Performance comparison#
This notebook compares the performance of cuDF and pandas. The comparisons performed are on identical data sizes. This notebook primarily showcases the factor
of speedups users can have when the similar pandas APIs are run on GPUs using cudf.
The hardware details used to run these performance comparisons are at the end of this page.
Note: This notebook is written to measure performance on NVIDIA GPUs with large memory. If running on hardware with lower memory, please consider lowering the num_rows values. Performance results may vary by data size, as well as the CPU and GPU used.
import os
import time
import timeit
from io import BytesIO
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import cudf
np.random.seed(0)
Concat, count & joins performance#
num_rows = 300_000_000
pdf = pd.DataFrame(
{
"numbers": np.random.randint(-1000, 1000, num_rows, dtype="int64"),
"business": np.random.choice(
["McD", "Buckees", "Walmart", "Costco"], size=num_rows
),
}
)
pdf
| numbers | business | |
|---|---|---|
| 0 | -316 | Costco |
| 1 | -441 | Costco |
| 2 | 653 | Buckees |
| 3 | 216 | Buckees |
| 4 | -165 | Walmart |
| ... | ... | ... |
| 299999995 | -395 | Walmart |
| 299999996 | -653 | Buckees |
| 299999997 | 364 | Buckees |
| 299999998 | 159 | Buckees |
| 299999999 | -501 | Walmart |
300000000 rows × 2 columns
gdf = cudf.from_pandas(pdf)
gdf
| numbers | business | |
|---|---|---|
| 0 | -316 | Costco |
| 1 | -441 | Costco |
| 2 | 653 | Buckees |
| 3 | 216 | Buckees |
| 4 | -165 | Walmart |
| ... | ... | ... |
| 299999995 | -395 | Walmart |
| 299999996 | -653 | Buckees |
| 299999997 | 364 | Buckees |
| 299999998 | 159 | Buckees |
| 299999999 | -501 | Walmart |
300000000 rows × 2 columns
def timeit_pandas_cudf(pd_obj, gd_obj, func, **kwargs):
"""
A utility function to measure execution time of an
API(`func`) in pandas & cudf.
Parameters
----------
pd_obj : Pandas object
gd_obj : cuDF object
func : callable
"""
pandas_time = timeit.timeit(lambda: func(pd_obj), **kwargs)
cudf_time = timeit.timeit(lambda: func(gd_obj), **kwargs)
return pandas_time, cudf_time
pandas_value_counts, cudf_value_counts = timeit_pandas_cudf(
pdf, gdf, lambda df: df.value_counts(), number=30
)
pdf = pdf.head(100_000_000)
gdf = gdf.head(100_000_000)
pandas_concat = timeit.timeit(lambda: pd.concat([pdf, pdf, pdf]), number=30)
cudf_concat = timeit.timeit(lambda: cudf.concat([gdf, gdf, gdf]), number=30)
pandas_groupby, cudf_groupby = timeit_pandas_cudf(
pdf,
gdf,
lambda df: df.groupby("business").agg(["min", "max", "mean"]),
number=30,
)
num_rows = 1_000_000
pdf = pd.DataFrame(
{
"numbers": np.random.randint(-1000, 1000, num_rows, dtype="int64"),
"business": np.random.choice(
["McD", "Buckees", "Walmart", "Costco"], size=num_rows
),
}
)
gdf = cudf.from_pandas(pdf)
pandas_merge, cudf_merge = timeit_pandas_cudf(
pdf, gdf, lambda df: df.merge(df), number=30
)
performance_df = pd.DataFrame(
{
"cudf speedup vs. pandas": [
pandas_value_counts / cudf_value_counts,
pandas_concat / cudf_concat,
pandas_groupby / cudf_groupby,
pandas_merge / cudf_merge,
],
},
index=["value_counts", "concat", "groupby", "merge"],
)
performance_df
| cudf speedup vs. pandas | |
|---|---|
| value_counts | 282.901300 |
| concat | 203.624680 |
| groupby | 138.495762 |
| merge | 136.519031 |
ax = performance_df.plot.bar(
color="#7400ff",
ylim=(1, 400),
rot=0,
xlabel="Operation",
ylabel="Speedup factor",
)
ax.bar_label(ax.containers[0], fmt="%.0f")
plt.show()
# Cleaning up used memory for later benchmarks
del pdf
del gdf
import gc
_ = gc.collect()
Strings Performance#
num_rows = 300_000_000
pd_series = pd.Series(
np.random.choice(
["123", "56.234", "Walmart", "Costco", "rapids ai"], size=num_rows
)
)
gd_series = cudf.from_pandas(pd_series)
pandas_upper, cudf_upper = timeit_pandas_cudf(
pd_series, gd_series, lambda s: s.str.upper(), number=20
)
pandas_contains, cudf_contains = timeit_pandas_cudf(
pd_series, gd_series, lambda s: s.str.contains(r"[0-9][a-z]"), number=20
)
pandas_isalpha, cudf_isalpha = timeit_pandas_cudf(
pd_series, gd_series, lambda s: s.str.isalpha(), number=20
)
performance_df = pd.DataFrame(
{
"cudf speedup vs. pandas": [
pandas_upper / cudf_upper,
pandas_contains / cudf_contains,
pandas_isalpha / cudf_isalpha,
],
},
index=["upper", "contains", "isalpha"],
)
performance_df
| cudf speedup vs. pandas | |
|---|---|
| upper | 1832.120875 |
| contains | 1311.758332 |
| is_alpha | 5752.301339 |
ax = performance_df.plot.bar(
color="#7400ff",
ylim=(1, 7000),
rot=0,
xlabel="String method",
ylabel="Speedup factor",
)
ax.bar_label(ax.containers[0], fmt="%.0f")
plt.show()
User-defined function (UDF) performance (with JIT overhead)#
The first UDF runs include JIT compilation overhead, due to which the performance of first run and average of next few runs are compared separately.
num_rows = 10_000_000
pdf_age = pd.DataFrame(
{
"age": np.random.randint(0, 100, num_rows),
}
)
pdf_age
| age | |
|---|---|
| 0 | 87 |
| 1 | 71 |
| 2 | 63 |
| 3 | 40 |
| 4 | 92 |
| ... | ... |
| 9999995 | 4 |
| 9999996 | 28 |
| 9999997 | 31 |
| 9999998 | 4 |
| 9999999 | 47 |
10000000 rows × 1 columns
gdf_age = cudf.from_pandas(pdf_age)
gdf_age
| age | |
|---|---|
| 0 | 87 |
| 1 | 71 |
| 2 | 63 |
| 3 | 40 |
| 4 | 92 |
| ... | ... |
| 9999995 | 4 |
| 9999996 | 28 |
| 9999997 | 31 |
| 9999998 | 4 |
| 9999999 | 47 |
10000000 rows × 1 columns
def age_udf(row):
if row["age"] < 18:
return 0
elif 18 <= row["age"] < 20:
return 1
elif 20 <= row["age"] < 30:
return 2
elif 30 <= row["age"] < 40:
return 3
elif 40 <= row["age"] < 50:
return 4
elif 50 <= row["age"] < 60:
return 5
elif 60 <= row["age"] < 70:
return 6
else:
return 7
pandas_int_udf, cudf_int_udf = timeit_pandas_cudf(
pdf_age, gdf_age, lambda df: df.apply(age_udf, axis=1), number=1
)
def str_isupper_udf(row):
if row.isupper():
return 0
else:
return 1
pd_series = pd.Series(
np.random.choice(["ABC", "abc", "hello world", "AI"], size=100_000_000),
name="strings",
)
pd_series
0 AI
1 ABC
2 hello world
3 abc
4 hello world
...
99999995 AI
99999996 AI
99999997 abc
99999998 abc
99999999 hello world
Name: strings, Length: 100000000, dtype: object
gd_series = cudf.from_pandas(pd_series)
gd_series
0 AI
1 ABC
2 hello world
3 abc
4 hello world
...
99999995 AI
99999996 AI
99999997 abc
99999998 abc
99999999 hello world
Name: strings, Length: 100000000, dtype: object
pandas_str_udf, cudf_str_udf = timeit_pandas_cudf(
pd_series, gd_series, lambda s: s.apply(str_isupper_udf), number=1
)
performance_df = pd.DataFrame(
{
"cudf speedup vs. pandas": [
pandas_int_udf / cudf_int_udf,
pandas_str_udf / cudf_str_udf,
]
},
index=["Numeric", "String"],
)
performance_df
| cudf speedup vs. pandas | |
|---|---|
| Numeric | 362.091673 |
| String | 204.865789 |
Below is the plot showing performance speedup in case of Numeric UDFs & String UDFs on their first runs.
ax = performance_df.plot.bar(
color="#7400ff",
ylim=(1, 550),
rot=0,
xlabel="UDF Kind",
ylabel="Speedup factor",
)
ax.bar_label(ax.containers[0], fmt="%.0f")
plt.show()
User-defined function (UDF) performance (without JIT overhead)#
pandas_int_udf, cudf_int_udf = timeit_pandas_cudf(
pdf_age, gdf_age, lambda df: df.apply(age_udf, axis=1), number=10
)
pandas_str_udf, cudf_str_udf = timeit_pandas_cudf(
pd_series, gd_series, lambda s: s.apply(str_isupper_udf), number=10
)
performance_df = pd.DataFrame(
{
"cudf speedup vs. pandas": [
pandas_int_udf / cudf_int_udf,
pandas_str_udf / cudf_str_udf,
]
},
index=["Numeric", "String"],
)
performance_df
| cudf speedup vs. pandas | |
|---|---|
| Numeric | 95448.144630 |
| String | 2587.570338 |
Below is the plot showing performance speedup in case of Numeric UDFs & String UDFs on their consequent runs. In this case the speedup is massive because of no JIT overhead present.
ax = performance_df.plot.bar(
color="#7400ff",
ylim=(1, 100000),
rot=0,
xlabel="UDF Kind",
ylabel="Speedup factor",
)
ax.bar_label(ax.containers[0], fmt="%.0f")
plt.show()
UDF Performance in GroupBy#
num_rows = 100_000_000
pdf = pd.DataFrame()
pdf["key"] = np.random.randint(0, 2, num_rows)
pdf["val"] = np.random.randint(0, 7, num_rows)
def custom_formula_udf(df):
df["out"] = df["key"] * df["val"] - 10
return df
gdf = cudf.from_pandas(pdf)
pandas_udf_groupby, cudf_udf_groupby = timeit_pandas_cudf(
pdf,
gdf,
lambda df: df.groupby(["key"], group_keys=False).apply(custom_formula_udf),
number=10,
)
performance_df = pd.DataFrame(
{"cudf speedup vs. pandas": [pandas_udf_groupby / cudf_udf_groupby]},
index=["Grouped UDF"],
)
performance_df
| cudf speedup vs. pandas | |
|---|---|
| Grouped UDF | 423.83606 |
ax = performance_df.plot.bar(
color="#7400ff", ylim=(1, 500), rot=0, ylabel="Speedup factor"
)
ax.bar_label(ax.containers[0], fmt="%.0f")
plt.show()
System Configuration#
CPU Configuration#
!lscpu
Architecture: x86_64
CPU op-mode(s): 32-bit, 64-bit
Byte Order: Little Endian
Address sizes: 46 bits physical, 48 bits virtual
CPU(s): 80
On-line CPU(s) list: 0-79
Thread(s) per core: 2
Core(s) per socket: 20
Socket(s): 2
NUMA node(s): 2
Vendor ID: GenuineIntel
CPU family: 6
Model: 85
Model name: Intel(R) Xeon(R) Gold 6230 CPU @ 2.10GHz
Stepping: 7
CPU MHz: 800.049
CPU max MHz: 3900.0000
CPU min MHz: 800.0000
BogoMIPS: 4200.00
Virtualization: VT-x
L1d cache: 1.3 MiB
L1i cache: 1.3 MiB
L2 cache: 40 MiB
L3 cache: 55 MiB
NUMA node0 CPU(s): 0-19,40-59
NUMA node1 CPU(s): 20-39,60-79
Vulnerability Itlb multihit: KVM: Mitigation: Split huge pages
Vulnerability L1tf: Not affected
Vulnerability Mds: Not affected
Vulnerability Meltdown: Not affected
Vulnerability Spec store bypass: Mitigation; Speculative Store Bypass disabled v
ia prctl and seccomp
Vulnerability Spectre v1: Mitigation; usercopy/swapgs barriers and __user
pointer sanitization
Vulnerability Spectre v2: Mitigation; Enhanced IBRS, IBPB conditional, RS
B filling
Vulnerability Srbds: Not affected
Vulnerability Tsx async abort: Mitigation; TSX disabled
Flags: fpu vme de pse tsc msr pae mce cx8 apic sep mtr
r pge mca cmov pat pse36 clflush dts acpi mmx f
xsr sse sse2 ss ht tm pbe syscall nx pdpe1gb rd
tscp lm constant_tsc art arch_perfmon pebs bts
rep_good nopl xtopology nonstop_tsc cpuid aperf
mperf pni pclmulqdq dtes64 monitor ds_cpl vmx s
mx est tm2 ssse3 sdbg fma cx16 xtpr pdcm pcid d
ca sse4_1 sse4_2 x2apic movbe popcnt tsc_deadli
ne_timer aes xsave avx f16c rdrand lahf_lm abm
3dnowprefetch cpuid_fault epb cat_l3 cdp_l3 inv
pcid_single intel_ppin ssbd mba ibrs ibpb stibp
ibrs_enhanced tpr_shadow vnmi flexpriority ept
vpid ept_ad fsgsbase tsc_adjust bmi1 avx2 smep
bmi2 erms invpcid cqm mpx rdt_a avx512f avx512
dq rdseed adx smap clflushopt clwb intel_pt avx
512cd avx512bw avx512vl xsaveopt xsavec xgetbv1
xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm
_mbm_local dtherm ida arat pln pts hwp hwp_act_
window hwp_epp hwp_pkg_req pku ospke avx512_vnn
i md_clear flush_l1d arch_capabilities
GPU Configuration#
!nvidia-smi
Mon Feb 6 17:43:52 2023
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 525.60.04 Driver Version: 525.60.04 CUDA Version: 12.0 |
|-------------------------------+----------------------+----------------------+
| GPU Name Persistence-M| Bus-Id Disp.A | Volatile Uncorr. ECC |
| Fan Temp Perf Pwr:Usage/Cap| Memory-Usage | GPU-Util Compute M. |
| | | MIG M. |
|===============================+======================+======================|
| 0 H100 80GB HBM2e On | 00000000:1E:00.0 Off | 0 |
| N/A 30C P0 60W / 700W | 0MiB / 81559MiB | 0% Default |
| | | Disabled |
+-------------------------------+----------------------+----------------------+
| 1 H100 80GB HBM2e On | 00000000:22:00.0 Off | 0 |
| N/A 30C P0 60W / 700W | 0MiB / 81559MiB | 0% Default |
| | | Disabled |
+-------------------------------+----------------------+----------------------+
+-----------------------------------------------------------------------------+
| Processes: |
| GPU GI CI PID Type Process name GPU Memory |
| ID ID Usage |
|=============================================================================|
| No running processes found |
+-----------------------------------------------------------------------------+