Intel Core i7 Review Thread

[Astratuner]

guys I've got a question for you :

What is SMT Enabled? What does SMT stands for? What does it do?


Can anybody please explain me cause I can't find it on the web :$

[ap4lifetn]

SMT = simultaneous multi threading, aka hyper-threading on P4's

[Spectrobozo]

guys I've got a question for you :

What is SMT Enabled? What does SMT stands for? What does it do?


Can anybody please explain me cause I can't find it on the web :$

is the "hyper threading", like in the Pentium 4, or the Atom...

[Blauhung]

toms link.

It is not temp downclocking. You can see these options in the bios screenshots as well, on the EE edition you can override them.
What is hardwired into the chip, is it is limited to powerdraw as well, not only temp. So if you buy a i920, and you luck out, like i did with my q6600 and your tdp is way down lets say 72 watts for arguements sake, you ve got massive headroom and the volts would probably get to high before you reach that wall. However, there is a spread , like in all processors, so if you get an i920 that has 110W tdp at default volts you wont get very far before that wall comes into play. Since the i920 i would assume is the lower bins, the luck of getting a low TDP chip is not on your side, and this will otherwise limit oc'ing for a chip, that might have to be pushed harder than another.
The benefit or protection element doesnt make sense to me. If you have temp based down clocking.. i dont see the necessity for this, as you void your rma abilities if you go beyond spec anyways... so either they found a high inciedence in fried chips.. meaning this can't handle volts well (which i doubt) or they wanted to stop budget chips from getting up to 4 - 4.5 on air... it makes perfect business sense, just hoops us budget oc'rs who cant drop 1k on an EE chip.

I'm not 100%, but I'm pretty sure it's not power based downclocking, it's merely the basis for when turbo mode is allowed to kick. The override on the 965 allows you to set that limit so that turbo mode functions nearly all the time when the cpu is in use. On the locked chips, turbo mode only kicks in when this headroom is available. There is no forced thermal throttling when these chips reach this point, just turbo mode won't kick in.

OC'ing will still work, just if you get close to the TDP, you probably won't get that extra oomph to kick in.

......


Most reviews conclude that tri-channel provides close to zero benefits, is there a reason why people still think they need tri-channel?

Right now, the cpu probably doesn't have quite enough execution throughput to fully utilize the full memory bandwidth that is available to it when all 8 virtual cores are running full bore. There's also the issue that most programs can't even get close to running these chips full bore, so there's plenty there to be used to fulfill the cpu's every whim. Prefetching probably also puts some extra use on the memory, but not much. As you can see from some benchmarks, there is a boost, just only when the cpu is pushing all 8 threads to the max.

Currently the real reason for the extra bandwidth comes from the server space. Since gainstown introduces a NUMA (non-uniform memory access) type platform on intel's line, there will be calls from one cpu to the other to get stuff stored in either chips local memory. This is the case where the bandwidth comes into play. Say CPU0 is running full speed working on some verry threaded server workload, sucking up as much memory bandwidth as it can take. Then CPU1 comes along after not finding what it needs in the memory attached to it, and requests it from the memory attached to CPU0. The bandwidth is still available to fulfill this request without causing any slowdown in what CPU0 is doing.

The real boom for tripple channel probably won't happen till westmere comes out with it's 12 virtual cores that will be even more hungry for bandwidth. Right now, the most people here with bloomfield won't really see much gain unless you're using very threaded programs, but I'm pretty sure that all the crunchers here who are thinking about a gainstown platform will see big improvements when filling all 3 channels to both sockets.

[villa1n]

I'm not 100%, but I'm pretty sure it's not power based downclocking, it's merely the basis for when turbo mode is allowed to kick. The override on the 965 allows you to set that limit so that turbo mode functions nearly all the time when the cpu is in use. On the locked chips, turbo mode only kicks in when this headroom is available. There is no forced thermal throttling when these chips reach this point, just turbo mode won't kick in.

OC'ing will still work, just if you get close to the TDP, you probably won't get that extra oomph to kick in.

Ahh, well that seems much more reasonable and logical. So Tom's sound like they dramatized what this was. If i can push my chip to 3.8, I m not really going to miss my turbo that much hehe good that this seems a lot clearer now

[Astratuner]

is the "hyper threading", like in the Pentium 4, or the Atom...

So it was that easy after all

Thanks mate!

[initialised]

If the limit is 3.8-4.0ghz for budget chips, then it's not much of a protection anyway.. Anyone with some first hand knowledge?

4.0GHz Stable, 4.2GHZ into Windows (poor cooler), 4.44GHz POST on Ci7 940 on P6T

My feeling at 4.2GHz was that there was something getting in the way but given that on removal of the core contact, three HPC cooler I noticed that only one of the HPCs was making good contact so probably poor cooling. It would happen at ~80-115C depending on the voltage so my guess is that if the temperature swing between idle and load was smaller the lock up would not happen.

We have a Titan cooler to try next, the base in smoother so I'm expecting better results.

[Blauhung]

Ahh, well that seems much more reasonable and logical. So Tom's sound like they dramatized what this was. If i can push my chip to 3.8, I m not really going to miss my turbo that much hehe good that this seems a lot clearer now

yeah, as I've said before, I just make the things, but I'm pretty sure that setting only applies to turbo mode. Thermal throttling only kicks in when the chip is thinking about burning a hole in something, and EIST only kicks in when cores aren't in use.

[vengance_01]

The cost of the boards, DDR3 and the chips will prevent people from getting this right off the bat. Its clear this chip was designed for the server market where it will excel vs current Xeon Core 2 based Quads. For 95% of the desktop market Dual Core Core 2's like the 8400 will offer the best bang for your buck. Even so the platform costs are just to high(mainly the 300$ X58 boards)

[rge]

That would be nice if the 130W, 100amp limits only applied to turbo mode. That is not what the toms article says, but they could be wrong. All those i940's reaching 4ghz at idle is not relevant, as no chip will reach 130W at idle. The claim by tom's is once the chip is loaded (prime etc) if load exceeds those limits the "cpu reduces its clock speed" . What I would like to see is several of those i940's at load with prime or linpack with cpuz showing the clock speed to make sure it does not downclock on load.

[Helmore]

The cost of the boards, DDR3 and the chips will prevent people from getting this right off the bat. Its clear this chip was designed for the server market where it will excel vs current Xeon Core 2 based Quads. For 95% of the desktop market Dual Core Core 2's like the 8400 will offer the best bang for your buck. Even so the platform costs are just to high(mainly the 300$ X58 boards)

It's not meant to be the mainstream platform, for that you will have to wait for Lynfield and Havendale on socket LGA1156 (around summer 2009).

EDIT: Tom is simply talking crap, besides, turbo-mode can be turned off as far as I know.
Just check the last BIOS screen on this page:
http://techreport.com/articles.x/15816/4
They also state this on the following page:

We also kept the Core i7's Turbo mode enabled for testing, although we did disable it when overclocking the boards.

[bowman]

Besides 1 or 2 games. i7 beats Core 2 in games...specially in highres!

You say that as if it's a good thing.. The 'revolutionary' new microarchitecture is slower in certain cases than the old one?

I was fully prepared for it not being any faster in games than Penryn, that's to be expected (except for those silly, irrelevant 640x480 benches), but slower by double-digit frames per second in real-use high-res tests? Jesus, Maria, that's pretty ing retarded.

I guess that's what happens when you take the work of the geniuses in Israel and hand it to the NetBurst designers for polishing..

Now to decide whether to get a Penryn, wait this out even more and get a Sandy Bridge, or what..

[Macadamia]

I know Tom's article mentions that, but I'm not sure this is entirely true, as Anand has overclocked an i7 920 using an ungodly amount of voltage up to 3.8ghz, which has been pointed out in their comments. http://anandtech.com/cpuchipsets/int...px?i=3448&p=15
THG is not the most credible site IIRC, so there's still hope.
If the limit is 3.8-4.0ghz for budget chips, then it's not much of a protection anyway.. Anyone with some first hand knowledge?

Isn't that an ES?

AFAIK everyone reviewing these chips have an ES.

[dinos22]

i think Anandtech sums it up best in their conclusion

Expecting a sequel to be a reincarnation of the original is just setting yourself up for disappointment. A good sequel will be able to stand on its own, independent of whatever may have come before it. Nehalem is Intel's Dark Knight, it lacks the reinvention that made Conroe so incredible, but it continues what was started in 2006.

they really did a good review on Nehalem

hey Gary how many of these did you guys kill j/k

[JumpingJack]

I know Tom's article mentions that, but I'm not sure this is entirely true, as Anand has overclocked an i7 920 using an ungodly amount of voltage up to 3.8ghz, which has been pointed out in their comments. http://anandtech.com/cpuchipsets/int...px?i=3448&p=15
THG is not the most credible site IIRC, so there's still hope.
If the limit is 3.8-4.0ghz for budget chips, then it's not much of a protection anyway.. Anyone with some first hand knowledge?

Tom's is incompetent.

EDIT: This is why ... http://www.tomshardware.com/reviews/...e-i7,2063.html the dimwits didn't even think about flipping that bios option "over current protection" off to try over-clocking.

[Hornet331]

ouch, when a game engien is only decent programmed for multithreading even a 920 destroys a QX9750.

Anyway thats just bonus, can't wait till i see how that monster does at distributed computing... imagine 3,8GHz 8 threads.

[Toysoldier]

I see that "some people" are eating up the news that Core i7 performs only a little better than Core 2 at games (even though there are huge increases in multi-GPU scenarios).

Suddenly game benchmarks are all that matters.

Well, I know what I'm upgrading to next month, and it definitely isn't an AMD platform.

+1

[JumpingJack]

ouch, when a game engien is only decent programmed for multithreading even a 920 destroys a QX9750.

Anyway thats just bonus, can't wait till i see how that monster does at distributed computing... imagine 3,8GHz 8 threads.

When the GPU horsepower is there, Nehalem shows it's muscle:
http://www.guru3d.com/article/core-i...ance-review/19

You could also look at this processor as a good investment for future (much faster) graphics cards.

I warned people about this, that most reviewers would run it up against the GPU wall, and then conclude the CPU was not a 'gaming cpu'. For SLI/xFire setups, this will be the CPU to pair it with.

[Metroid]

Overclocking A Core I7

We started off overclocking the Extreme Edition to 4.1GHz, raising the overall score to a phenomenal 2,037 – the fastest ever recorded. However, the 2.66GHz Core i7-920 also proved highly overclockable, as we could also boost its frequency to 4.1GHz by raising the base clock from 133MHz to 205MHz. At these settings, the overclocked Core i7-920 proved incredibly fast, running Crysis at a minimum frame rate of 51fps, and scoring 2,005 in our Media Benchmarks.

In comparison, even when overclocked to 4GHz, the QX9770 could only manage a minimum of 39fps in Crysis and 1,679 in the Custom PC Media Benchmarks. The 6.218 seconds taken to run WPrime on the Core i7-920 overclocked to 4.1GHz was the second fastest time on the planet at the time of writing.

Conclusion

The graphs show the new range of Core i7 CPUs in direct comparison to what was previously the fastest processor in the world, the Core 2 Extreme QX9770. Our testing proves conclusively that Core i7 is significantly faster clock for clock than any other CPU series, and great for everything from video encoding to gaming.

Source Custom PC

They summed up pretty good.

[Hornet331]

When the GPU horsepower is there, Nehalem shows it's muscle:
http://www.guru3d.com/article/core-i...ance-review/19



I warned people about this, that most reviewers would run it up against the GPU wall, and then conclude the CPU was not a 'gaming cpu'. For SLI/xFire setups, this will be the CPU to pair it with.

yeah i saw that, nearly double the performance when the gpu isn't the limit.

But since im not into SLI/CF that don't really bothers me.

[fragmasterMax]

if ya got a endless pocket ur laughing lol upgrade for me next year wen everything matured abit more prices dropped abit so i will wait. my set up can play any game on the planet at max settings pritty much crysis struggles abit but with a hack i put in maxed out 40fps thumbs up tho for this thumbs down price

learn some english bro.
seriously, that was one big indecipherable fragment.

[JumpingJack]

yeah i saw that, nearly double the performance when the gpu isn't the limit.

But since im not into SLI/CF that don't really bothers me.

Yeah, the scatter of data we saw today ... it would be hard to recommend a pure gamer to use Core i7 unless they were really interested in having just a 'future' upgrade path with just the GPU or they are going to plunge all he way into massive multi-GPU setups.

Most all GPUs today are not well paired with this CPU, the current crop of cheaper 45 nm or Phenom/Deneb will do fine.

[Metroid]

i think Anandtech sums it up best in their conclusion



they really did a good review on Nehalem

hey Gary how many of these did you guys kill j/k

It was an interesting review and what got my attention was just one important thing [as the performance increased as the power got hungrier] on a side note they said 32nm Westmere will cure most of these problems.

[Movieman]

Source Custom PC

They summed up pretty good.

Air,TRUE in a 72F room..
Not quite.. SP1M 9.812s.. Cinebench10-23,034..

[villa1n]

Isn't that an ES?

AFAIK everyone reviewing these chips have an ES.

Interesting point, so you think aside an unlocked multi, they would also have inbuilt protections disabled that wouldnt be in a retail?