Search the Community

Summary As part of today’s AMD’s 2022 Financial Analyst Day, the company is offering a short update on their forthcoming Zen 4 CPU architecture. Addressing some post-Computex questions around IPC expectations, AMD is revealing that they expect Zen 4 to offer an 8-10% IPC uplift over Zen 3. AMD is projecting a >25% increase in performance-per-watt with Zen 4 over Zen 3 (based on desktop 16C chips running CineBench). Meanwhile the overall performance improvement stands at >35%, no doubt taking advantage of both the greater performance of the architecture per-thread, and AMD’s previously disclosed higher TDPs. Lastly, Zen 4 V-Cache chips are expected to come in the future too. Quotes My thoughts I think it's great for AMD to clear the air on the post Computex qualms. While 8-10% increase in IPC sounds underwhelming, greater than 35% overall performance improvement does not, IMO. Combine that with future V-Cache chips and Zen 4 will be a highly competitive product against Intel. Sources https://www.anandtech.com/show/17441/amd-zen-4-update-8-to-10-ipc-uplift-25-more-perfperwatt-vcache-chips-coming https://www.tomshardware.com/news/amd-zen-4-ryzen-7000-has-810-ipc-uplift-more-than-35-overall-performance-gain https://www.pcgamer.com/amd-provides-new-zen-4-details-and-touts-a-greater-than-25-performance-per-watt-gain/

source: https://forums.anandtech.com/threads/ryzen-strictly-technical.2500572/ "The Slit", established member over Anand's forums has done some testing that feature more technical benchmarks that you don't usually see go check the source for all the benches and results plus in depth technical details on the CPU and chipset (Zeppelin; the chipset is inside the CPU) the tests were conducted at 3.5Ghz mobo used ASUS CROSSHAIR VI HERO results with ER mean extremes removed (highest and lowest) let's talk IPC: back in 2015, AMD estimated a 40% IPC increase over Excavator, they achieved that and 6.6% over it - that's good what's not so good is that Zen's IPC is practically lower than Haswell's, a microarchitecture that's already 4y old - AMD took 5y in development to produce a CPU that's slower than an already 4y old microarchitecture [extremes removed] IPC gains: Zen over Excavator: 46.62% Zen over Haswell: -7.03% KabyLake over Haswell: 14.29% (my math, take with salt) Zen over KabyLake: -25.56% the FMA set implementation in Zen seems very problematic as tests with Linpack and Himeno shows: Himeno benchmark (evaluates performance of incompressible fluid analysis code - takes in measurements to precede major loops in solving the Poisson's equation solution using the Jacobi iteration method) shows particularly bad results: and before you start blaming the compiler, here's what "The Slit" said: GCC - GNU Compiler Collection MSVC - Microsoft Visual C++ compiler ICL - Intel C++ compiler --- AMD needs to get their house in order, because after seeing what happened with this launch, I don't see a bright future for AMD on the market Intel has nothing to fear from Zen - AMD's products for enterprise (Naples), based on the current IPC, have no reason to exist other than being a cheap alternative

I built my PC in June of last year with a GTX 960 4 gig and a FX 4350. Back when I started, It performed very well. Now though it appears to have slowed down significantly. the same games (GTA V, ROTR, The Witcher 3, etc.) perform worse with the same graphics settings. When I look at the Afterburner overlay it shows my CPU at or near 100% all the time, however when I go into task manager it shows only the game I am playing as any significant draw. This did not used to happen, so the question is: What would affect my CPU to make it perform worse in the same conditions? Full parts list: 128 gig ADATA SSD 750 gig Caviar Blue HDD Cooler Master Hyper T2 EVGA 600 Watt Bronze PSU MSI 970A-G43 Plus Mobo FX 4350 GTX 960 4 gig

Source: http://www.anandtech.com/show/11464/intel-announces-skylakex-bringing-18core-hcc-silicon-to-consumers-for-1999 If you're looking for the strangely positioned X299 4 core parts I have another thread. Welcome to the era of 12 cores and 24 threads being 'low core count' I was skeptical about those high turbo speeds in the leaks but they appear to be true. These things are absolute monsters, compared to the 6950X which tubo-ed to 3.5ghz and 4.0ghz boost this is an 800mhz and 500mhz improvement on the equivalent 10 core part. I wonder what the turbo breakdown per number of active cores is. Shame to see that 40 lanes is only available starting at $1000, last gen it was $617 6850K to get the full 40. The old lineup Comparison with enthusiast Ryzen As people predicted the i7-7820K competes directly with the Ryzen 7 1800X The High-core-count chips Little official details on the 2 flagship chips, only prices announced so far. However going by the leaks being as accurate as they are... Anandtech notes that the silicon these top 3 are harvested from appears to be a 20 core die. A response to Threadripper? Kinda. Anandtech notes that Skylake-X was only ever expected to be up to 12 cores, however as the industry started to hear rumors of Threadripper Intel had to react. This is why the 2 HCC chips don't have finalised specifications. Intel is waiting for the Threadripper specs to drop. Show me the money (Cache) Intel have tweaked their cache hieracy again, this means that like Skylake-S (read i7-77xx) the L3 cache is reduced and L2 cache increased. The L3 now acts like a pseudo-victim cache with pre-fetching. This increased lower level cache means that the cache miss rate, which slow down operations are reduced and the cache hit rate is increased. ELI5: IPC is increased over Skylake-S Anandtech predicts around 11-18% IPC increase. This is more than a recent generational leap, Broadwell --> Skylake was ~3% gain, Skylake --> Kabylake was 0% gain. Other stuff AVX-512 instructions supported. Turbo boost 3.0 supported. Anandtech predicts we'll see all products from both sides by the end of the summer. Overclocking-ness Unfortunately it looks like Intel has gone for TIM rather than solder for the IHS. Heavy stuff before 9am here in the UK, probably missed a ton or misinterpreted some stuff, let me know and I'll update it

So i am working on my technical seminar Paper and its about how Moore's law is coming to an end. While i was looking into the transistor counts of some processors i noticed that the R7 1800x has significantly more transistors than the 6950x. What i am confused about is , how does a processor(1800x has 4.8 bln transistors vs the 6950x has 3.2 bln) with significantly more transistors and higher clock rate(base 3.0 vs 3.6ghz ) fail to out perform the 6950x.Transistor count and their switching frequency are two main factors defining IPC.The architecture does play an important role in IPC but to what extent ? Can someone experienced in the field of VLSI engineering please elaborate on this? Thanks in advance. https://en.wikipedia.org/wiki/Transistor_count *MODS , please move the thread to where it needs to be if it already isn't!

I watched a couple of streamers playing League of Legends with specs of either: A GTX 1060 6GB w/ Ryzen 1600. An RX 580 8GB w/ Ryzen 1600. A GTX 1060 6GB w/ i5 4590. #1 and #2 are having an FPS of 75-110~ while #3 is having at around 110-144+. Questions: Is Ryzen's IPC that low? From my research, League of Legends only handles 2 cores or 4 threads and scales better with a higher IPC. This is quite concerning to me as I am planning to stream soon. So, if the 4590 gives higher FPS than 1600 in this specific game while streaming, does that mean it will be better if I bought an 8400 instead of 1600?(I am currently choosing between 8400 and 1600) How is the 8400's IPC compared to 4590? Thank you.

Hi everyone, Could someone plz give me a good explenation on what CPU IPC is?

How would the single core performance of a i7 first gen and i5 2450m compare ?

Hey guys I just wanted to know something about gaming requirements and processors, some games require that you have a core i5-6600k or a core i7-4790k e.g. battlefield 1. link: https://www.systemrequirementslab.com/cyri/requirements/battlefield-1/13404 The i7-4790k and i5-6600k processors are undoubtedly some really high-end processors that cost a fair bit of money, yet I see gamers playing some really demanding games (on medium to low settings) with a low end i3 8100 or even a pentium G4560 paired up with a decent graphics card like 1050ti or 1060. So my question is this: when a game says it requires a 3.5GHz to 3.9GHz core i5-6600k processor (minimum requirements) how important is that for you to match up your processor? Could you play that game with a core i5-4670? Could you maybe play that game with a core i3-7300? Or maybe even a pentium G4560? Would there be any stuttering or lag? Would there be any issues at all with your gameplay? Thank you and God bless you all.

I'm buying a new custom PC should I buy Intel i7 5820k or Intel i7 6700k. Is Skylake the enthusiast platform surpassing Haswell-e? Let me now... Thanks.

A Silicon Valley startup Soft Machines has raised over $125 million in funding from -among others- Samsung and AMD for further development of their VISC architecture that could bring 3x improvement in IPC ( single threaded performance). The architecture uses virtual cores and sounds a bit like reverse hyper-threading. http://techreport.com/news/27259/cpu-startup-claims-to-achieve-3x-ipc-gains-with-visc-architecture http://www.eetimes.com/document.asp?doc_id=1324364 A description of (another) VISC architecture from an abstract of a research paper http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=1214337

Hey all, I've been into tech only around two years ago. So I don't know much hardware before then. When I started researching and learning it was the deneb, thuban, and sandy bridge time. At the time bulldozer and ivy bridge were very close to launching. By then it was very clear Intel's core and single thread performance were years ahead of AMD's. Interestingly on forums I've see people mentioning how back in the old days AMD were the leading cpu manufacturer and dominated the cpu market. Saying how AMD were the one who came up with extreme edition processors and stuff. So can someone enlighten me with a brief summary of the cpu war back then? And perhaps how did AMD fall so far behind?

A benchmark result was found by an OCN member and reported by WCCFTECH . The 7850K was overclocked to 4.9Ghz and at that clock speed was 35% faster than a 6800K overclocked to 5.1Ghz. This puts Steamroller cores at a 36-37% performance advantage per clock over Piledriver which is mighty close to Sandy Bridge & Ivy Bridge i5 processors. As with all leaked benchmarks take it with a grain of salt. UPDATE : The image is most likely fake as it was posted on 4Chan and no longer can be found. A rough examination under MS paint reveals possible evidence of manipulation.