphic check so see what level is playable. With my IGP 785 graphics the game is only playable with minimum settings. I can check this again with HD 4650, HD 5770 and HD 5770 Crossfire. To be honest, I don't expect to see much difference between the i5 and PII 965 systems. I'll also compare with my X3 720 to see if having another core makes any difference. PassMark has Peformance test which also includes both 2D and 3D graphic tests. I can say that my 785 graphics fail miserably on the last two tests. I'll give these a try but I wouldn't be surprised if they are low enough stress that even the 4650 card passes. I'm hoping that Dawn of War will be require a bit more although it too may top out before reaching the level of Crossfire.

For Integer testing I'm thinking about something based on GMP since this library shouldn't be tuned for either AMD or Intel. Use of the Intel Compiler is obviously out since this would contaminate the metrics. I have Visual Studio but this compiler is only middle of road in terms of what it produces. The next version is looking much better and it is just now available in Beta so we'll see. Better code would be nice but bugs in the Beta version could also contaminate the metrics. However, even the current version should be adequate with Integer code; it is the SSE code that is more of a concern. SSE2 is getting a bit dated. SSE3 is about the minimum level that would be nice to test. Better would be SSE4a versus some or all of Intel's SSE4. I don't have a requirement for full SSE4 testing since a fair bit of this will be replaced with Intel's next upgrade much as SSE became less important as wider SSE functions were added.

My operating system is 64 bit. I'm using 8 GB's memory and see no reason to waste time with 32 bits. Anything that I compile will be 64 bit. I do plan to test with both 2 DIMMs and 4 DIMMs to see if there is any difference. With my system I haven't seen any significant difference in timing or top speed. The two standard cases that I'll be using are both CoolerMaster cases with 200mm fans in front and top and a 120mm fan in the back. I do have a smaller case that only has a single 120mm fan in the back which I could test with. Personally, the notion of putting a $200 processor in a $60 case seems a little goofy but this could show what type of environment would be acceptable. Frankly I wouldn't be surprised if the 125 watt PII 965 were too much for that case.

I'm also glad that I got the X3 720 first since it is rated at 95 watts just like the i5-750. This should give me a pretty comparison of two 95 watt systems although since the i5 is a quad core I would expect it to be more powerful. I suppose if the i5 turned in thermals similar to my 720 while matching the performance of the 965 that would be quite a feather in Intel's cap since it would mean that i5's could be used in smaller cases with less cooling. Overall, it wouldn't be any great victory though since there is no price advantage. Of course, it has been suggested that Intel's power rating is bogus and is actually higher than they claim. Others have tested and insisted that Intel draws less power. Again, I don't really care about previous power draw tests since I have a 95 watt X3 and a 125 watt X4 to compare with. I suspect that since i5 is on the lower end of the Nehalem range it will actually fall in between these two but again I don't know without testing.

And, I have two graphics-free motherboards so I can test without the contaminating effect of integrated graphics. Given the huge gap between AMD and Intel integrated graphics there really is no way to directly compare them. The simplest solution is to discard the integrated graphics and use the same discreet graphic card. For lower level tests or small case tests I would use the HD 4650 which is a pretty solid, middle of the road card. I wouldn't expect a small case with one 120mm fan to be able to handle one 5770, much less two; nor would I expect it to handle an overclocked 125 watt processor. I know that I haven't had any thermal problems with my X3 720 but the case is well ventilated and it is only a 95 watt processor with integrated graphics. If i5 with 4650 does not pass a small case test then I can still project how well i5 would do with integrated graphics by comparing with 720 and the 785 motherboard. And, if the tests are borderline I picked up a heftier, 40 CFM, 120mm Scythe fan which would boost cooling in the small case. This should allow a pretty good inference for cooling with two regular 120mm fans. At any rate, thermal comparisons should settle any question of thermal issues either for Intel or AMD.

I still don't know if my thoughts about case testing are clear. I take issue with the open case, huge cooler testing that they do over at Anandtech. Likewise I take issue with the schizophrenic testing they do over at Toms Hardware Guide. I mean, who in their right mind would put four graphic cards in a small case? I don't really care that much about testing power draw. If electicity is really a concern you could always buy a lower power system with slower memory, 65 or even 45 watt processor and use integrated graphics. But most people are not that concerned about it. Of more concern is whether or not a given case will work with a given system. Generally, everything that people do to increase speed also increases heat. Voltage is increased on the memory, on the CPU, and even on the graphics. Higher clock speeds, more memory, and faster graphics all use more power. We all know that, at various times in the past, thermals were an issue. AMD had K7's before Barton that ran hot, Intel had Prescott which ran very hot and begat the BTX case as a desperate solution. There had been rumors of higher clocked AMD dual and Intel quad core 65nm processors running hot. Of course, now everyone is on 45nm but the top end chips are still rated at 125 or even 140 watts. My results should have much more practical value to people who would like to build lower and midrange systems rather just people at the very top end.

Keep in mind that thermal testing is somewhat separate from performance testing. You can't really begin benchmarking unless you know a given system is reliable. Too often, it seems that reviewers achieve a hasty estimate of maximum clock and then run their benchmarks without really knowing how stable the system is (unless it crashes during the tests). I suppose that and time pressure is why they take so many shortcuts. It takes hours just to run memory stability tests and hours more to run system and stress tests. And, this has to be repeated when trying to find a maximum overclock. Lots of variables like memory voltage, northbridge speed, CPU voltage, base clock versus multiplier all add up to hours and hours of added testing. And this before any actual benchmarks are run. I am confident in the settings on my system. I run the CPU at 3.4 Ghz. I've tested it much higher. I run with auto voltage on both the CPU and NB. I run with the base clock at 200 Mhz and the NB at 2.6 Ghz. I have the memory at 1333 Mhz with CL 7 and 1.545 volts. Auto doesn't work with the memory since auto is 1.5 volts and it will get errors at 1.515 volts with these settings. I'm confident that this is maximum performance for my system. I've tested overclocking the graphics but this isn't really worthwhile since you can get many times better performance without stressing the chipset just by putting in a moderate graphic card like the HD 4650. I expect both the i5 and 965 to show improved performance over my current system.

Overclocking

I've done a fair amount of testing with my X3 720 Black Edition. My results should be more typical since I'm using the stock heatsink and fan

As I mentioned in my last article people who like Intel's new i5 and i7 processors could argue that the increased wear caused by heat stress doesn't matter because they'll scrap their system in three years anyway. At least, I assume that is what they would say; I haven't heard from anyone who actually uses an i5 or i7 and wants to talk about heat testing with stock HSF. So, it is possible that they are doing better than I have heard. However, with mine I prefer that my system can pass a heavy thermal load test.

I'm currently running:
Phenom II X3 720 Black Edition - 3.4 Ghz; Northbridge - 2.6 Ghz, auto voltage
I cannot hit 3.5 Ghz stable nor can I reach 2.7 Ghz with the NB setting.

OCZ AMD Gold Edition, 7-7-7-19-31, 1333 Mhz, 1.545 volts
To run with these timings I need at least 1.53 volts to be stable.

The standard test program is Prime95. However, this program doesn't give any indication of what is going on with the temperatures so you have to run something else. I ran AMD Overdrive on mine so that I could monitor the temperatures while Prime95 was running. At the very least I would reccomend running the AOD stability test. It is a good test but it doesn't raise the temperature like Prime95. I also tried Intel Burn Test which raises the temperature even more than Prime95 (but not 20 C as the author claims). Mine idles at about 25 C, hits 51 C with Prime95, and reaches 55 C with Intel Burn Test. However, IBT is so buggy that even if your computer crashes it may simply be due to IBT and not due to instability in your system.

For those looking for something simpler I would suggest OCCT 3.1. The regular test on this application heats the same as Prime95 while the Linpack test has heating similar to IBT. However, it also displays core temperatures in real time just like AOD. I would still suggest using a boot CD with MemTest86 to verify the memory. Early on my system would eventually BSOD and reboot even at 3.2 Ghz. So I had wondered if maybe the X3 720 wasn't as good as I had hoped. But, it turned out that the processor was fine and one of my 4 DIMMs was bad. I only confirmed this by running MemTest86. It is also very good at verifying the stability when you change timing or NB speed. I suppose the only thing I haven't tried is using a larger CPU cooler however I wouldn't really expect to get more than maybe 100 Mhz out this. I'm not sure a larger Cooler is really worth it for just 100 Mhz. I'm currently considering getting the new C3 stepping of Phenom II X4 965. With the heat from an extra core a larger HSF might be more worthwhile.

I also tried increasing the Integrated Graphics Processor speed but realistically I don't think you can get much more than 10% this way before your chip gets hot. You might even get 20%. However, I've tried running the free evaluation copy of PassMark's Performance Test and you can really see the graphics bog down. The early tests hit 140 FPS then as the tests get more difficult they go down to 60 FPS, 20 FPS, and 10 FPS. No amount of tweaking of the IGP is going to improve 20 FPS enough to be acceptable, much less 10 FPS. This requires a beefier GPU. So, I'm looking at HD 5770 cards right now. These are pretty close to the older HD 4890 cards so they should be able to handle most graphic loads.

About Intel's i7

So why is it that I seem to be so down on Intel when others think Intel and especially i5/i7 is the greatest thing since sliced bread? Well, I don't have an i5/i7 to test but there are people who do a reasonably good job of testing and don't have a glassy eyed love affair with Intel (like Anand Lal Shimpi). The truth is that Intel's C2D was an excellent processor. However, the initial batch of Kentsfield quads ran hot. In fact, they ran so hot that you could not clock even to 3.0 Ghz without exceeding the rated temperature when using the stock HSF. Ouch. However, the G0 stepping did fix this. And remember that this was at a time when AMD was struggling to hit 2.3 Ghz stock with its own quad core. Then in early 2008 the 45nm Penryns came out and this cut the temperatures even further. The FSB was a serious bottleneck with these quads but again AMD was only creeping up to 2.4 Ghz so it didn't seem to matter. So, Intel enthusiasts had every reason to feel a bit smug. However, by the end of Summer of 2008, AMD was at 2.6 Ghz and using the tweaked 750 southbridge the overclocks were no longer embarrassing. Instead, Intel's severe FSB bottleneck began to be an embarrassment. But, with i7 just around the corner, Intel enthusiasts were able to grin and bear it.

Unfortunately, for i7, it has been a two edged sword. Now that the memory controller is on the die instead of in a separate chip i7 fixes the FSB bottleneck, but i7 also has to handle the entire heat load just as AMD's quad Phenom had to all along. And, just whe i7 was picking up extra heat, AMD had the nerve to release a 45nm Phenom II that actually worked and reduced power draw (just as Penryn did for Intel). Today, heat is once again a serious problem for Intel. But, don't take my word for it. Here at
Benchmark Reviews, Cooling i7 you can see that even undervolted to 1.16 volts, an i7 920 running stock at 2.67 Ghz is 38.5 C over ambient using the stock HSF. So, if your house is 72 F you'll hit 60 C on your processor. Unfortunately, the reviewers were using liquid cooling on both the chipset and video card so most likely your case will be 5 C warmer. In Indiana in the Summer 85 F would not be unsual if you don't have air conditioning. Add 5 C for the case and you are just over the 70 C max that Intel specifies. This is without overclocking. In contrast, my X3 720 running 600 Mhz overclocked would run under 70 C with the same conditions.

Figure in overclocking and it gets far worse. Looking at Benchmark Reviews, i7 Cooling Overclocked we see that running i7 920 at 3.8 Ghz and 1.4 volts increases the temperature by 20 C. With the stock HSF, you would be 10 C over max even at 22 C ambient without a video card. Add in the video card and you are easily 15 C over max. Ouch. But its worse than that. According to the testing, even at 22 C ambient you could easily see 10 C over max even when using a ZeroTherm NV120. A Xigmatek HDT-1S283 or Tuniq Tower 120 will keep you within 70 C just as long as you don't try to push it higher than 3.8 Ghz. And, keep in mind these temperatures are after they laboriously polished the i7's integrated heat spreader to a mirror surface and used a Yate Loon D12SH-12 cooling fan on each product tested. The D12SH-12 cooling fan forces an impressive 88 CFM of air at a moderately noisy 40 dBA. Without these extras your results may be worse.

I know that at this point there will be Intel enthusiasts who will be in full blown denial. They will insist that going over 70 C is nothing and that you can of course go over 1.4 volts and they may even insist that it is impossible for a Phenom II to run cooler since they know that it draws more power. The sad truth though is this quote from the same article:

"The Phenom II processor series from AMD offer a very large 37.31 x 37.31mm (1392.04mm total area) integrated heat-spreader surface, which is the largest processor surface I can recall since the original Intel Pentium (I) days. Compared to Intel's Core 2 Duo and Quad processors which measure 28.5 x 28.5mm, the Phenom II offers over 71% more contact surface area. If you compare the latest Intel Core i7 processors which measure 32 x 35mm, then the Phenom II series offers 24% more contact surface area. For overclockers, this will mean a much larger area to cool, but also much more manageable temperatures."

And, there it is. If you are really committed to Intel and you don't mind spending time polishing the heat spreader and you don't mind the decibel roar of a high volume fan and you don't mind the extra cost of a bolt through kit and you make damn certain that your case is well ventilated then you can indeed get an i7 up to an impressive overclock. Or you could do something similar with a Phenom II with a fraction of the effort because of its 24% larger heat spreader. And, adding insult to injury, AMD just released the C3 Stepping of PII 965 which gives AMD another 100 Mhz bump in overclocking to 4.0 Ghz. Can you match this with an i5 750 or i7 860? Yes, you can but it sure won't be easy. To paraphrase Robert Heinlein, the Intel i7 is a harsh mistress.

Update on new system

Even with the stock HSF, AMD's Phenom II X3 720 is a nice overclocker.

Case - Cooler Master Storm Sniper
Power Supply - Corsair Professional series 850 watts


Motherboard - Asus M4A785TD-V EVO
BIOS 4.10

Processor - Phenom II X3 720 BE, stock HSF
Bus, 200 * 17 = 3.4 Ghz, auto voltage
NB, 200 * 12 = 2.4 Ghz, auto voltage

Memory - OCZ AMD Gold Edition, 4 DIMMs, Unganged
Total, 4 * 2 GB = 8 GB's
Clock, 667 Mhz, 1.45 volts
Timing, 7-7-7-19-30-2T

Memory tested with a two full cycles of MemTest86 4.0 from boot CD. Processor tested with Prime95 using one thread per core of In-Place, Large FFT (maximum heating) which hits just over 50 C. Also tested with two passes of Intel Burn Test, 10 cycles each. IBT is only useful to raise the temperature; you have to ignore the numeric results since IBT has errors in calculation.

I had one bad DIMM from OCZ that I RMA'd without too much trouble and got my replacement. The Case had a bad circuit board which prevented the pretty blue fan lights from coming on. They sent me a new top which I swapped out with the old one and the lights work fine now. I haven't had any other problems with hardware. With Windows Vista though there were a couple of things. When it first installed some of the menu items like Accessories were missing. Later it developed a Component Store error which could not be fixed. This prevented upgrading to service pack 2. In the end I had to reformat and reinstall Vista. This install seems to be working just fine and it successfully installed service pack 2. I will upgrade to Windows 7 as soon as Microsoft sends me my copy. Theoretically it was shipped today.

I am curious though about Intel Nehalem based systems. I've seen people who claimed with a straight face that 85 C was okay because the CPU wouldn't be damaged until the temperature hit 100C. I'm baffled where this notion comes from. It seems to be a naiive assumption that since thermal throttling occurs at 95 C that anything below that must be safe. Thermal throttling seems analogous to me to a rev limiter on an engine. Engines have a green band where the engine normally runs, a yellow band where damage will eventually occur and a red band where damage can occur at any time. Rev limiters are normally set to hold RPMs within the yellow band and prevent moving up into the red. However, I have seen limiters which even allowed moving slightly into the red. If you aren't familiar with the notion of green, yellow, and red bands for RPM range this is an excerpt from an engine manual:


ENGINE OPERATING RANGE
Idle 1300-1600 RPM
Continuous cruise 3200-4800 RPM Peak torque @ 3600/100% throttle
Max continuous 5000 RPM Peak in-flight HP/100% throttle
Max peak / Time 5600 RPM / 5 Minutes Ignition rev limiter set @ 5900
Never exceed 6000 RPM Engine mechanical limit.

Here we have a direct comparison to processor temperatures. Obviously Idle would be the no load temperature with minimum clock. Continuous would be up to Intel's maximum temperature which on i7 920 would be 70 C. This is the green band where the processor was designed to run. Next we have the yellow band which for this engine runs from 4800 - 5900 RPM where the rev limiter kicks in. This is directly analogous to the temperature band on i7 920 which runs from 70 - 95 C where thermal throttling occurs. The never exceed limit is 6000 RPM which again is directly comparable to Intel's 100 C limit.

Now, notice that the maximum peak is 5600 RPM even though the rev limiter doesn't kick in until 5900 RPM. Peak is not an RPM that you can run on a continuous basis; the maximum continuous is only 5000 RPM. Increased wear occurs if you run above 5000 RPM for any length of time. The same thing happens when running a CPU above rated temperatute. The big difference though is that the engine can be overhauled whereas the procesor will have to be scrapped. I've seen overclockers blandly assume that Intel temperatures were fine because "thermal throttling didn't occur". This is almost certainly incorrect. Occasional jumps into the band above 70 C shouldn't cause any noticeable effects just as they don't when briefly revving into the yellow band with a engine. However, continuously running above 70 C is not just asking for trouble; it is sending out an engraved invitation.

So, what exactly does this mean in real terms? First of all, both Intel and AMD expect CPU's to have a lifespan of 30,000 - 50,000 hours. Secondly, for every 10 C you go over 70 C you decrease life expectancy by half. So if you ran something like Prime95 continuously with it stressing your processor to 85 C, life expectancy would be 1.2 - 2.0 years. The problem is that it would be difficult to reach this temperature with something like Folding@Home and few other applications would run continuously. Routine tasks like web browing and word processing will hardly stress the CPU above idle temperatures. In other words, if your temps stayed under 70 C then the couple of hours a day you spent playing Call of Duty with your temps hitting 85 C would be neglible on CPU life.

If you ran your CPU at 75 C continuously doing something like encoding and then hit 90 C for three hours a day playing Crysis then your CPU life would be 2.1 - 3.5 years. If you planned to upgrade in three years it might be worth the risk. However, keep in mind that your video card and memory (and sometimes the power supply) get their cooling air from inside the case. If the case temps are high because of heat from the CPU then it would be a good idea to either upgrade the case fans or to avoid overclocking memory or GPU. And since the power supply typically has twice the life expectancy of the CPU I would avoid using a case where the PS has to draw from the interior. For example, in my case, the PS draws cooling air through a screen on the bottom so it doesn't get warm air from inside the case. Over at Toms Hardware you can see a perfect example of a poorly design system where they tried to stuff a hot i7 920 into a micro-ATX case with two GTX 295 graphic cards while using a PS that draw from inside the case:


The problem was that the cooler’s heatpipes would interfere with the DFI LANParty Jr. motherboard’s heat sinks, making it impossible to mount the CPU cooler in such a way that would force CPU-heated air towards the rear case fan, and therefore, out of the case.

We were left with no choice except to channel heated air upwards (toward the PSU) or downwards (toward the video cards). Given these choices, we would prefer to force it upwards toward the PSU so that it could be channeled through the power supply and out of the rear of the case.

Unfortunately, in this specific application, the PC Power and Cooling S75QB PSU does not pull air from below like a lot of ATX PSUs do–one fan pulls air in from the front of the PSU, and then another fan forces it out the rear. The bottom of the PSU is solid, and pushing CPU heated air upwards into a brick wall isn’t our idea of a good time.

We were left with no choice at all, really: we had to push the CPU-heated air downward, towards the graphics cards.


Contrary to what the fools at THG suggest, the last thing you want a good power supply to be doing while under heavy load is acting as an exhaust fan for the CPU. This irrational power-supply-as-cooling-fan notion began with Intel's desperate BTX case. Likewise, if your CPU runs hot then avoid using the the cheap video cards that exhaust inside the case. In another THG build they put four cheap, internally exhausting video cards inside the case. However, the NZXT Tempest case they used includes two front fans, one rear fan, and two top fans so it would be able to handle the case cooling with no problem. However, once again we see the completely bogus link with the power supply:

It's a little ironic that the reason a PC Power & Cooling PSU wasn't ideal for our previous microATX build is precisely what makes it so attractive this time around. In our last SBM, we experienced less-than-ideal airflow out of our CPU cooler because the power supply didn't pull air from below.

This is absurd. The cooling air will be drawn in the front and exhausted out the three fans in the top rear corner. Having the power supply sucking in air on the bottom of the case would actually reduce the cooling efficiency since you want it to exhaust out the top. The builder even recognizes this point in opposition to his previous statement about the power suppply:


The Xigmatek Dark Knight can dissipate a good amount of heat for the $37 price tag, which we’ll use to push the hot CPU and GPU heated air upwards and towards the NZXT Tempest case's upward-facing exhaust fans.



I'm not against overclocking. I specifically bought a BE processor to make overclocking easier and the system more stable. The highest clock I can reach and still boot Windows is 3.6225 Ghz (17.5 x 207 Mhz). I can boot, run SuperPi, open a paint program, paste, clip and save the image. I can open a browser, go to a website, and upload the resulting image. I can do this at 3.6225 Ghz and with the stock HSF. I don't consider this practical though. The highest clock I can hit and still pass Prime95 is 3.485 Ghz. Again, on the stock HSF. So, I run mine at 3.4 Ghz to have some margin of stability. Temperature has never been a problem. The max temperature for my processor is 73 C. With Prime95 I can hit about 51 C and with Intel Burn Test I can hit 55 C. Both are well below the maximum.

I guess reasons like these are what turned me off from the idea of running a Nehalem system. You never hear Intel enthusiasts talk about the stock HSF; they always use a premium cooler like a Zalman or Thermaltake. In fact, when Intel sent out socket 1156 i5's and i7's for review they included Thermalright MUX 120 coolers. This heatsink is 160mm's tall which means it would fit in my case as long as I don't have a fan on the side panel but would not fit in most cases. When the manufacturer has to include a premium cooler something is definitely amiss.

The bottom line is that the days of the cool running Penryn's are gone. However, we also know that Penryns ran cool because the memory controller was in a separate chip. Now that Intel has the memory controller on the CPU die the entire thermal bill must be paid from one socket instead of going Dutch with the Northbridge. I would guess that Intel enthusiasts want to pretend that they can get the benefits of the IMC while keeping the low thermals of Penryn. Since this doesn't work so well, they resort to downplaying the notion of going over the rated temperature, and they seem to avoid doing stress testing with programs like Prime95. That all seems very strange to me. You pretend you have a robust processor and then handle it with kid gloves. I think the denial started when reviewers found out that Turbo often wouldn't work in a standard case with a video card. Also strange is the notion of touting Turbo as a great feature but then having to turn it off to overclock. I like AMD's Cool and Quiet and it stays on and running all the time. This allows the processor to clock down to 800 Mhz at idle.

Now, I have encountered the same attitude about ignoring common sense limits on AMD hardware but you have to search a lot harder to find it. For example when I was installing the processor on the motherboard the stock HSF seemed pretty hefty to me. I had a hard time imagining using something much heavier. However, I was wrong; the actual AMD spec for socket AM2/AM3 indicates a weight limit of 500 grams which is quite a bit heavier than the stock unit. 500 grams should be plenty but these example heatsinks are compatible with AM2/AM3 and heavier:

Sunbeam CR-CCTF - 590 grams
XIGMATEK HDT-D1284 - 667 grams
ASUS Royal Knight - 790 grams
Xigmatek THOR'S HAMMER HDT-S126384 - 800 grams
COOLER MASTER V8 RR-UV8-XBU1-GP - 867 grams
ZALMAN CNPS10X Extreme - 920 grams

All sorts of nasty things can happen if you flex the motherboard too much. Just as we find Intel enthusiasts who rationalize running over temperature limits I'm sure the AMD enthusiasts who buy these heatsinks would have similar rationalizations for going over the socket weight limit. However, I'm not looking for denial, excuses, or rationalization. A decent processor should be capable of running in a standard case with the stock HSF and still passing Prime95. If you have to make excuses why your system can't do this then you don't have much of a system.

While I'm at it I might as well talk about the AMD 785 integrated graphics. Dawn of War is almost more than the graphics can handle; I have to run it on absolute minimum settings. Sins of a Solar Empire actually has the same graphic requirement so again turning things off is a good idea. DemiGod has higher graphic requirements and is jittery at 1280 x 1024 even with minimum settings so you'll be near the bottom in graphic resolution. All of these games are enhanced for multi-core but apparently not enough to offset slower graphics. In other words, I'm not sure that my X3 processor is noticeably faster than it would be with an X2 and I'm pretty sure that an X4 wouldn't make much difference. They are playable but more resolution would be nice. Presumably these would be unplayable with Intel X58 graphics or older ATI 690G. Of course, low end discreet cards are not that expensive unless you are trying for a budget system. I had been looking at AMD 4890 cards before the 5000 series was released. Now it looks like the 5770 Juniper cards have displaced the 4890's and have Direct X11 support as well. I had considered 5850 but the cost is higher than I want to go. I'll have to revisit the performance when I upgrade.

New System

I waited to get enough spare money and then waited because of rumors of the 3.4 Ghz Phenom II 965. So, I decided to stop waiting and put together an interim system.

I didn't really care for the wattage bump on Phenom II 965 to 140 watts. I figure that AMD will probably release a new one at 125 watts like they did with Phenom so I'll wait for that one before getting a quad core. I did briefly consider Intel but unfortunately the Penryn based quad cores are pretty much obsolete because of the FSB bottleneck. That only left i7 920 but I didn't feel like spending all the extra money for little additional CPU power. Also tipping the scale was that I decided to wait for AMD's 5000 series GPU's to see how good they are. But with an Intel motherboard I would be stuck with some truly second rate graphics. Maybe if i5 were out there would be a genuine option from Intel but right now, it doesn't exist. I picked up:

CPU: Phenom II X3 720 Black Edition (tri-core)

Motherboard: Asus M4785TD-V EVO

Memory: 4 x 2GB DDR3 OCZ Gold AMD edition (capable of DDR3-1600)

Power Supply: Corsair HX 850 W Professional Series (modular)

Harddrives: Western Digital 640 GB, 1 TB, and a Western Digital Elements USB 1.5 TB external

DVD drive: Lite On DVD reader

Monitor: Asus 21" LCD

OS: Vista Home Premium and OpenSuse Linux


I'm using the stock HSF so overclocking is limited primarily by cooling. That is fine since I'm not really interested in any severe overclocking. To test stability I run Prime95 with maximum heating. From every test I've done Prime95 always comes up as the toughest. I've found that running windows is not as stressful as running the stability tests in AMD OverDrive and the AOD tests are not as stressful as Prime95. This makes the common practice among overclockers of "being stable enough to get a SuperPi score" something of a joke. I wouldn't trust a system without running the toughest stressor to see exactly where I stand.

The M4A785TD motherboard has had a few BIOS updates. I installed the latest one. I wasn't really interested in trying to unlock the 4th core as some have done since I knew that this would be a weak core. However, others asked me about it so I did give it a try. I was unable to show a 4th core with any configuration using Asus' Unleashed mode. Also, even though the BIOS has an option for setting the NorthBridge multiplier independently of the HT clock it doesn't work. Maybe in the next BIOS update. So, in order to increase the NB speed I have to bump the base clock just as people would have to if they didn't have a Black Edition with unlocked multiplier. I'm running memory in Auto which defaults to ganged mode but I don't know that that makes any difference in stability. Perhaps it would at 1600 but with only three cores I don't really need more than DDR3-1200. With CPU and NorthBridge voltage on auto, memory set to 1.59 volts and 533 Mhz, I ended up with:

250 Mhz base x 13 = 3.25 Ghz is stable. However, the 2.5 Ghz NB is overkill. With all 4 DIMMs I get 667 Mhz with a latency of 7 and 1T command rate.

238 Mhz base x 14 = 3.332 Ghz is stable. The 2.38 Ghz NB is still higher than it needs to be. With all 4 DIMMs I get 634.7 Mhz with a latency of 7 and 1T command rate. If the multiplier were locked this would be the fastest configuration.

231 Mhz base x 14.5 = 3.35 Ghz is stable. The 2.31 Ghz NB is okay. With all 4 DIMMs I get 618 Mhz with a latency of 7 and 1T command rate.

225 Mhz base x 15 = 3.375 Ghz is stable. The 2.25 Ghz NB is about right. With all 4 DIMMs I get 600 Mhz with a latency of 7 and 1T command rate. This one or the following one seem to be about the best all around configurations.

219 Mhz base x 15.5 = 3.395 Ghz is stable. The 2.19 Ghz NB is not too bad. With all 4 DIMMs I get 584 Mhz with a latency of 7 and 1T command rate.

212 Mhz base x 16 = 3.392 Ghz is stable. The 2.12 Ghz NB is about the lowest I would want to go. With all 4 DIMMs I get 565 Mhz with a latency of 7 and 1T command rate.

206 Mhz base x 16.5 = 3.398 Ghz is stable. The 2.06 Ghz NB is a bit slow. With all 4 DIMMs I get 549 Mhz with a latency of 7 and 1T command rate. If I specify 667 Mhz in the BIOS the memory defaults to a latency of 9 and 2T command rate. I might be able to tweak the settings back to 7 and 1T, but with the slower NB it wouldn't be worthwhile.


Update:

i5 750, i7 860, and i7 870 were not out yet when I ordered my components. However, now that I've seen the reviews it is clear that these processors wouldn't have mattered anyway.

Remember back when AMD released the B3 stepping of Phenom in early 2008? AMD discovered something that they had overlooked. Their processor wasn't playing nicely with the 700 series southbridge. So, AMD released the new 750 southbridge; and, when these Phenoms were used with motherboards having the 750 southbridge, it made a difference. You could easily get 200 - 300 more Mhz on an overclock. With that painful lesson under their belts, AMD upgraded the phase lock loop in the 45nm Phenom II's so that they got the same benefit whether you used the new 750 or the old 700 southbridge.

However, Intel has apparently fallen into the same experience trap with the newest processors. These chips have PCI-e on the die itself. Great for reducing cost but not so great for overclocking. Intel ties the PCI-e clock to the BCLK much as AMD has tied the HyperTransport frequency to its base clock since K8. Even Anandtech admits that getting away from 133 Mhz multiples will cause PCI-e problems. In contrast I had no trouble running the base clock up from 200 Mhz to 250 Mhz; HyperTransport still worked fine. Secondly, as you increase the frequency on Intel's newest processors it destabilizes PCI-e because the drive transistors cannot keep up at stock voltage. The solution would normally be to overvolt but unfortunately this can't be done very well with Intel's stock HSF. In fact, Anandtech used the word "sucks" several times in describing overclocking with the stock HSF. Anandtech claims a top clock of 3.37 Ghz for the i7 870 with stock HSF. This would be the same as what I'm getting however given Anandtech's checkered history with testing I have to assume that they did not try running Prime95 on all four cores with maximum heating as I did. This sounds like it would easily knock their claim down to the same 3.2 Ghz stock that Chile Hardware was able to get.

This is a problem for Intel if you really want more performance. Anandtech also says that you need to disable Turbo if you want the maximum clock without crashing. However, they then say that you need to leave Turbo on to let the system clock down the cores individually to avoid wasting power. And, even though Intel quietly suggested that reviewers use the Thermalright MUX 120 premium cooler to solve Lynnfield's thermal woes, Xbit Labs saw temperatures of 93 C under load at 4.07 Ghz. No thanks. Maybe Intel will fix these hardware issues in the newer 32nm i5's and i7's but for now these chips have issues. Nor have I even mentioned the problems encountered in trying to get Turbo to work under Linux or the way Turbo is unlikely to work in a standard case with a nice video card installed.

To get the current speed on my system I only had to change the base frequency, the multiplier, and bump the DIMM voltage to 1.59. The CPU base and NB voltages and DIMM timing are all on auto. Even with the bump in speed however I'm still running both the CPU and memory under their maximum rated voltage. My processor idles at 900 Mhz at 25 C and runs up to 3.375 Ghz with about 48 C under Prime95 load. If Intel were able to do this then they would have something to brag about.

I don't know what some of the reviewers have been thinking but worse overclocking than Bloomfield with 33% less memory bandwidth hardly sounds like a revolution. PCI-e onboard to reduce cost while also sabotaging overclocking sounds more like a reluctant compromise than a feature. And less functionality with a higher price tag than Phenom II is just not my idea of a good deal. Intel has a few obstacles to overcome (or a sharp price cut) before the new i5's and i7's will be genuine competition for AMD's Phenom II's. However for anyone with money to burn, a premium cooler, and a calm disposition the new Lynnfields could provide endless hours of fun trying for a high SuperPi score. I would stay away from Prime95 though at least until Intel comes up with a solution to the PCI-e problem.