Intel Xeon E5-2670 Deal and Price Tracking

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.

wildpig1234

Well-Known Member
Aug 22, 2016
2,198
443
83
49
would the 2667 v2 hash better than a 2680 v2?
yes... its faster than even 2690 v2 in single thread and only minimally slower than 2690 v2 in multithread. so it is def faster than 2680 v2... might be worth it now that 2680 v2 are no longer $170-180..... 2667 v2 is more reasonably priced than 2690 v2
 

alex_stief

Well-Known Member
May 31, 2016
884
312
63
38
You might want to keep an eye on power consumption. Generally, the higher clocked CPUs are less power efficient. At least from my experience, this holds true for the E5-2667v2.
 

alex_stief

Well-Known Member
May 31, 2016
884
312
63
38
Although TDP does not equal power consumption, these figures support the point I was trying to make. The dual 2667v2 I briefly got my hands on mined at around 300W. I did not put much effort into optimization, but still... higher frequency SKUs are always less power efficient than lower frequency ones. In simple terms: a hypothetical 12-core CPU at 2GHz will draw less power than a 6-core CPU at 4GHz with the same architecture while both may put out the same computing power for an algorithm that scales linearly.
My view on power consumption may be skewed by the electricity costs in my country approaching 30ct per kWh. I understand that power consumption is less of an issue in other parts of the world with much lower electricity costs.
 

wildpig1234

Well-Known Member
Aug 22, 2016
2,198
443
83
49
try 2687w v2 at 150w...lol for $400+... ....Seems like that's the end for now of any good bargain 2011 v1 or v2
 

wildpig1234

Well-Known Member
Aug 22, 2016
2,198
443
83
49
If you can find E5-2673 v2, it is a 2667 v2 refresh @ 110W
Doesn't seem like it's even available much at all. If you can even find it I bet you can probably get it at a very good price since you will be the only one bidding on it.....Sadly I think it will probably before offer at outrageous price like $400....
 

Joel

Active Member
Jan 30, 2015
850
191
43
42
Doesn't seem like it's even available much at all. If you can even find it I bet you can probably get it at a very good price since you will be the only one bidding on it.....Sadly I think it will probably before offer at outrageous price like $400....
I don't think it's worth going out of one's way for; as I repeatedly remind people, TDP has little to do with efficiency. It's basically the equivalent of a rev limiter in a car.

I suspect the two will perform identically except the 73 will start power throttling sooner. Useful in a datacenter mostly for power & cooling design at the chassis/rack/datacenter levels, but in a homelab: meh. Comparing the datasheet, the TDP is literally the only difference.

Intel® Product Specification Comparison
 

Stereodude

Active Member
Feb 21, 2016
453
88
28
USA
Be careful with the lower TDP chips. They may or may not run any faster than a lower clocked chip with the same TDP depending on the workload. I found that a E5-2689 v1 was barely any faster than a 2670 v1 during the 2nd pass of an encode with x264 (saturating the CPU) despite what looked like should have been a 10% advantage in clock speed. In the first pass (much lower CPU usage) it was about 10% faster.
 

Stereodude

Active Member
Feb 21, 2016
453
88
28
USA
I suspect the two will perform identically except the 73 will start power throttling sooner. Useful in a datacenter mostly for power & cooling design at the chassis/rack/datacenter levels, but in a homelab: meh. Comparing the datasheet, the TDP is literally the only difference.
It is entirely dependent on the specific workload and how much TDP headroom the chip has. I found with x264 video compression (fully saturating all the cores) the TDP would slow a v1 E5-2689 even if the CPU was kept cool (mid 50's). There was virtually no performance improvement vs. a 2670 (which has the same TDP) even though the e5-2689 can run 10% faster (with all cores loaded).
 

Robert Fontaine

Active Member
Jan 9, 2018
113
28
28
57
At the end of the day I will likely end up with a couple of 2790's because I can't afford to make the switch to v3/v4 and ddr4 and for the price of a couple of chips it's a decent upgrade for the 2760's but the darn things have spoiled me forever when it comes to value for money.
 

Stereodude

Active Member
Feb 21, 2016
453
88
28
USA
At the end of the day I will likely end up with a couple of 2790's because I can't afford to make the switch to v3/v4 and ddr4 and for the price of a couple of chips it's a decent upgrade for the 2760's but the darn things have spoiled me forever when it comes to value for money.
I replaced one with a E5-2687W v2 and a pair of them in a dual processor box with E5-2690 v2's.
 

alex_stief

Well-Known Member
May 31, 2016
884
312
63
38
Can confirm, the E5-2687w v2 are ridiculously overpriced for no obvious reason other than being one of the highest end SKU. E5-2667 v2 are virtually identical. 4GHz single-core turbo, 3.6GHz all-core turbo
 
Last edited:

Stereodude

Active Member
Feb 21, 2016
453
88
28
USA
That's about $1000+ worth of cpu?
Around there.

I probably would get 3x 2667 v2. get you nearly the same performance for significantly cheaper...
Well, if that's your criteria keeping your E5-2670 v1 will get you nearly the same performance for even cheaper. It all depends how we're defining "nearly the same performance". The E5-2670 v1 chips were absurdly cheap for what they were/are. Getting 3 Xeon's that used to be $2k a piece for not too much more than $1k total doesn't seem like a bad deal to me even if they're not the cheapest MIP per dollar.
 
  • Like
Reactions: Joel

einstein

Member
Jan 30, 2016
31
2
8
I've also ended up upgrading to a pair of E5-2667 V2 and they are a beast! Really happy with them even if they don't impress in benchmarks or in value for money.
 

wildpig1234

Well-Known Member
Aug 22, 2016
2,198
443
83
49
I've also ended up upgrading to a pair of E5-2667 V2 and they are a beast! Really happy with them even if they don't impress in benchmarks or in value for money.
I wouldnt say that now... given the recent increase in price in 2680v2... actually 2667 v2 (if $250 or less) is probably a good deal now given the 2680 v2 current price of $210+..

Everything is relative I guess
 

einstein

Member
Jan 30, 2016
31
2
8
It's not that relative though, paid 150$ for a pair of E5-2670 + 225$ for a brand new Tyan S7050 Motherboard and 100$ for 64GB of ram. Paying 500$ for a 20% increase in cpu performance isn't great value for money in comparison, and they also have higher power consumption...but i love them because i use them for running Qubes OS (based on Xen) and they are so much faster that it feels like a brand new system!
 

Boddy

Active Member
Oct 25, 2014
772
144
43
I wonder if all the interest in crypto currency is creating greater demand for cpu"s and servers that's pushing up prices these days? Or perhaps lack of flood of parts from end of service contracts? Kind of weird that old technology like 2011 socket cpu"s and ddr3 ram is actually going up in price.
 

Joel

Active Member
Jan 30, 2015
850
191
43
42
That's about $1000+ worth of cpu?

I probably would get 3x 2667 v2. get you nearly the same performance for significantly cheaper...
I think the math changes once you already have processors on hand, installed in a system. Then it becomes net cost of switching, plus futz factor (value of your time, if you prefer).

I wonder if all the interest in crypto currency is creating greater demand for cpu"s and servers that's pushing up prices these days? Or perhaps lack of flood of parts from end of service contracts? Kind of weird that old technology like 2011 socket cpu"s and ddr3 ram is actually going up in price.
Mining. 100%.