AMD offers the assurance of 'ready to game out of the box at 1080p' performance levels. 1080p remains the most common resolution among gamers by far, as indicated by the latest Steam Hardware Survey. However, it's important to note that these chips are not tailored for high-end AAA gaming at 1080p. Users looking to play more graphically demanding games should be prepared to adjust the quality settings accordingly or opt to purchase a discrete graphics card – though if you're planning on using a discrete card from the start, you may want to look at AMD's traditional CPUs.
I changed my Main Gaming/Office monitor from a 4K 42" gaming monitor to the 4K 32" curved (1000R) gaming monitor (in signature).
Viewing the 42" from so close (about 30 inches) had greatly damaged my vision over the last few years -- from nominally 20/20 (before starting to use 42") to 20/70 and bad double vision at times.
Over a year of eye exercises and monitor changes around the house seem to have greatly fixed my eyes. I still need to do the exercises to "cement" the changes.
One of my "pet peeves" regarding AMD competition and why I use AMD.
The competition often makes good stuff; but, the competition also often "misinforms/misleads" the public. I.e. Its not as good as they would have you believe.
Quote
Industry group invalidates 2,600 official Intel CPU benchmarks — SPEC says the company's compiler used unfair optimizations to boost performance
SPEC says it will no longer be publishing SPEC CPU 2017 results for Intel CPUs running a specific version of the Intel compiler, citing displeasure over an apparent targeted optimization for a specific workload ... that essentially amounts to cheating.
While it's expected that compilers should be optimized since more performance is obviously better, optimizing specifically for benchmarks is controversial and frowned upon. SPEC wants its benchmarks to reflect the real-world performance of hardware and to provide a standardized way to compare different processors.
AMD Zen 5 CPUs Reportedly Made On TSMC’s 3nm Process In Q2, Mass Production In Q3
The Zen 5 CPU architecture is going to be a big deal for AMD in 2024 as it will power multiple lines of CPU families including Granite Ridge "Ryzen Desktop", Strix Point "Ryzen Mobile" & Turin "EPYC Server". All of these CPU families will play a major role and it looks like these chips are entering the TSMC fabs for production next quarter followed by volume production in Q3 2024, as reported by UDN.
Intel puts 1nm process (10A) on the roadmap for 2027 — also plans for fully AI-automated factories with 'Cobots'
Intel's previously-unannounced Intel 10A (analogous to 1nm) will enter production/development in late 2027, marking the arrival of the company's first 1nm node, and its 14A (1.4nm) node will enter production in 2026. The company is also working to create fully autonomous AI-powered fabs in the future.
Intel to produce AI to run entire factories?
In other news, Nvidia CEO says no-one should study "Computer Programming" for use as a "Profession". He says AI will be doing virtually all the Programming in a few years.
It sounds like "human experts" are in the process of being replaced wholesale by AI. We'll see. But, it is what I expect to happen within the lifetime of current young folks.
Decades from now, when Humans are virtually no longer necessary, then what???
Intel CEO Says He's Betting the Entire Company's Future on its 18A Process
Intel's upcoming 18A process, short for 18 Angstroms, .. will be the culmination of many years of engineering work by Intel and billions of dollars of R&D. It is theoretically the process node that will allow Intel to leapfrog rivals TSMC and Samsung in the nanometer arms race. Intel's CEO is being extremely blunt about this node's importance to the company's future, stating that he has bet the entire company on 18A's success.
AMD Teases Next-Gen FSR “FiedlityFX Super Resolution” Upscaling Tech With AI, Following DLSS & XeSS
"2024 for us is a huge year because we have spent so many years developing hardware and software capabilities for AI. We've just completed AI enabling our entire portfolios so cloud, edge, PCs, our embedded devices, our gaming devices, we're enabling our gaming devices to upscale using AI and 2024 is really a huge deployment year for us so now the bedrock's there, the capabilities are there."
AMD has incorporated special AI circuitry in virtually all its upscale products. That includes the GPUs. Now, they plan to "turn it on" in the GPUs.
Survey Reveals AI Professionals Considering Switching From NVIDIA To AMD, Cite Instinct MI300X Performance & Cost
Jeff Tatarchuk, who is the co-founder of TensorWave, revealed that in an independent study of 82 engineers and AI professionals, around 50% of them expressed their confidence in utilizing AMD Instinct MI300X GPU, considering that it offers a better price-to-performance ratio, along with having extensive availability compared to counterparts such as NVIDIA H100s. Apart from that, Jeff also says that TensorWave will employ the MI300X AI accelerators, which is another promising news for Team Red [AMD]
For a quick rundown on the MI300X Instinct AI GPU, it is designed solely on the CDNA 3 architecture, and a lot of stuff is going on. ...Here is how it compares to NVIDIA's H100:
2.4X Higher Memory Capacity 1.6X Higher Memory Bandwidth 1.3X FP8 TFLOPS 1.3X FP16 TFLOPS Up To 20% Faster Vs H100 (Llama 2 70B) In 1v1 Comparison Up To 20% Faster Vs H100 (FlashAttention 2) in 1v1 Comparison Up To 40% Faster Vs H100 (Llama 2 70B) in 8v8 Server Up To 60% Faster Vs H100 (Bloom 176B) In 8v8 Server
Crypto Mining. Who needs an expensive GPU?? (a slight exaggeration )
Quote
AMD Ryzen 9 7950X CPU More Profitable In Mining Than Fastest GPUs As Crypto Once Again Sees Rise After Bitcoin Jump
This has also led to the rise in demand for AMD's Ryzen 7000 CPUs in certain currencies such as Qubic which takes advantage of AVX2 or AVX512 instructions which the Zen 4 architecture is very good at since it is the only client and widely available option that supports the ISA. Intel gave up on offering AVX512 support
If you search up the current mining profit rates, the Qubic (QBIC) coin generates a mining profit of around $3 US (24 hours) using the chip which includes the electricity cost when running the chip at 170W. Given the fantastic efficiency that AMD's Zen 4 CPUs have to offer and their ability to offer the same or very similar performance when down-clocked to 65-100W, the 16-core behemoth and the rest of its siblings become a viable solution for such algorithms.
With AMD Ryzen 9 7950X CPUs, you can generate almost twice as much profit in crypto mining versus using the flagship NVIDIA GeForce RTX 4090 GPU which offers a maximum profit of $1.7-$1.9 US in currencies such as PYI and NEX. The [Nvidia] GPU has a maximum power draw of 450W & also features a 3.5x higher price ($559 US vs $2000 US).
AMD Navi 48 “RDNA 4” GPU Confirmed In ROCm Patches, Coming To Radeon RX 8000 Gaming Cards This Year
One thing to note is that AMD's RDNA 4 GPU architecture will be a brand new design so we should also consider the fact that the red team can fine-tune its performance within ray tracing titles and also offer additional goodies such as an AI-infused FSR upscaling technology which rivals NVIDIA's DLSS even though FSR 3 in its current state is a very good alternative for gamers. AI acceleration can sort some of the shortcomings of FSR and bring enhanced image quality in areas where it currently lacks against DLSS.
we can expect the next-generation lineup to feature support for GDDR7 memory which will initially come in 28/32 Gbps dies & with 16Gb
Rumor and reasonable speculation regarding how AI is quickly changing CPUs (and GPUs). Designs are being altered "mid-stream" as opposed to "designs from the ground up".
Quote
Microsoft's Demand For Faster AI Capabilities Leads To Severe SOC Nerfs Across AMD & Intel Camps As NPUs Take Prioritization Over Other Aspects
We have recently seen an AI explosion in the PC segment with all chipmakers talking about the respective capabilities of their chips and platforms. The segment is driven by a range of software innovations and Microsoft's Windows Copilot which has some hefty requirements to support its AI functionality. Chipmakers are now heavily betting on the AI craze and it looks like some have gone outside of their traditional chip development plans to prioritize AI over other parts of their newest SOCs that will be coming to market later this year.
TSMC gets $6.6 billion in cash and $5 billion in loans from CHIPS Act, plans third US fab
The largest CHIPS and Science Act support package to date
The U.S. government has announced plans to provide TSMC financial support totaling $11.6 billion, comprising $6.6 billion in grants and up to $5 billion in loans. TSMC will use the funding to build three semiconductor production facilities in Arizona, with a total investment exceeding $65 billion. This is the largest financial package the U.S. government approved under the CHIPS and Science Act.
Repair shop continues to replace 200 RTX 4090 power connectors each month
As noted above, the 5000 series of Nvidia GPUs is rumored to come out by the end of the year. Presumably, melting connectors will not be an issue for the 5000 series.
9000 series CPUs coming soon, perhaps. The BIOS update is already on-line. Some speculation says Official Announce in less than 2 months. Top of the line to come in 2024. Uses same Motherboards as 7000 CPUs.
Quote
AMD Ryzen 9000 “Zen 5” Desktop CPUs Confirmed As Gigabyte Lists Official Support On Its AM5 Motherboards
The interesting thing is that Gigabyte mentions that the BIOS only provides boot-up support for the AMD Ryzen 9000 Desktop CPUs. This may still be a very early BIOS release
Not a lot about AMD's Zen 5 core architecture is known at the moment. Rumors are stating that the core-to-core performance should see a >40% performance uplift & there are changes to the cache structure too but from what the company has officially stated, it will offer:
Enhanced performance and efficiency Re-pipelined front-end and wide issue Integrated AI and Machine Learning optimizations
PC builders buying an AM5 platform now are assured drop-in-compatibility with the next-gen AMD Ryzen CPUs. The company has committed to a 2025+ plan for its AM5 socket and as we have seen on the AM4 platform, AMD supports its desktop platforms more than Intel with a very strong 7-year+ dedication to its previous platform.
The rumored 40 percent uplift in performance interests me. I just may buy one.
Analyst Forecasts $40 Billion AI GPU Sales For NVIDIA, $3.5 Billion For AMD & Just $500 Million For Intel In 2024
Bloomberg Technology reports that NVIDIA's AI accelerators are expected to generate almost $50 billion, whilst competitors like AMD and Intel will only receive a minimal portion, coming at $3.5 billion and $500 million, respectively. The information was disclosed with Intel's CEO Pat Gelsinger onboard after the announcement of Intel's latest earnings.
The future of the AI markets is indeed interesting to witness, and despite complete dominance by NVIDIA, there is still hope and optimism left for other companies as well. However, it's fascinating to see how Team Green has led itself up to the very top, and with the debut of the Blackwell AI GPU architecture, I don't see their dominant position getting broken anytime soon.
IBM to invest $730 million in Canadian semiconductor sector: Report
IBM is set to channel over $730 million in expanding its semiconductor packaging and testing plant in Bromont, Quebec, aiming to strengthen its semiconductor supply chain. This investment, spread over the next five years, will enhance the largest facility of its type in the region and is anticipated to generate 280 new skilled jobs, Bloomberg reports.
The Bromont facility, occupying 800 acres approximately 50 miles east of Montreal, is an important player in semiconductor production in North America as far as its applications are concerned. It also hosts Canada’s first universal quantum computer
)