Why AMD’s Bad Benchmarks Are BAD! Investigating The Lie

Summary

The video script discusses AMD's Zen3 architecture, comparing its gaming performance to Intel's newer Raptor Lake CPUs. It criticizes AMD for misleading benchmarks using an entry-level GPU, which mask the true performance differences. The script also reviews AMD's pricing strategies and the introduction of the 5900 XT and 5800 XT processors. The host, Steve, retests these CPUs with a more powerful GPU, revealing significant performance gaps in favor of Intel. The video concludes by emphasizing the importance of accurate CPU benchmarking and AMD's marketing missteps.

Takeaways

• 😀 AMD's Zen3 architecture, introduced in 2019, is claimed to be faster for gaming than Intel's Raptor Lake architecture for 13th and 14th generation Core processors.
• 📊 AMD presented graphs showing the Ryzen 9 5950X (previously 5900 XT) outperforming Intel's Core i7 13700K by up to 4% in some scenarios.
• 💰 The Ryzen 9 5950X is positioned as a more cost-effective option compared to the Intel Core i7 13700K, which costs at least $330 USD, while the 5950X is priced at $360 USD.
• 🤔 AMD initially announced a price for the upcoming 5950X, but later retracted it, possibly due to realizing the impracticality of announcing pricing for unreleased products.
• 🐻 The video criticizes AMD's marketing tactics, accusing them of misleading consumers with performance claims that do not hold up under scrutiny.
• 🎮 The video discusses the use of GPU-limited CPU tests, which can obscure the true performance capabilities of CPUs, especially when paired with older GPUs like the Radeon RX 6600.
• 🕹️ The script mentions several games used for benchmarking, including 'Party Animals' and 'Cyberpunk 2077', but criticizes the choice of games and testing conditions as not being representative of real-world CPU performance.
• 📉 The video demonstrates that when tested with a more powerful GPU like the Radeon RX 7900 XT, Intel's processors outperform AMD's Zen3 CPUs by a significant margin.
• 💡 The script emphasizes the importance of using appropriate benchmarking methods to accurately assess CPU performance, particularly in gaming scenarios.
• 🛒 The video suggests that AMD's new 5950X might make sense for productivity tasks if priced competitively, but for gaming, other options like the Ryzen 7 5800X3D could offer better value.
• 📝 The author expresses frustration with AMD's marketing approach, which necessitates additional benchmarking to correct misconceptions about CPU performance.

Q & A

What is the main topic of the video script?

The main topic of the video script is a critique of AMD's marketing claims regarding the performance of their Zen 3 processors compared to Intel's Core processors in gaming scenarios.

When was AMD's Zen 3 architecture released?

AMD's Zen 3 architecture was released back in late 2019.

What is the issue with AMD's performance claims for their Zen 3 processors?

The issue is that AMD's performance claims for their Zen 3 processors are considered misleading because they were based on tests that heavily GPU limited the CPUs, using an older Radeon RX 6600 GPU, which does not accurately represent real-world gaming performance.

What is the significance of the Ryzen 9 5900 XT in the script?

The Ryzen 9 5900 XT is mentioned as an example of AMD's Zen 3 processor that was claimed to be competitive with Intel's Core i7 13700 K, but the script argues that this claim is misleading due to the testing methodology used by AMD.

Why is the Radeon RX 6600 not suitable for CPU benchmarking in this context?

The Radeon RX 6600 is not suitable for CPU benchmarking in this context because it is an older, entry-level GPU that can create a heavy GPU bottleneck, thus not accurately reflecting the performance capabilities of modern CPUs in gaming scenarios.

What is the role of the game 'Party Animals' in the script?

The game 'Party Animals' is used as an example of a game that is not CPU demanding and therefore not suitable for CPU benchmarking, as it does not provide meaningful data about CPU performance in gaming.

What is the criticism regarding AMD's pricing strategy for the 5800 XT?

The criticism is that AMD initially announced the 5800 XT at a price of $250 US, which was later walked back, and the script suggests that this price point was not justified given the minimal performance increase over the 5800 X.

What is the purpose of the sponsor spot featuring Thermogrizzly's Cryos Sheet Graphene Thermal Pads?

The purpose of the sponsor spot is to promote Thermogrizzly's Cryos Sheet Graphene Thermal Pads as an alternative to thermal pastes, highlighting their benefits such as high thermal conductivity and durability without the risk of drying out.

What is the script's stance on benchmarking CPUs with low-end GPUs?

The script's stance is that benchmarking CPUs with low-end GPUs is a flawed approach that can mislead consumers about the true performance capabilities of CPUs in gaming scenarios.

What is the conclusion of the video script regarding AMD's Zen 3 processors and their gaming performance?

The conclusion is that AMD's Zen 3 processors, as represented by the 5900 XT and 5800 XT, do not perform as well in gaming as claimed by AMD when tested with more demanding games and a higher-end GPU like the Radeon RX 7900 XT.

Outlines

00:00

🤖 AMD Zen3 vs. Intel Raptor Lake Gaming Performance

This paragraph discusses AMD's claim that their Zen3 architecture, originally released in 2019, still outperforms Intel's Raptor Lake architecture for gaming. AMD presented graphs showing the Ryzen 9 550x XT, which is a rebranded 5900 XT, outperforming Intel's Core i7 13700 K by up to 4% in gaming. The paragraph also touches on the pricing of these processors, with AMD's initially announced prices being walked back, suggesting a potential for more competitive pricing with the upcoming 5500 XT. Additionally, the script mentions a sponsorship by Thermal Grizzly for their Cryos sheet graphene thermal pads, which are an alternative to thermal pastes with the advantage of not degrading over time.

05:01

🎮 Debunking AMD's Gaming Benchmarks with GPU Bottlenecks

The second paragraph critiques AMD's methodology in benchmarking their new Zen3 processors against Intel's offerings. AMD paired their CPUs with the Radeon RX 6600, an older GPU, to create a heavy GPU bottleneck, which the script argues is an inaccurate way to measure CPU performance in games. The script then presents retested data using a more current GPU, the Radeon RX 7900 XT, across a variety of games, showing significant performance differences between AMD's and Intel's processors that were not evident in AMD's original tests. The paragraph also highlights the use of certain games that are not CPU-intensive, further questioning AMD's benchmark choices.

10:04

📊 A Deeper Dive into AMD's Benchmarking Methods and Results

This paragraph continues the analysis of AMD's benchmarking methods, pointing out the discrepancies between their claims and the script's own testing results. It discusses the use of the RX 6600 as a limiting factor in the tests and the subsequent retesting with the RX 7900 XT, which revealed more significant performance advantages for Intel's processors. The script also addresses AMD's claims of small performance advantages in certain games, which are contradicted by the retest results. The paragraph emphasizes the importance of using more demanding games and appropriate hardware for accurate CPU benchmarking.

15:05

🚫 The Flaws of AMD's Benchmarking Approach and Future Testing

The final paragraph concludes the critique of AMD's benchmarking approach, arguing that testing CPUs with low-end GPUs is misleading and does not provide useful information about real-world gaming performance. It suggests that AMD's marketing strategy was unnecessary and potentially damaging to their reputation. The script also expresses the author's reluctance to benchmark the new CPUs upon release due to the obvious performance differences already established. However, the author commits to providing benchmarks to clarify the situation for viewers. The paragraph ends with a call to action for viewers to support the channel and a sign-off from the host, Steve.

Keywords

💡Zen3 Architecture

Zen3 Architecture refers to the microarchitecture design by AMD, first introduced in 2019, which is the basis for their Ryzen 5000 series processors. In the video, it is highlighted that despite being older, Zen3 performs better in gaming scenarios compared to Intel's newer Raptor Lake architecture, as demonstrated through various benchmarks.

💡Ryzen 9 5900 XT

Ryzen 9 5900 XT is a high-end processor from AMD's Ryzen 5000 series, which is mentioned in the script as being compared with Intel's 13700 K. The video discusses its gaming performance and how it fares against its competitors, emphasizing the processor's continued relevance despite newer models being available.

💡Core i7 13700 K

Core i7 13700 K is a processor from Intel's 13th generation Core series, which is used as a comparison point in the video against AMD's Zen3 architecture. The term is significant as it represents Intel's attempt to compete with AMD in gaming performance, a claim that the video scrutinizes.

💡GPU Bottleneck

A GPU bottleneck occurs when the graphics processing unit (GPU) cannot keep up with the processing demands of the central processing unit (CPU), effectively limiting the overall performance in gaming or other graphically intensive tasks. The video criticizes AMD for using a GPU that creates a bottleneck in their benchmarks, which can misrepresent CPU performance.

💡Thermog Grizzly

Thermog Grizzly is a brand mentioned in the video that produces thermal pads and pastes for CPU cooling solutions. The video discusses their Cryos sheet graphene thermal pads as an alternative to traditional thermal pastes, emphasizing their durability and high thermal conductivity.

💡Zen 5 Processor Series

Zen 5 Processor Series is AMD's upcoming generation of processors, which were announced during a trade show and are part of the video's discussion on AMD's marketing practices. The series represents a significant advancement in AMD's processor technology, moving beyond the Zen3 architecture.

💡Benchmarking

Benchmarking in the context of the video refers to the process of testing and comparing the performance of different processors using specific software or games. The script criticizes certain benchmarking practices, particularly when they are conducted under conditions that do not accurately represent real-world usage or performance.

💡Radeon RX 6600

Radeon RX 6600 is an entry-level graphics card from AMD, which is used in the video's benchmarks to illustrate the concept of GPU bottlenecks. The video argues that using this particular GPU for testing can lead to misleading conclusions about CPU performance.

💡Productivity CPU

A productivity CPU is a term used to describe a processor that is optimized for tasks beyond gaming, such as video editing, 3D rendering, and other compute-intensive applications. The video suggests that the 5900 XT might be a viable option for a productivity CPU if it is priced competitively.

💡Ryzen 7 7800X 3D

Ryzen 7 7800X 3D is a specific model of AMD's processor, which includes 3D V-Cache technology to enhance performance. The video mentions this processor to highlight that even with the latest technology, certain benchmarking conditions can still result in misleading performance comparisons.

💡Marketing Blunder

A marketing blunder refers to a mistake or misstep in a company's promotional or advertising efforts that can lead to negative consequences or misrepresentation of a product. The video uses this term to describe AMD's approach to showcasing the performance of their new processors, suggesting it was both embarrassing and unnecessary.

Highlights

AMD's Zen3 architecture, released in 2019, is claimed to be faster for gaming than Intel's Raptor Lake architecture in their 13th and 14th generation Core series processors.

AMD provided graphs showing Ryzen 9 5900 XT outperforming Intel Core i7-13700K in certain scenarios by up to 4%.

The price of the Ryzen 9 5900 XT is $360, compared to the Intel Core i7-13700K at $330, suggesting potential cost-effectiveness for AMD's processor.

AMD initially announced a $359 price for the upcoming Ryzen 9 5900 XT but retracted this, possibly due to strategic pricing considerations.

Thermog Grizzly's Cryos sheet graphite thermal pads are highlighted as an alternative to thermal pastes, offering high thermal conductivity without degradation over time.

The video discusses AMD's misleading marketing practices, particularly regarding the performance claims of their Zen 5 processor series and the 5900 XT and 5800 XT CPUs.

AMD's use of the Radeon RX 6600 for CPU testing is criticized for creating a GPU bottleneck that does not accurately reflect CPU performance in gaming.

The video includes testing of a range of CPUs with the Radeon RX 7900 XT to provide a more realistic assessment of CPU performance in gaming.

Games such as 'Party Animals' and 'NRAA Blade Point' are shown to be poor choices for CPU benchmarking due to their low CPU demand and heavy GPU dependency.

The video demonstrates that AMD's claims of the 5800 XT being 12% faster than Intel's 13600 K in 'Cyberpunk 2077' are not supported by more rigorous testing conditions.

The use of the RX 6600 in testing is shown to be inadequate for accurately benchmarking modern CPUs, with the Intel 13600 K outperforming AMD's CPUs by a significant margin in 'Forspoken'.

The video argues against the practice of benchmarking CPUs with low-end GPUs, stating that it can mislead consumers about actual gaming performance.

AMD's benchmarking methods are criticized for being 'bad and unnecessary', with the suggestion that they should focus on pricing and value rather than performance claims.

The video concludes that AMD's Zen3 processors, despite being older, do not offer better gaming performance than Intel's latest architectures, contrary to AMD's claims.

The Ryzen 7 7800X 3D is shown to not have a performance advantage over the Core i3-12100 when tested with the RX 6600, highlighting the importance of accurate benchmarking.

The host, Steve, expresses frustration with AMD's marketing approach and the need to benchmark the new CPUs to correct misconceptions about their performance.

The video encourages viewers to subscribe and join the channel's Discord server for exclusive content and live streams.

Transcripts

[Music]

did you know that amd's old zen3

architecture released back in late 2019

is actually faster for gaming than

Intel's rapake architecture used for

their 13th and 14th generation core

series processors yeah now it is though

and AMD has some graphs to prove it have

a look at these see the ryzen 9 550x

which is now also a 5900 XT is worst

case a core i7 13700 K and best case

it's like 4% faster that's pretty

awesome right well hell yeah that's

awesome the 13700 Cas still costs at

least $330 us while the 550x is $360 us

therefore we can probably assume that

the upcoming $500 XT will be even

cheaper maybe not sure AMD did initially

say that it was going to be

$359 but then they quickly walked that

back probably after realizing that

announcing pricing for upcoming

processes while as high as a is a bad

idea if only the radon division was that

measured AMD also announced the 5800 XT

at $250 us but they also walked that one

back again they were obviously as high

as a kite when they came up with that

price that's the only explanation for

why they would charge 25% more than the

5800 X for a mere 2% increase in clock

speed that's okay though we've all done

it and it's why we're going to allow AMD

to come back at a later date with a less

batshit crazy price but before they do

today's sponsor spot is brought to you

by thermog Grizzly and their cryos sheet

graphine thermal pads which are an

excellent alternative to Thermal pastes

they offer very high thermal

conductivity with no liquid components

so they can't dry out and therefore

don't degrade over time like pastes and

even liquid Metals Crow sheet is very

easy to use it's extremely durable and

is available in a range of sizes to suit

most applications I've personally done

some high-end GPU testing with Crow

sheet and the results were impressive

very similar and back to that of liquid

metal but without the mess and of course

no risk of drying out so for more

information please check the link in the

video

description okay so if you watched our

recent Q&A Series where we pulled AMD up

on their insanely misleading an

anti-consumer marketing BS you might be

thinking Steve stop flogging that dead

horse we get it AMD was naughty and to

that I say that dead horse we're going

to flog it some more and it's going to

be educational now if you weren't

fortunate enough to have watch our Q&A

Series this month let me quickly fill in

the blanks during the computex trade

show a few weeks back AMD announced

their upcoming Zen 5 processor series

and in a press deck that was sent to

Media they also announced new but not

really new zen3 processors basically bin

versions of silicon that they were

already selling which they called the

5900 XT and 5800 XT these CPUs on their

own are fine they're not exciting we

don't need need them but whatever sure

AMD release them the problem is AMD

included two performance slides for

these CPUs which I've already shown you

and as you saw they claimed that these

Bend zen3 parts were as fast or really

slightly faster for gaming than Intel's

core i7 13700 K and core i5 at 13600 KF

and that's a whopping big lie in order

to achieve this deception AMD heavily

GPU limited the CPU tests pairing all

CPUs with the radon RX 6600 an

entry-level previous generation GPU

that's 3 years old now essentially what

this does is level the playing field

neutralizing any potential performance

difference between these CPUs resulting

in nothing more than an RX 6600

Benchmark benchmarking CPUs with a heavy

GPU bottleneck is a bad idea at the best

of times as it tells you nothing about

how the CPU really performs in games and

it's made even worse when you don't give

basic information like what the average

frame rate was for all we know the RX

6600 could have been rendering just

30fps so I've decided to look into what

exactly was going on but more than that

I've included some data with a more

reasonable GPU the radon RX 7900 XT

which might be a highend product right

now but by early next year we have it on

good authority that this is going to be

mid-tier

performance so for this testing I'll be

using the games that AMD bench marked

with at 1080p using the high or highest

quality preset but with upscaling and

any R tracing disabled AMD has included

an odd range of games here such as party

animals which I had to purchase on Steam

specifically for this video I promise I

didn't wasn't playing party animals

before this video but not to worry my

11-year-old daughter now is and she

quite enjoys it I've also gone and

tested a dozen very different CPUs as I

believe this illustrates very well just

how silly testing CPUs with a strong GPU

bottleneck is so let's get into it okay

let's start with everyone's favorite

video game party

animals using the radon RX 6600 which

probably unsurprisingly is powerful

enough to deliver High refresh rate

light performance in this title so I

guess in that sense AMD hasn't been too

dodgy here except for the fact that this

is a terrible game for CPU benchmarking

as it barely uses the CP

and the RX 6600 is still a very strong

bottleneck for any relatively modern SL

highend CPU now AMD didn't test the 5900

XT with party animals but they did with

the 5800 XT and they claimed that the

5800 XT was 2% faster than the core i5

13600 K but in our testing we found the

core i5 to be 2% faster maybe it's the

2% boost of core clocks that helped the

5800 XT here but also I don't suspect

that'll be enough to pull it ahead of

the core i5 also if we look at the 1%

lows which AMD didn't test or at least

they didn't provide the data for the

core i5 is 6%

faster but what if we retest party

animals with the Radeon RX 7900 XT while

doing so changes quite a bit and now the

13600 K is 133% faster than the 5800 X

When comparing the average frame rate

this is still a poor title for testing

the gaming performance of CPUs because

as I said it's just not very CPU

demanding but at least this more CPU

limited data provides us with a bit more

insight into how these CPUs really

perform nraa blade point is another game

that I haven't used for testing before

but this one seems a bit more legitimate

as a CPU Benchmark that is assuming that

you're not using a radon RX 6600

seriously though this one is super dodgy

by AMD using an RX 6600 you're about as

GPU bound as you could possibly be even

with a ryzon 5 2600 it's able to match

the 7800x 3D and 14900 K so this data is

extremely useless and it blows my mind

how often we see requests for this sort

of testing even with the 1700 XT

installed the game is still for the most

part GPU limited so nraa blade point

doesn't appear to be a great CPU

Benchmark oddly AMD used this game to

claim that the 5900 XT is 4% faster than

the 13700 K which is really odd because

with an RX 6600 there's simply no chance

that's true and with a faster GPU like

the 17900 XT the Intel processor is

actually a few per faster but yeah not a

great CPU Benchmark this one Tony 10 is

Wonderland is a game that we have used

in the past for CPU benchmarking but

then quickly dropped it after it became

apparent that it just wasn't demanding

enough for testing modern processors and

that's especially true when using an RX

6600 AMD did claim no performance

difference between the upcoming 5900 XT

and 13700 K in this title and we see

that's certainly true here again when

using a previous generation entry level

GPU but even with the 7900 XT installed

we see that Tiny Tina's wonderlands just

isn't that CPU demanding with no real

performance drop off until we get right

down to the super old ryzen 5 2600 so

again AMD used another game that doesn't

really stress the CPU for a CPU

benchmark and then proceeded to ensure

that it was as GPU limited as possible

by using an RX

6600 okay cyberpunk 2077 finally we have

a game that is actually quite demanding

on the CPU or really very demanding or

at least it would be when going above 60

FPS we're not even using the ultra

preset here as that hammered the frame

rate down into the 40s so with the RX

6600 I decided to drop the quality

preset to high and even then we're left

with totally useless data again oddly

AMD has claimed under these conditions

that the 5800 XT and I can't believe I'm

saying this is 12% faster than the 13600

K what the actual fbom AMD how could

that be even remotely true I'd love to

know what settings they're using and how

they're testing the game to make the

5800 XT 12% faster than a 13600 K that

is pretty wild stuff and making amd's

12% claim even more insane are the 7900

XT results here the 13600 K is 27%

faster than the 5800 X and 35% faster

we're looking at the 1% lows granted we

are testing the Intel CPUs with

high-speed ddr5 memory but even with the

slowest possible ddr4 memory you could

find the 13600 K should still be able to

beat the 5800 X in fact with a GPU such

as the 1700 XT memory speed really

shouldn't be that crucial and really

this shouldn't be shocking information

21 it would be crazy if zen3 was as fast

or faster than Intel's latest generation

architecture that competes very well

with Zen

4 now for testing f122 I dropped down to

the high preset which disables RAID

tracing and allows older gpus like the

rx600 to render pretty high frame rates

at 1080p we see that only the ryzen 5

2600 drops off here the rest of the pack

is heavily GPU limited so once again

we're not really learning anything here

in regards to CPU performance AMD

claimed to have a 1% Advantage for both

matchups but we're not really seeing

that here okay so retesting with the

17900 XT changed things quite a bit the

13700 K is now 28% faster than the 5950

X while the 13600 K is 177% faster than

the 5800 X so it's a bit odd that AMD

would claim to have a small performance

advantage in this title when that's not

at all the case finally we have shut off

the Tomb Raider and I'm going to presume

that AMD was using the built-in

Benchmark for this one which is a

terrible CPU test it's really just a GPU

Benchmark so we're testing ingame in the

village section which is substantially

more CPU demanding though you wouldn't

really know it with a radon RX 6600 as

all highend CPUs were limited to the

same 96 to 97

FPS again AMD claimed the upcoming 5900

XT which is really just a 5950 X is 1%

faster than the 13700 K in this title so

let's take a bit of a closer look at

that really the margin isn't that huge

here the 13700 K is 7% faster or 12%

faster if you compare the 1% lows the

5950 X actually does really well

relative to the 5800 X but yeah the

results are miles of what AMD claimed so

in conclusion you could say that the

5800 X and 13600 K are identical for

gaming if you wanted to grossly

misrepresent the gaming performance of

Intel's processor and if you go with a

radon RX 6600 using an odd batch of six

games this is what you get Intel wins by

a perc not AMD winning by 2% mind you

like they claimed but if you Benchmark

these CPUs using an odd batch of six

games but with a 7900 XT the Intel CPU

is 133% faster on average we've shown

using more demanding games that are

probably more suited to CPU Benchmark

and the average is more like 28% in

Intel's favor but even the 133% we're

seeing here is very different to what

AMD claimed now for the 595x versus

13700 K using the RX 6600 and again an

odd batch of six games they deliver the

same performance then with the 7900 XT

installed the 13700 K was on average 16%

faster not quite the 36% we see with the

RTX 490 in more demanding games but

again very different to what AMD showed

us so there you have it amd's bad

benchmarks are indeed bad and frankly

unnecessary AMD should have just

announced the 5900 XT and 5800 XT and

left it at that there's no need to show

gaming performance for zen3 processors

that we've had for like 3 years now

everyone knows what they are and without

a hefty price cut they're really not

worth buying for gaming the 5900 XT it

might make sense for a productivity CPU

assuming it is much cheaper than the

550x and you're already on the am4

platform but for gaming surely the 5700

x3d for $200 makes way more sense than

the 5800 XT as for benchmarking CPUs

with low-end gpus I hope we're starting

to make some Headway here with viewers

who believe testing with an RTX 490 at

1080p is misleading inaccurate or

whatever else they come up with the idea

is to see how many frames each processor

can throughput allowing you to compare

their performance and work out which one

of offers you the best value at a given

price point I get the idea of testing

with a more realistic GPU it makes sense

on the surface but it is a deeply flawed

approach that tells you nothing useful

and really if anything it only serves to

mislead pretending that a ryzen 75800 x

is just as fast as a core i733 700k for

gaming might make you feel good about

that ryzen processor but outside of GPU

limited gaming it's simply not true we

also found that the ryzen 7 7800x 3D was

no faster than the core I3 12100 when

using the RX

6600 but I'm sure you'll find the ryzen

7 processor is indeed much faster for

gaming and it won't take you too long to

discover this when you're actually

gaming anyway this is like my third

video on this subject so for those of

you who are yet to be convinced I doubt

I got you with this one as for AMD this

was an embarrassing and unnecessary

marketing blunder and I'm most annoyed

about the fact that I now have have to

Benchmark these CPUs when they're

released ideally I'd kind of just like

to ignore them call them what they are

the 5900 XT is a 5950 X and the 5800 XT

is a 5800 X but now I'll have to provide

you some benchmarks to prove the obvious

so thanks for that one AMD anyway I hope

you enjoyed these RX 6600 benchmarks I

know I did and don't forget to like

subscribe do all that stuff and if you'd

like to support the channel more

directly get some cool perks in return

we have float plane P

signing up to either one of those things

we'll give you access to our exclusive

Discord server monthly live streams

behind the scenes content and some Q&A

stuff so check that out if you're

interested but if not that is perfectly

fine and I would like to thank you for

watching this video I'm your host Steve

see you again next time

[Music]