| | | | | | |

Xeon E5-2670 v3 vs Ryzen 9 3900


Description
The E5-2670 v3 is based on Haswell architecture while the 3900 is based on Zen 2.

Using the multithread performance as a reference, the E5-2670 v3 gets a score of 452.9 k points while the 3900 gets 687.5 k points.

Summarizing, the 3900 is 1.5 times faster than the E5-2670 v3. To get a proper comparison between both models, take a look to the data shown below.

Specs
CPUID
306f2
870f10
Core
Haswell-EP
Matisse
Architecture
Base frecuency
2.3 GHz
3.1 GHz
Boost frecuency
3.1 GHz
4.3 GHz
Socket
LGA 2011-3
AM4
Cores/Threads
12/24
12/24
TDP
120 W
65 W
Cache L1 (d+i)
12x32+12x32 kB
12x32+12x32 kB
Cache L2
12x256 kB
12x512 kB
Cache L3
30720 kB
4x16384 kB
Date
September 2014
September 2019
Mean monothread perf.
36.41k points
74.97k points
Mean multithread perf.
452.9k points
687.5k points

Non-optimized benchmark
The benchmark in Mode 0 (FPU) measures cpu performance with non-optimized software. It uses the basic µinstructions from the i386 architecture with the i387 floating point unit. This mode is compatible with all CPUs so it's practical to compare very different CPUs
Monothread
E5-2670 v3
3900
Test#1 (Integers)
3.52k
4.39k (x1.25)
Test#2 (FP)
8.68k
16.99k (x1.96)
Test#3 (Generic, ZIP)
2.83k
7.74k (x2.73)
Test#1 (Memory)
3.13k
24.57k (x7.85)
TOTAL
18.17k
53.69k (x2.95)

Multithread

E5-2670 v3

3900
Test#1 (Integers)
32.7k
48.45k (x1.48)
Test#2 (FP)
105.18k
230.81k (x2.19)
Test#3 (Generic, ZIP)
35.18k
122.82k (x3.49)
Test#1 (Memory)
9.54k
54.73k (x5.74)
TOTAL
182.59k
456.8k (x2.5)

SSE3 optimized benchmark
The benchmark in mode I (SSE) is optimized for the use of SIMD instructions with 128 bits register and the SSE set up to version 3. Nearly every modern CPU has support for this mode.
Monothread
E5-2670 v3
3900
Test#1 (Integers)
11.08k
16.59k (x1.5)
Test#2 (FP)
9.57k
24.55k (x2.57)
Test#3 (Generic, ZIP)
3.18k
8.55k (x2.68)
Test#1 (Memory)
3.28k
26.28k (x8.01)
TOTAL
27.11k
75.97k (x2.8)

Multithread

E5-2670 v3

3900
Test#1 (Integers)
139.07k
181.33k (x1.3)
Test#2 (FP)
133.47k
261.01k (x1.96)
Test#3 (Generic, ZIP)
46.6k
126.62k (x2.72)
Test#1 (Memory)
6.65k
66.64k (x10.02)
TOTAL
325.78k
635.6k (x1.95)

AVX optimized benchmark
The benchmark in mode II (AVX) is optimized to used 256 bits registers beside the first version of the Advanced Vector Extensions (AVX). The first AVX compatible CPU was released in 2011.
Monothread
E5-2670 v3
3900
Test#1 (Integers)
11.18k
16.16k (x1.44)
Test#2 (FP)
10.06k
23.75k (x2.36)
Test#3 (Generic, ZIP)
3.21k
9.32k (x2.91)
Test#1 (Memory)
3.12k
23.17k (x7.42)
TOTAL
27.57k
72.4k (x2.63)

Multithread

E5-2670 v3

3900
Test#1 (Integers)
140.98k
174.73k (x1.24)
Test#2 (FP)
145.37k
275.63k (x1.9)
Test#3 (Generic, ZIP)
45.62k
125.65k (x2.75)
Test#1 (Memory)
6.75k
44.17k (x6.54)
TOTAL
338.73k
620.18k (x1.83)

AVX2 optimized benchmark
The benchmark in mode III (AVX2), like AVX1, is optimized to used 256 bits registers beside the second version of the Advanced Vector Extensions (AVX). The first AVX2 compatible CPU was released in 2013.
Monothread
E5-2670 v3
3900
Test#1 (Integers)
18.62k
16.85k (x0.91)
Test#2 (FP)
10.49k
26.03k (x2.48)
Test#3 (Generic, ZIP)
4.04k
9.54k (x2.36)
Test#1 (Memory)
3.26k
22.55k (x6.91)
TOTAL
36.41k
74.97k (x2.06)

Multithread

E5-2670 v3

3900
Test#1 (Integers)
236.25k
229.41k (x0.97)
Test#2 (FP)
151.05k
292.81k (x1.94)
Test#3 (Generic, ZIP)
58.9k
128.48k (x2.18)
Test#1 (Memory)
6.7k
36.8k (x5.49)
TOTAL
452.9k
687.5k (x1.52)

Performance/W
E5-2670 v3
3900
Test#1 (Integers)
1969 points/W
3529 points/W
Test#2 (FP)
1259 points/W
4505 points/W
Test#3 (Generic, ZIP)
491 points/W
1977 points/W
Test#1 (Memory)
56 points/W
566 points/W
TOTAL
3774 points/W
10577 points/W

Performance/GHz
E5-2670 v3
3900
Test#1 (Integers)
6006 points/GHz
3920 points/GHz
Test#2 (FP)
3385 points/GHz
6054 points/GHz
Test#3 (Generic, ZIP)
1303 points/GHz
2218 points/GHz
Test#1 (Memory)
1052 points/GHz
5243 points/GHz
TOTAL
11745 points/GHz
17435 points/GHz

Monothread performance graph
Monothread performance graphics gives the performance vs time. They are useful to measure the time it takes to the CPU to reach the maximum performance.

Usually, CPU's performance will be steady during these tests but if it has a slow frequency strategy, the first samples will show a lower score.


Test#1 (Integers) [points vs time]

grafica bm.hardlimit.com


Test#2 (FP) [points vs time]

grafica bm.hardlimit.com


Test#3 (Generic, ZIP) [points vs time]

grafica bm.hardlimit.com


Test#1 (Memory) [points vs time]

grafica bm.hardlimit.com

Multithread performance graph
Multithread graphs measure the performance against a heavy load during certain time.

If CPU's TDP doesn't limit the frequency and the machine is properly cooled, performance should remain steady vs time. Otherwise, the performance score will oscillate or decrease over time.


Test#1 (Integers) [points vs time]

grafica bm.hardlimit.com


Test#2 (FP) [points vs time]

grafica bm.hardlimit.com


Test#3 (Generic, ZIP) [points vs time]

grafica bm.hardlimit.com


Test#1 (Memory) [points vs time]

grafica bm.hardlimit.com

Hardlimit Benchmark Central - Ver. 3.11.4