Xeon Owners Club

[Edwired]

Everything can add up watt usage like gpu, cpu, ram, motherboard, fans, lighting option. Regarding of overclocking it will increase power consumption overall. Unless you want to save on power down clock and decrease voltages will decrease power consumption. Is that kind of answer you looking for? Also look for watt Meter plug that will help you to see the overall power consumption

 

[lexluthermiester]

I would like to ask, what's his actual power consumption in idle/load

Generally between 40 and 100 watts for just the CPU. The total system wattage depends on what else you have in the system.

 

[Warrax]

Everything can add up watt usage like gpu, cpu, ram, motherboard, fans, lighting option. Regarding of overclocking it will increase power consumption overall. Unless you want to save on power down clock and decrease voltages will decrease power consumption. Is that kind of answer you looking for? Also look for watt Meter plug that will help you to see the overall power consumption

I have watt-meter, but I'm deciding, which one to take for tight budget build. The computer is used many hours in idle or in desktop browsing, that means CPU will go like 10-20% most of time, which is more closer to idle, than to full load power consumption.
But if the differance in wattage is not big, or after overclocking even zero, then I would take E5450. If differance is about 20-30W, then I want L5430, because it's also cheaper on aliexpress.
Anyway, I wanted to find graps like
L5430 idle - 100W,
L5430@3 ghz idle - 110W,
L5430 load - 150W,
L5430 @3 ghz - 170W,
and same for E5450,

to compare them on equal frequencies. But nowhere on internet is L5430 tested overclocked. Only extreme overclocks like 3,6 Ghz on 1,40volts etc...but still cannot find direct comparsion of two. I am also curious, how they dropped TDP so much, if also by Vcore, or by technology improvement. Cannot find these information anywhere.

 

[DR4G00N]

The L5430 has a slightly lower stock voltage and a lower maximum temp than the E5430 at stock. Otherwise they are the same and so at the same frequency & voltage the power consumption will be within the margin of error of silicon quality.

In your application it won't matter which one you use as neither will run beyond idle for any significant length of time.

I'd just run an E8400 or something for that system since it's just for web use, it'll be faster and use even less power.

 

[Edwired]

Well each cpu will have difference in tjmax, tdp, vcore, L1/L2 cache. Some of them like lower vcore and some like higher vcore depends on overclock or underclock

On this site it shows the L5430 is 10 months younger than e5450, better for lower typical power consumption. The tdp is general heat that is generated from the die that just my guess

Decided to mess around with msconfig to disable 2 cores to just to see what windows 10, cpuid, and hwinfo and what sort of loss I see. Quite surprised that I lost 6mb of L2 will test it for gaming and browsing later to see how bad it hit and usual checks on typical consumption

With 2 cores disabled dirt rally 2.0 ran good with abit of stuttering also longer loading time, resident evil 2 remake on the other hand ran really bad stuttering and freezing a lot totally unplayable, Firefox is slower to load up.

 

[8BitZ80]

I have watt-meter, but I'm deciding, which one to take for tight budget build. The computer is used many hours in idle or in desktop browsing, that means CPU will go like 10-20% most of time, which is more closer to idle, than to full load power consumption.
But if the differance in wattage is not big, or after overclocking even zero, then I would take E5450. If differance is about 20-30W, then I want L5430, because it's also cheaper on aliexpress.
Anyway, I wanted to find graps like
L5430 idle - 100W,
L5430@3 ghz idle - 110W,
L5430 load - 150W,
L5430 @3 ghz - 170W,
and same for E5450,

to compare them on equal frequencies. But nowhere on internet is L5430 tested overclocked. Only extreme overclocks like 3,6 Ghz on 1,40volts etc...but still cannot find direct comparsion of two. I am also curious, how they dropped TDP so much, if also by Vcore, or by technology improvement. Cannot find these information anywhere.

Back on page 216 I tested the L5420 and E5450 for power consumption and thermals. Here are some posts I made:

Anyone else had experiences with the Xeon L5420? Mine isn't overclocked and runs at ~65C idle and touches 90C during a Cinebench run. I tried different coolers to make sure it wasn't a thermal contact issue, but nope, it just seems to be running terrible despite the 50W TDP. Not sure what the deal is.

The X3363 for comparison has an 80W TDP and idles in the high 40s with a max of almost 80. Performs better, too.

I finally went to the hardware store and bought a power outlet meter to get the power consumption sussed out with these Xeons.
Currently running a GTX650, 4GB of DDR2 and standard hard drive.
Power was measured for the tower only.
These are my results:

Power consumption for E5450 SLBBM at 3.0GHz:
Idle = 57 watts
Cinebench = 113 watts
IntelBurnTest = 157 watts

Power consumption for L5420 SLBBR at 2.5GHz:
Idle = 58 watts
Cinebench = 95 watts
IntelBurnTest = 124 watts

So the power consumption seems to favor the L5420, but clock-for-clock (E5450 is 20% faster) there's no significant difference in efficiency.
My guess is that the higher thermals of the L5420 is from the use of inferior solder due to the cheap price tag when new. Maybe.

The only major difference was thermals. For some reason, the L5420 counter-intuitively ran hotter than the E5450. The idle temperature of the L5420 was approaching the max temperature of the E5450. Same thing occurred with both SLBBR and SLARP versions, so higher temperatures just seem to be inherent to those CPUs.

I personally really like the E54xx variants (especially the E5450) for the cool temperatures and great overclocking. The L54xx variants are still alright considering they do consume less power, but they might be more limited with overclocking due to the higher stock thermals. Haven't tested overclocking myself yet, though.

 

[Edwired]

Back on page 216 I tested the L5420 and E5450 for power consumption and thermals. Here are some posts I made:





The only major difference was thermals. For some reason, the L5420 counter-intuitively ran hotter than the E5450. The idle temperature of the L5420 was approaching the max temperature of the E5450. Same thing occurred with both SLBBR and SLARP versions, so higher temperatures just seem to be inherent to those CPUs.

I personally really like the E54xx variants (especially the E5450) for the cool temperatures and great overclocking. The L54xx variants are still alright considering they do consume less power, but they might be more limited with overclocking due to the higher stock thermals. Haven't tested overclocking myself yet, though.

Dont forget the tjmax offset that affects the a lot of them for temperature reading

 

[Warrax]

Back on page 216 I tested the L5420 and E5450 for power consumption and thermals. Here are some posts I made:





The only major difference was thermals. For some reason, the L5420 counter-intuitively ran hotter than the E5450. The idle temperature of the L5420 was approaching the max temperature of the E5450. Same thing occurred with both SLBBR and SLARP versions, so higher temperatures just seem to be inherent to those CPUs.

I personally really like the E54xx variants (especially the E5450) for the cool temperatures and great overclocking. The L54xx variants are still alright considering they do consume less power, but they might be more limited with overclocking due to the higher stock thermals. Haven't tested overclocking myself yet, though.

This was very fruitful, thank you. It's almost exactly, what was I searching for. Interesting would be how much power would draw overclocked L5430 to 3ghz, if it would be at least few watts more effective. Anyway, reason for same idle consumption is, as I said, that both of them have 6x multipier as lowest, so in C-state they are exactly same.

What is interesting, there is a 65W TDP Q9550s, which is same as Q9550, except, it has 65W TDP instead 95W. I've found some testing https://www.anandtech.com/show/2714/2
It shows a little bit better power effectivity than Q9650 or Q9400. So maybe, same results would have that L5430 @ 3 ghz.

Edit:
So I've just found this article, that compares Q9550s and Q9550 directly. Don't understand why it didn't showed up before.

Intel Q9550S: A Greener Quad Core? - Silent PC Review

Intel markets the Q9550S quad core processor as an energy efficient variant of the Q9550. Our comparison seeks to answer two basic questions: Is the Q9550S’ power savings enough to justify its much higher price, and how does it compare to an undervolted Q9550? June 22, 2009 by Lawrence Lee...

www.silentpcreview.com

The issue is finaly explained I think, all power savings are done only by undervolt from default volts (and only in load, so that's why idle power consumption is same), but undervolt can be done also on original processor. So I guess, same will apply also to E5450/L5430 issue, that means, that E5450 is better to take, for higher multiplier and possibly better overclock, while it can be still undervolted for power consumption, if you let it on stock.

I have suspision, that Intel wanted to sell inferior chips, that could not be overclocked good, this way, it's look good on paper with lower TDP, so they just undervolted them. Seems, that original Q9550 could be even undervolted more, so maybe they were selling inferior chips for this, and even charged extra $$$ for it.

 

[8BitZ80]

This was very fruitful, thank you. It's almost exactly, what was I searching for. Interesting would be how much power would draw overclocked L5430 to 3ghz, if it would be at least few watts more effective. Anyway, reason for same idle consumption is, as I said, that both of them have 6x multipier as lowest, so in C-state they are exactly same.

What is interesting, there is a 65W TDP Q9550s, which is same as Q9550, except, it has 65W TDP instead 95W. I've found some testing https://www.anandtech.com/show/2714/2
It shows a little bit better power effectivity than Q9650 or Q9400. So maybe, same results would have that L5430 @ 3 ghz.

Edit:
So I've just found this article, that compares Q9550s and Q9550 directly. Don't understand why it didn't showed up before.

Intel Q9550S: A Greener Quad Core? - Silent PC Review

Intel markets the Q9550S quad core processor as an energy efficient variant of the Q9550. Our comparison seeks to answer two basic questions: Is the Q9550S’ power savings enough to justify its much higher price, and how does it compare to an undervolted Q9550? June 22, 2009 by Lawrence Lee...

www.silentpcreview.com

The issue is finaly explained I think, all power savings are done only by undervolt from default volts (and only in load, so that's why idle power consumption is same), but undervolt can be done also on original processor. So I guess, same will apply also to E5450/L5430 issue, that means, that E5450 is better to take, for higher multiplier and possibly better overclock, while it can be still undervolted for power consumption, if you let it on stock.

I have suspision, that Intel wanted to sell inferior chips, that could not be overclocked good, this way, it's look good on paper with lower TDP, so they just undervolted them. Seems, that original Q9550 could be even undervolted more, so maybe they were selling inferior chips for this, and even charged extra $$$ for it.

Sounds about right to me. I don't have a functional L5420 right now since the ones I had were destroyed for science. From my own testing they don't seem to be power efficient or have a low TDP. On paper they have a 50W TDP but for some reason they perform like a 120W TDP chip. I'll be getting a new L5420 this week to continue with testing but I can already say with confidence that the E54xx chips are more desirable, especially considering there's no cost difference on Ebay. Pretty much a no-brainer.

 

[Toothless]

Anyone get stuttering issues when running multiple virtual machines on their Xeon build? I've got between three and eleven going and it seems after about a week the entire system likes to stutter pretty badly. There's enough ram left open and it's purely running on SSDs except for the storage HDD which is for backups and misc files.

 

[Edwired]

What Xeon build you having stuttering with?

 

[8BitZ80]

Other people been noticing prices going up on imported CPUs due to the COVID-19? I could've sworn there was a listing for Xeon E5450's for like $5 each a few weeks ago but it disappeared and now it seems like prices on everything else is creeping upward. I guess it's to be expected since they usually come from China or South Korea.

 

[Toothless]

What Xeon build you having stuttering with?

The one in my system specs.

 

[Edwired]

The one in my system specs.

Might be windows 10 background services do a check around task manager to see what spiking when mutli vm

 

[Toothless]

Might be windows 10 background services do a check around task manager to see what spiking when mutli vm

Nothing is spiking. No disk usage or any kind of usage. It just likes to lock up.

 

[Edwired]

Is sysmain and windows search is running in the services.msc?

 

[8BitZ80]

Straight up, this'll be ironic after I said I wasn't particularly fond of L54xx Xeons.
But I got 30 of them (plus a handful of E54xx ones) for about US$1.00 a pop so I'm not complaining at all.

Edit: As for the purpose of them, I don't currently have a use for so many processors. Most of them are identical variants so it'd probably be ideal for someone who wants to fill a whole server cabinet with low power quad cores.

 

[lexluthermiester]

Straight up, this'll be ironic after I said I wasn't particularly fond of L54xx Xeons.
But I got 30 of them (plus a handful of E54xx ones) for about US$1.00 a pop so I'm not complaining at all.

Edit: As for the purpose of them, I don't currently have a use for so many processors. Most of them are identical variants so it'd probably be ideal for someone who wants to fill a whole server cabinet with low power quad cores.

View attachment 148402

Ebay?

 

[8BitZ80]

Ebay?

Yeah. Gave an offer though, it wasn't the original Buy It Now price.

 

[Edwired]

The x5675 just arrived this morning currently waitng on the asus p6x58d-e to arrive could take 2 weeks maybe. It was the cheapest board i got across the european

 

[lexluthermiester]

The x5675 just arrived this morning

That's a good CPU.

 

[Edwired]

Yeah seen good reviews on it. Just learning to get my feets wet on overclocking with newer cpu, motherboard and ram and work my way up the Xeon chain. Just hoping I don't kill it with a blink of an eye

 

[lexluthermiester]

Yeah seen good reviews on it. Just learning to get my feets wet on overclocking with newer cpu, motherboard and ram and work my way up the Xeon chain. Just hoping I don't kill it with a blink of an eye

With OCing, it's a universally held opinion that it is a very good CPU to OC. But I advise caution. Take it easy with the level of OC and voltages you push it to. The CPU will likely get an excellent result with a BCLK OC of 166mhz and even 200mhz. As you raise the BCLK remember to lower the multiplier. You should be able to easily get 4ghz out of it with careful tweaking and experimentation. 4.2 and even 4.4ghz are generally possible with higher voltages and with good cooling. 1.35volts is the upper range limit of Intel's VID spec(see link below) for that CPU. It can handle 1.4v but I would not recommend anything higher than 1.375v for daily use.

Product Specifications

quick reference guide including specifications, features, pricing, compatibility, design documentation, ordering codes, spec codes and more.

ark.intel.com

 

[8BitZ80]

This was quite interesting. I was modding the BIOS on my new Gigabyte P35 DS3P and had installed a Xeon E5440 with the un-modded BIOS. It seemed to perform pretty good and so I made some Minecraft 1.6 framerate benchmarks to test the before-and-after BIOS mod CPU performance. Got the R9 270X installed so there was no issues with GPU limitations.

Standard F12 BIOS:
TEST 1 - 87.6 avg, 46 min, 140 max
TEST 2 - 75.5 avg, 49 min, 92 max

Modded Xeon F12 BIOS:
TEST 1 - 112.6 avg, 53 min, 176 max
TEST 2 - 93.5 avg, 79 min, 121 max

Never expected to see that kind of performance boost. A lot of those numbers went up 30fps (including minimum FPS for Test 2 which was crazy) and it's like a totally different machine now. Also tested Cinebench R15 but it remained identical at 300cb, but not surprising behavior from a synthetic benchmark.
Minecraft isn't the ideal benchmark tool, but it's the only decent CPU-intensive game I've got and it's great for testing LGA775 systems.

 

[RadeonProVega]

Question is my CPU speed correct?
Intel Xeon 2630 v2. When its at full throttle i was wondering was it suppose to be higher, the core speed, looks like its staying at 2.8ghz. But i thought it uses turbo boost 2.0 and it should be reading 3.10ghz or something like that.