Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

twitter.com

Post reply on HN
Print view View original

61–70 of 139 posts

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#61

Post by tails4e »

AlexandrB wrote: post #9

> SUPERIOR PRODUCTS LOSE TO SUPERIOR DISTRIBUTION LOCK-INS & GTM. This takeaway was a little odd to me in the context of 2008. I had been an AMD stalwart in my PCs since about 2000 (Athlon Thunderbird), but IIRC in 2008 Intel had the better processor. Better single core performance, better performance/watt, and I think AMD processors tended to have stability issues around this time. I remember I built a PC in 2009 wi…

Agree. It took a truly superior product at lower cost to make a dent in Intel's dominance in server, all the while Intel tried their best to flex their lock in muscle.

That happened well after 2008, with the advent of Zen and chiplet bases tech and better perf/W

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#62

Post by diamond559 »

belter wrote: post #15
basilgohar wrote: post #7

I love this insider view into this interesting point in computing history, especially about AMD. However, I was a little put off by the glorification of nVidia's shady practices and lock-in policies as key to their current leading position. While technically true, I dislike "ends justify the means"-style thinking. All this as the OP glorifies AMD's engineering and grit-based culture to drive through all though tough…

On the GPU area AMD lost, and will continue to lose to Nvidia, because they don't seem to get a grip on Software and Drivers. And that does not bode well for their long time CEO.

The more this is blindly repeated the more you know it's bs

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#63

Post by lotsofpulp »

polymatter wrote: post #50
lotsofpulp wrote: post #25

Earlier quoted context omitted.

I have been reading this comment since early 2000s. If I was old enough, I probably would have heard this comment being made in the late 1990s. As well as the 1980s. And probably before that too.

That doesn’t make it less true! It’s a bet because it’s risky “capital is at risk”,“value of investments can go down as well as up” etc. As opposed to a savings account which is far far less risky enough that it’s not really a bet.

That use of bet would make it a meaningless comment.

Presumably, elzbardico’s use of “bet” meant something akin to betting in a casino or lottery, where the goal is to get high from the rush of sudden, big, improbable wins.

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#64

Post by quesera »

polymatter wrote: post #50
lotsofpulp wrote: post #25

Earlier quoted context omitted.

I have been reading this comment since early 2000s. If I was old enough, I probably would have heard this comment being made in the late 1990s. As well as the 1980s. And probably before that too.

That doesn’t make it less true! It’s a bet because it’s risky “capital is at risk”,“value of investments can go down as well as up” etc. As opposed to a savings account which is far far less risky enough that it’s not really a bet.

No, it's just wordplay.

Capital is always at risk in financial investments.

If there is a semantic difference, I'd say you "invest" when you have a historical expectation of future positive returns, and you "bet" when you're taking a contrarian approach or just going with a gut feeling when data isn't available or known.

Anecdotally, and personally, I've had better luck with "bets" than "investments". But they're fundamentally the same thing.

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#65

Post by Zambyte »

> I seriously wish Nvidia and AMD could merge now – a technology cross-licensing that takes advantages of each other’s fab capabilities is going to help a lot in bringing the cost of GPU cycles down much further!

It's interesting that they see such a monopoly as something that would bring costs down. It seems more to me like competing with AMD does much more to keep Nvidias costs down (if they can be described as "down") than combining resources would.

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#66

Post by CoolGuySteve »

woooooo wrote: post #33
basilgohar wrote: post #7

I love this insider view into this interesting point in computing history, especially about AMD. However, I was a little put off by the glorification of nVidia's shady practices and lock-in policies as key to their current leading position. While technically true, I dislike "ends justify the means"-style thinking. All this as the OP glorifies AMD's engineering and grit-based culture to drive through all though tough…

Nvidia was pushing CUDA forward for over a decade before it started getting serious commercial traction. It's not like they blocked anyone else from developing viable GPGPU tech, they were just the only ones pushing it. For like 8 years their drivers on Linux were a nightmare and AMD could have come in and done better.

AMD and Apple tried to push OpenCL but the design of it, a C-like kernel compiled to the GPU with LLVM and managed by the Khronos consortium, tended to lag in absolute performance to CUDA which was able to take advantage of evolutions in GPU design more closely.

Nowadays almost nobody cares about OpenCL.

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#67

Post by badsectoracula »

AlexandrB wrote: post #42
luyu_wu wrote: post #38

Earlier quoted context omitted.

The obvious issue with both your points is that NVidia's competitors did do as such. AMD has had workable Linux drivers for many years now and there were numerous alternatives to CUDA pushed.

A common talking point is that CUDA is a formidable moat for Nvidia, but - as someone who has never done AI dev - I'm curious to understand what makes CUDA so sticky. From an outsider perspective it looks like a re-run of DirectX vs. everything else but AI is not like gaming and end users often don't have to run the model themselves. So it seems like the network effects should be less than that for a graphics APIs.

I don't know how it is nowadays but i remember trying CUDA back when GeForce GTX 280 was still a high end GPU. I didn't do anything fancy, i just tried to write a simple raytracer to get a feel of how it'd work.

The experience was incredibly simple: write C like usual but annotate a few C functions with some extra keywords and compile using a custom frontend/preprocessor/whatever-nvcc-was instead of gcc (i was on Linux - and BTW i heavily contest the notion that Nvidia drivers on Linux were "nightmare", they always worked just fine with both performance and features comparable to their Windows counterparts while ATi/AMD had buggy and broken drivers for years). Again, the experience was very simple, i even just copy/pasted a bunch of existing C code i had and it worked.

Later i tried to use OpenCL which was supposedly the open alternative. That one felt way more primitive and low level, like writing shaders without the shading bits.

In a way, as you wrote, it was kinda like DirectX: that is, CUDA was like using OpenGL 1.1 with its convenient and straightforward C API and OpenCL was like using DirectX 3 with its COM infested execute buffer nonsense.

After that i never really used CUDA (or OpenCL for that matter) but it gave me the impression that Nvidia did put way more effort on developer experience.

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#68

Post by deredede »

AlexandrB wrote: post #42
luyu_wu wrote: post #38

Earlier quoted context omitted.

The obvious issue with both your points is that NVidia's competitors did do as such. AMD has had workable Linux drivers for many years now and there were numerous alternatives to CUDA pushed.

A common talking point is that CUDA is a formidable moat for Nvidia, but - as someone who has never done AI dev - I'm curious to understand what makes CUDA so sticky. From an outsider perspective it looks like a re-run of DirectX vs. everything else but AI is not like gaming and end users often don't have to run the model themselves. So it seems like the network effects should be less than that for a graphics APIs.

The difference is not just about APIs; CUDA has a single source file model that is dead easy to use whereas last I checked every competitor still had an outdated manual loading process that adds significant friction.

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#69

Post by prewett »

polymatter wrote: post #50
lotsofpulp wrote: post #25

Earlier quoted context omitted.

I have been reading this comment since early 2000s. If I was old enough, I probably would have heard this comment being made in the late 1990s. As well as the 1980s. And probably before that too.

That doesn’t make it less true! It’s a bet because it’s risky “capital is at risk”,“value of investments can go down as well as up” etc. As opposed to a savings account which is far far less risky enough that it’s not really a bet.

Pedantically, any storage of money is a bet, because it could change in value. However, to the Buffett-style investor, you think about whether you want to buy the entire company, even if you can only afford one billionth of it. You look at a reasonable projection of earnings growth--and don't buy companies that are unpredictable (like early stage tech companies). You try to buy at a discount ("margin of safety") in case you are wrong in some fashion. And so forth.

So for example, Coca-Cola (KO) is pretty predictable. Absent any major blunders by management, KO is going to grow roughly the size of the economy, and it's going to put out 3% a year in dividends. So the fair market price of KO is reasonably determinable, and you wait until you can buy it at or less than it's fair price.

This is usually contrasted against technical traders, momentum traders, etc., who are not investing in the fundamentals of the business and assuming the price will follow good fundamentals, but rather they are betting on how the price will change.

So "investing" is seen as buying fundamentals and "betting" (or "gambling") is seen as buying on expected price changes.

Re: I was at AMD in the mid-late 2000s helping design CPU/APU/GPUs

#70

Post by kkielhofner »

disgruntledphd2 wrote: post #45
AlexandrB wrote: post #42

Earlier quoted context omitted.

A common talking point is that CUDA is a formidable moat for Nvidia, but - as someone who has never done AI dev - I'm curious to understand what makes CUDA so sticky. From an outsider perspective it looks like a re-run of DirectX vs. everything else but AI is not like gaming and end users often don't have to run the model themselves. So it seems like the network effects should be less than that for a graphics APIs.

Nvidia have invested a lot in CUDA, and they have C & Fortran bindings for a lot of scientific stuff, apart from all the DL/Gen AI stuff that's super hot right now. Like, I started using CUDA (through frameworks) over ten years ago, and basically nobody has come up with anything competitive since then.

> Nvidia have invested a lot in CUDA,

This is a significant understatement. For quite some time Jensen has been saying repeatedly that 30% of their R&D spend is on software. With the money-printing machine that is Nvidia if that holds they're going to continue to rocket ahead of competitors in terms of delivering actual solutions.

The "What are you talking about? AMD/Intel runs torch just fine!" crowd clearly haven't seen things like RIVA, Deepstream, Nemo, Triton Inference Server/NIM, etc. Meanwhile AMD (ROCm) still struggles with flash attention...

What these hardware-first (only?) companies like AMD don't seem to understand is that people buy solutions, not GPUs. It just so happens that GPUs are the best way to run these kinds of workloads but if you don't have a wholistic and exhaustive overall ecosystem you end up in single digit market share vs Nvidia at ~90%.

Post reply on HN