Corei-7920 Bloomfield

[lanceor]

Hello I bought this processor(Corei-7 920)last 2009, do you think its time for me to upgrade to Ivy Bridge? or wait for Haswell next year?

How big is the difference between bloomfield and ivy bridge in terms of percentage? any opinion and suggestion is most welcome!

 

[Supernova1138]

It greatly depends on what you are doing with your computer. Another thing to keep in mind is that the i7 920 has very good overclocking potential, with 4 GHz quite possible if you have a D0 stepping 920. For gaming, it's not worth upgrading right now unless you have a high end dual GPU setup, and you aren't comfortable with overclocking, or have such a high end multi GPU setup that even a heavily overclocked i7 920 can't keep up anymore. Productivity software also runs faster with Ivy Bridge over Bloomfield, but not necessarily to the point that it is really worth dropping the money for a new CPU and motherboard.

In terms of clock for clock performance, going from Bloomfield to Ivy Bridge would be a 20% boost at best. Sandy was a 10 to 15% boost over Bloomfield, Ivy nets you another 5% maybe.

I would say wait for Haswell, Ivy wasn't a big jump over Sandy Bridge, with most of the improvements going into power efficiency and integrated graphics. Sandy Bridge wasn't a big jump over Bloomfield/Lynnfield, aside from higher clockspeeds. Right now unless you absolutely need bleeding edge CPU performance, it isn't really worth dropping the money for Sandy Bridge or Ivy Bridge if you have a first gen i5 or i7.

 

[amuffin]

It's not really worth it AT ALL.

 

[amuffin]

The 920 will be able to keep up with the 7870, even in crossfire mode.

 

[joedjnpc]

Wait for Haswell for sure, I'm in the same boat, and with a decent overclock Ivy just isn't worth the new mobo / CPU cost and hastle of reinstalling windows.

Have you considered OCing your 920? I found 3.6Ghz is very easy to reach, 3.8 not too hard either, and gives a great performance boost in CPU intense games.

 

[lanceor]

The 920 will be able to keep up with the 7870, even in crossfire mode.

I have a Sapphire HD7870 OC right now, and I havent overclocked any part of my PC.



Current

CPU : Intel Core i7 920 2.66Ghz (Quad Core) (Bloomfield)
HDD : Seagate 500GB 7200rpm
PSU : Cougar CMX 700w
MOBO : lga1366 Intel DX58SO
VGA : Sapphire HD 7870 2gb/256bit ddr5 (OC Ed)
Ram : Kingston 4GB DDR3 2x
Casing : CoolerMaster USP100 Mid-Tower
Heatsink : CoolerMaster X6


I don't know why but my FPS on BF3 MP is around 40-50, i thnk my processor is bottlenecking my vga


I'm not sure if I wanna OC my processor, my idle heat on four cores(CoreTemp) are around 40-48c

Playing BF3 my coretemp reaches 75c -.-

 

[amuffin]

It isn't.

Of course overclocking will give you a decent performance boost in BF3

 

[Guest]

You have a pretty decent hardware, your temps is more a problem of your Case CoolerMaster USP100 Mid-Tower.

Installing a full size card like Sapphire HD 7870 2gb and some hard disks you are blocking a faster airflow to your cooler by the bottom and the front fans. The low side intake may also trap the air to your GPU first and then heated to go to your cooler that more likely dragging air from your dvds. The upper intake is too close to the rear and some cool air coming from there gets immediately out! Just to make things worst you may have the rear of your box facing the wall some of the hot air may return back inside to your box by the low side intake. But even in the case you don’t and operating in wide open windows in your room a small drift of air may force even more hot air to get back inside. Your cooler looks pretty decent but also looks desperate for some cool air.

I strongly suggest Raven Rv03 you can see the review here in Tom’s hardware:

http://www.tomshardware.com/reviews/dragon-rider-chaser-mk-i,2992-5.html

As you may see in photo the hard disks are in separate room, in ambient temperature may have 35 c to 41 c eatch they are cooled if you install fan separately and their heat is not going to your cpu cooler or your card! The really good cable management is not blocking the air that goes more focused from the bottom directly to your CPU cooler and also colling the ram and the other component around the CPU more effectively! Also your full size card will be cooled by another big size separate intake. This box is taking the air from the bottom and releasing to the top. The hot air is always goes up that is making almost impossible to be back inside and I say almost, because there is always a small possibility with a drift of air to get so low even if you have it at the floor. With this Case is more than ok to have it facing the wall will help the hot air climb higher as always does.

Bottom line don’t be surprised if you get from -10 C to -15 C to your core's or more if you put all your hardware in there. Your cooler maybe proved good enough to run overclocked at 3 Ghz even the ambient temperature of 35 C you may bypass even the upcoming processors if you like for sure.

Paying attention in any detail will help you make the BIG difference.

 

[jaguarskx]

I wouldn't bother upgrading to Ivy Bridge. Assuming the same clockspeed, the performance difference will be about 18% on average. Of course if you think your PC is too slow for your liking, then you can upgrade.

I'm waiting for Haswell to come before upgrading. Maybe Broadwell in 2014....

 

[Guest]

Yes with Reaven Rv03 you can build a raid 0 array with 2 satta iii disks or 6 satta ii cheap disks you can find and you can go up to 385 Mbytes/sec increasing the performance that you will feel even in idle.

The serious Heat load is not effecting your Motherboard components, not directly anyway, I use wide open all the windows in my room to remove the hot air will less disturbance as I can.

I mean they fixing processors with more cores for what? You will not find many multithreading applications or x64 programs that using more than 3,5 Gbytes anyway, not to mention the games that are using only 1 core and the hyper threading doesn’t help always.

I believe we done here?

 

[Guest]

I don’t have much time to do a deep analysis for every case like I did over here and I think is better to make this thread more complete and adding a why i said no the new ivy bridge CPU?

Many people think as it consumes les energy will run cooler they take it and find them self’s in a situation to find a new CPU cooler thinking it will solve their problem well it is not!!

Older Intel Chips had the CPU core attached to heat spreader with a fluxed soldier as you can see here:

http://img138.imageshack.us/img138/4836/sb20ihs.jpg

That soldier had heat conductivity 80 W/mK

Ivy CPU’s are using a plain thermal paste as you can see here:

http://img15.imageshack.us/img15/3070/ivyy.jpg

With thermal paste that has thermal conductivity at 5 W/mK

Isn’t safe to assume that an older CPU was sending 16 times faster the heat to CPU cooler because the attached surface is a bit bigger but ivy chip is found running +20 C hotter in many different websites. Is not only an overclock problem is running hotter by the hour and I doubt someone with Ivy chip can run a 2 hours stability test of the machine like this:

http://img338.imageshack.us/img338/3953/prime956gbytesfurmark18.jpg

Prime95 torture mode 4096MB minimal FFT (half the CPU cache size) and 6 GBytes ram (more than half total ram of the system) while at the same time furmark 1.8 extreme burn was running. I believe they will have a problem even with 1 hour test of prime95 all CPU cache FFT size and half the ram!

With the cores at 90 c and above 100 c for long time you have the risk the glue that used to attach the heat spreader to start evaporating inside the chip! More problems can be created of this heat starts spreading in back of the board and though the copper all over the motherboard for those who doubt such a thing can happen I have these pictures taken by a thermal camera:

http://img4.imageshack.us/img4/5849/266pxheatdiffusion.png

You can see the first image the heat spreads to a surface all over and covers everything, always running faster to the cooled area, you can also see the stable heat flow to the heat sinks. The cooling effect itself seems to appears more random, the 2 pictures that follow, due to surface imperfections microstructure (some not detectable by the eye edges) and the air movement that is not 100% predictable.

Many crazy solutions can be found to the internet about cooling this chip removing the heat spreader and applying better thermal paste directly to the cooler, melting down even the attached surface! I assume not many will risk as you losing the guarantee.

Many people don’t seem to know that many motherboards were coming with heat sinks at the back like GA-P35-DQ6 with a system called Crazy Cool because is very effective. By removing this system and adding a bigger heat sink like HR-05/IFX as you can see here:

http://www.thermalright.com/products/index.php?act=data&cat_id=8&id=107

You can leaving it hanging out of the box parallel to the sealing and have a passive system that is actually dragging the heat out of the box. Have no doubt that is a very effective method and as the heat gets bigger to one side the other side is in ambient the performance is increased. You can add a small 80x80 fan and increase the performance even in idle mode.

I have serious doubts even this method proved to be effective enough to the ivy as I have no enough data of anyone preforming the hard stabilities tests that I making.

Bottom line I have to turn down this CPU as unusable in hard working data processing environments (hard multitasking), I suggest to anyone to avoid it.

 

[Guest]

Update: Someone with ivy cpu and the same cpu cooler with me NH-D14 tried something I told him:

http://www.tomshardware.co.uk/forum/forum2.php?config=tomshardwareuk.inc&cat=10&post=341689&page=1&p=1&sondage=0&owntopic=1&trash=0&trash_post=0&print=0&numreponse=0&quote_only=0&new=0&nojs=0

He got 90 c to the cores and system went above 100 c much faster than I expected the test terminated in les than 10 minutes! No way the 2 hours test, above, can happen to an ivy! Don’t even try prime95 1 hour test with minimal fft and all the cpu cache, half the ram usage! The NH-D14 is the best cpu cooler 5 years in a raw at Tom’s hardware proved very pure to handle this chips!!!

The faster spread to the system can only be explained by the larger amount of copper gathering around the cpu! Imagine is connected with ram, graphics card, peripheral devices chip unlike mine that only the chipset is connected!

If your motherboard has japanise capacitors are fully operational at 105 above that you risking malfunction, hardware failure, that lowers the volts all over the motherboard making even the coolers that are connected on board to run lower! In my test above you may notice I have only the 140x140 of the NH-D14 connected to power connector that usually runs at 1250 rpm if I see a drop at 1240 is sure there a drop at the voltage in the motherboard, is a way to detect it. All the other fans are connected to PSU directly (5 NF-P12), this way I’m making sure no fan will lose speed from a possible capacitor over heating failure.

The one that tried the test was a moderator, member of the staff at Tom’s hardware and we will try to find a solution.

Have in mind also is not a problem of OC the 90 c is something like standard to these chips read this one, didn't oc :

http://www.tomshardware.co.uk/forum/341561-10-temps-3770k

You will you run the risk to be the next one searching for new cooler? oc or not they ended up downclocking these chips at 1.6Ghz idle using Intel Speed step, Thermal monitor, cpu halt, or what even they can find in bios just to get some low temprature readings!! I run my cores at 3 Ghz idle....

These cpu's are sending arround 16 times slower the heat to the cpu cooler , i was cut clean above, NH-D14 5 years in raw no 1 to Tom's hardware failed, i have the same.

 

[Guest]

Update:

BlizzardGamer did a really good job researching the issue over the net:

http://www.tomshardware.co.uk/forum/341758-10-2500k-2600k-3770k-p8z77-ripjaws-2133mhz

But they don’t touch to the heart of the issue. The fluxed solder looked really small but it had thermal transfer conductivity 16 times faster than any thermal paste!
So why Intel didn’t use the same material attached to all of the cores driving the heat out faster and keeping the core temperatures to the lowest point? Is not hard Imagine the oc capabilities way above 5 Ghz if they do such a move. They use a black box that offers no thermal advantage conductivity to cover the cores even deeper! More strange is they didn’t even bother to spread this thermal paste all over the back of the heat spreader!? It would be much more effective this way. Point is even you try cut the heat spreader how much do expect to win applying better thermal paste all over your cooler bottom (is not easy thing to do)? -2 -4c? You can’t solve the fact that is getting hotter by the hour. I suggested try dragging out the heat from behind that will be proved much more effective, I have 2 at the back chipset, CPU. It isn’t easy thing to do either feating, mounting problems must be solved, cutting the box to drive the heat out, but in the end is worth the trouble keeping your system temperatures at the lowest point on heavy load witch is our target. I guess i was lucky that ga-p35-dq6 already came with a simular system at the back and the holes for the screws existed on board...

No water cooling method exist better than air they even try Freon friendly once (search Tom’s hardware) and they never even made it at the top of the charts. My NH-D14 is giving me +5-6 C above the ambient temperature truly madly deeply Noctua, there is no substitute.

 

[Guest]

Update:

in here someone claimed that an ivy CPU is running cooler than any existed CPU:

http://www.tomshardware.co.uk/forum/341835-10-bridge-wait

But in end admits his ivy cpu is incapable to handle half of his rams size at once, with only 4 threads for the cores!! The claims these chips are a small workstations, can be used in hard data processing environments, the multithreading is better are false alarms! The user in end chosen right not to try a test like this. Someone with an older cpu, sandy dosen't have a problem to perform something like this, is nothing to him.

 

[Guest]

Udate:

There are some ‘veterans’ that making the easily suggestion to someone to get Ivy:

http://www.tomshardware.co.uk/forum/forum2.php?config=tomshardwareuk.inc&cat=10&post=341756&page=1&p=1&sondage=0&owntopic=1&trash=0&trash_post=0&print=0&numreponse=0&quote_only=0&new=0&nojs=0

They seems to ignore the fact no cooler can make them smile for a minute or 2! Even when I end up telling them they will forced to operate at 1.6 Ghz in real word computing (idle), while any existed ‘old’ chip on line doesn’t drop under 2Ghz in idle they end up saying even this is NOT a problem!! You end up surfing, listen music in YouTube, reading sending mails, at lower speed than ever before.

If I make a mistake is more than ok for me to accept it and say ok bro you are right I had no IDEA about this issue. With the above attitude nobody wins in Tom’s Hardware accept Intel and the corporation’s behind all those ‘false’ moves.

In here also a useful link from someone that needed to upgrade now:

http://www.tomshardware.co.uk/forum/forum2.php?config=tomshardwareuk.inc&cat=10&post=341835&page=1&p=1&sondage=0&owntopic=1&trash=0&trash_post=0&print=0&numreponse=0&quote_only=0&new=0&nojs=0

I ended up suggesting for socket LGA 2011 the sandy bridge processors i7 3820 or i7 3930K, they are fluxed soldier condactivity processor, and you getting a good chance for a possible upgrade to a new upcoming processors hopping Intel will not such ‘foolish’ moves again.

 

[Guest]

Update:

Bloomfield is Back alive and kicking! Some of my links all over the forum are not working, I pissed really bad as all my work is in here!! I need time to reupload all the images such a huge mistake from me will never happen again I promise.

For the one who deleted is ok no big damage done feel free, not only you, but enyone to contact me prive and ask me enything you beliave is offencive or inaccurate to be changed.
Nobady can be perfect and such a thing dosen't exist in nature anyway! *** happens to anyone you can read my perfect screwup in here:

http://www.tomshardware.co.uk/forum/forum2.php?config=tomshardwareuk.inc&cat=10&post=341877&page=1&p=1&sondage=0&owntopic=1&trash=0&trash_post=0&print=0&numreponse=0&quote_only=0&new=0&nojs=0

It's was funny =;-)

 

[Mousemonkey]

This topic has been closed by Mousemonkey

 

[lanceor]

Best answer selected by lanceor.